]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
2a4c8c883ad52cf72931b27d06b6857d45fb74de
[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 HAS_LEGACY as compat_has_legacy,
31 compat_get_terminal_size,
32 compat_os_name,
33 compat_shlex_quote,
34 compat_str,
35 compat_urllib_error,
36 compat_urllib_request,
37 )
38 from .cookies import load_cookies
39 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
40 from .downloader.rtmp import rtmpdump_version
41 from .extractor import gen_extractor_classes, get_info_extractor
42 from .extractor.openload import PhantomJSwrapper
43 from .minicurses import format_text
44 from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
45 from .postprocessor import (
46 EmbedThumbnailPP,
47 FFmpegFixupDuplicateMoovPP,
48 FFmpegFixupDurationPP,
49 FFmpegFixupM3u8PP,
50 FFmpegFixupM4aPP,
51 FFmpegFixupStretchedPP,
52 FFmpegFixupTimestampPP,
53 FFmpegMergerPP,
54 FFmpegPostProcessor,
55 MoveFilesAfterDownloadPP,
56 get_postprocessor,
57 )
58 from .update import detect_variant
59 from .utils import (
60 DEFAULT_OUTTMPL,
61 LINK_TEMPLATES,
62 NO_DEFAULT,
63 NUMBER_RE,
64 OUTTMPL_TYPES,
65 POSTPROCESS_WHEN,
66 STR_FORMAT_RE_TMPL,
67 STR_FORMAT_TYPES,
68 ContentTooShortError,
69 DateRange,
70 DownloadCancelled,
71 DownloadError,
72 EntryNotInPlaylist,
73 ExistingVideoReached,
74 ExtractorError,
75 GeoRestrictedError,
76 HEADRequest,
77 ISO3166Utils,
78 LazyList,
79 MaxDownloadsReached,
80 Namespace,
81 PagedList,
82 PerRequestProxyHandler,
83 PlaylistEntries,
84 Popen,
85 PostProcessingError,
86 ReExtractInfo,
87 RejectedVideoReached,
88 SameFileError,
89 UnavailableVideoError,
90 YoutubeDLCookieProcessor,
91 YoutubeDLHandler,
92 YoutubeDLRedirectHandler,
93 age_restricted,
94 args_to_str,
95 date_from_str,
96 determine_ext,
97 determine_protocol,
98 encode_compat_str,
99 encodeFilename,
100 error_to_compat_str,
101 expand_path,
102 filter_dict,
103 float_or_none,
104 format_bytes,
105 format_decimal_suffix,
106 format_field,
107 formatSeconds,
108 get_domain,
109 int_or_none,
110 iri_to_uri,
111 join_nonempty,
112 locked_file,
113 make_dir,
114 make_HTTPS_handler,
115 merge_headers,
116 network_exceptions,
117 number_of_digits,
118 orderedSet,
119 parse_filesize,
120 platform_name,
121 preferredencoding,
122 prepend_extension,
123 register_socks_protocols,
124 remove_terminal_sequences,
125 render_table,
126 replace_extension,
127 sanitize_filename,
128 sanitize_path,
129 sanitize_url,
130 sanitized_Request,
131 std_headers,
132 str_or_none,
133 strftime_or_none,
134 subtitles_filename,
135 supports_terminal_sequences,
136 timetuple_from_msec,
137 to_high_limit_path,
138 traverse_obj,
139 try_get,
140 url_basename,
141 variadic,
142 version_tuple,
143 windows_enable_vt_mode,
144 write_json_file,
145 write_string,
146 )
147 from .version import RELEASE_GIT_HEAD, __version__
148
149 if compat_os_name == 'nt':
150 import ctypes
151
152
153 class YoutubeDL:
154 """YoutubeDL class.
155
156 YoutubeDL objects are the ones responsible of downloading the
157 actual video file and writing it to disk if the user has requested
158 it, among some other tasks. In most cases there should be one per
159 program. As, given a video URL, the downloader doesn't know how to
160 extract all the needed information, task that InfoExtractors do, it
161 has to pass the URL to one of them.
162
163 For this, YoutubeDL objects have a method that allows
164 InfoExtractors to be registered in a given order. When it is passed
165 a URL, the YoutubeDL object handles it to the first InfoExtractor it
166 finds that reports being able to handle it. The InfoExtractor extracts
167 all the information about the video or videos the URL refers to, and
168 YoutubeDL process the extracted information, possibly using a File
169 Downloader to download the video.
170
171 YoutubeDL objects accept a lot of parameters. In order not to saturate
172 the object constructor with arguments, it receives a dictionary of
173 options instead. These options are available through the params
174 attribute for the InfoExtractors to use. The YoutubeDL also
175 registers itself as the downloader in charge for the InfoExtractors
176 that are added to it, so this is a "mutual registration".
177
178 Available options:
179
180 username: Username for authentication purposes.
181 password: Password for authentication purposes.
182 videopassword: Password for accessing a video.
183 ap_mso: Adobe Pass multiple-system operator identifier.
184 ap_username: Multiple-system operator account username.
185 ap_password: Multiple-system operator account password.
186 usenetrc: Use netrc for authentication instead.
187 verbose: Print additional info to stdout.
188 quiet: Do not print messages to stdout.
189 no_warnings: Do not print out anything for warnings.
190 forceprint: A dict with keys WHEN mapped to a list of templates to
191 print to stdout. The allowed keys are video or any of the
192 items in utils.POSTPROCESS_WHEN.
193 For compatibility, a single list is also accepted
194 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
195 a list of tuples with (template, filename)
196 forcejson: Force printing info_dict as JSON.
197 dump_single_json: Force printing the info_dict of the whole playlist
198 (or video) as a single JSON line.
199 force_write_download_archive: Force writing download archive regardless
200 of 'skip_download' or 'simulate'.
201 simulate: Do not download the video files. If unset (or None),
202 simulate only if listsubtitles, listformats or list_thumbnails is used
203 format: Video format code. see "FORMAT SELECTION" for more details.
204 You can also pass a function. The function takes 'ctx' as
205 argument and returns the formats to download.
206 See "build_format_selector" for an implementation
207 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
208 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
209 extracting metadata even if the video is not actually
210 available for download (experimental)
211 format_sort: A list of fields by which to sort the video formats.
212 See "Sorting Formats" for more details.
213 format_sort_force: Force the given format_sort. see "Sorting Formats"
214 for more details.
215 prefer_free_formats: Whether to prefer video formats with free containers
216 over non-free ones of same quality.
217 allow_multiple_video_streams: Allow multiple video streams to be merged
218 into a single file
219 allow_multiple_audio_streams: Allow multiple audio streams to be merged
220 into a single file
221 check_formats Whether to test if the formats are downloadable.
222 Can be True (check all), False (check none),
223 'selected' (check selected formats),
224 or None (check only if requested by extractor)
225 paths: Dictionary of output paths. The allowed keys are 'home'
226 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
227 outtmpl: Dictionary of templates for output names. Allowed keys
228 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
229 For compatibility with youtube-dl, a single string can also be used
230 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
231 restrictfilenames: Do not allow "&" and spaces in file names
232 trim_file_name: Limit length of filename (extension excluded)
233 windowsfilenames: Force the filenames to be windows compatible
234 ignoreerrors: Do not stop on download/postprocessing errors.
235 Can be 'only_download' to ignore only download errors.
236 Default is 'only_download' for CLI, but False for API
237 skip_playlist_after_errors: Number of allowed failures until the rest of
238 the playlist is skipped
239 force_generic_extractor: Force downloader to use the generic extractor
240 overwrites: Overwrite all video and metadata files if True,
241 overwrite only non-video files if None
242 and don't overwrite any file if False
243 For compatibility with youtube-dl,
244 "nooverwrites" may also be used instead
245 playlist_items: Specific indices of playlist to download.
246 playlistrandom: Download playlist items in random order.
247 lazy_playlist: Process playlist entries as they are received.
248 matchtitle: Download only matching titles.
249 rejecttitle: Reject downloads for matching titles.
250 logger: Log messages to a logging.Logger instance.
251 logtostderr: Log messages to stderr instead of stdout.
252 consoletitle: Display progress in console window's titlebar.
253 writedescription: Write the video description to a .description file
254 writeinfojson: Write the video description to a .info.json file
255 clean_infojson: Remove private fields from the infojson
256 getcomments: Extract video comments. This will not be written to disk
257 unless writeinfojson is also given
258 writeannotations: Write the video annotations to a .annotations.xml file
259 writethumbnail: Write the thumbnail image to a file
260 allow_playlist_files: Whether to write playlists' description, infojson etc
261 also to disk when using the 'write*' options
262 write_all_thumbnails: Write all thumbnail formats to files
263 writelink: Write an internet shortcut file, depending on the
264 current platform (.url/.webloc/.desktop)
265 writeurllink: Write a Windows internet shortcut file (.url)
266 writewebloclink: Write a macOS internet shortcut file (.webloc)
267 writedesktoplink: Write a Linux internet shortcut file (.desktop)
268 writesubtitles: Write the video subtitles to a file
269 writeautomaticsub: Write the automatically generated subtitles to a file
270 listsubtitles: Lists all available subtitles for the video
271 subtitlesformat: The format code for subtitles
272 subtitleslangs: List of languages of the subtitles to download (can be regex).
273 The list may contain "all" to refer to all the available
274 subtitles. The language can be prefixed with a "-" to
275 exclude it from the requested languages. Eg: ['all', '-live_chat']
276 keepvideo: Keep the video file after post-processing
277 daterange: A DateRange object, download only if the upload_date is in the range.
278 skip_download: Skip the actual download of the video file
279 cachedir: Location of the cache files in the filesystem.
280 False to disable filesystem cache.
281 noplaylist: Download single video instead of a playlist if in doubt.
282 age_limit: An integer representing the user's age in years.
283 Unsuitable videos for the given age are skipped.
284 min_views: An integer representing the minimum view count the video
285 must have in order to not be skipped.
286 Videos without view count information are always
287 downloaded. None for no limit.
288 max_views: An integer representing the maximum view count.
289 Videos that are more popular than that are not
290 downloaded.
291 Videos without view count information are always
292 downloaded. None for no limit.
293 download_archive: File name of a file where all downloads are recorded.
294 Videos already present in the file are not downloaded
295 again.
296 break_on_existing: Stop the download process after attempting to download a
297 file that is in the archive.
298 break_on_reject: Stop the download process when encountering a video that
299 has been filtered out.
300 break_per_url: Whether break_on_reject and break_on_existing
301 should act on each input URL as opposed to for the entire queue
302 cookiefile: File name or text stream from where cookies should be read and dumped to
303 cookiesfrombrowser: A tuple containing the name of the browser, the profile
304 name/pathfrom where cookies are loaded, and the name of the
305 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
306 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
307 support RFC 5746 secure renegotiation
308 nocheckcertificate: Do not verify SSL certificates
309 client_certificate: Path to client certificate file in PEM format. May include the private key
310 client_certificate_key: Path to private key file for client certificate
311 client_certificate_password: Password for client certificate private key, if encrypted.
312 If not provided and the key is encrypted, yt-dlp will ask interactively
313 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
314 At the moment, this is only supported by YouTube.
315 http_headers: A dictionary of custom headers to be used for all requests
316 proxy: URL of the proxy server to use
317 geo_verification_proxy: URL of the proxy to use for IP address verification
318 on geo-restricted sites.
319 socket_timeout: Time to wait for unresponsive hosts, in seconds
320 bidi_workaround: Work around buggy terminals without bidirectional text
321 support, using fridibi
322 debug_printtraffic:Print out sent and received HTTP traffic
323 default_search: Prepend this string if an input url is not valid.
324 'auto' for elaborate guessing
325 encoding: Use this encoding instead of the system-specified.
326 extract_flat: Do not resolve URLs, return the immediate result.
327 Pass in 'in_playlist' to only show this behavior for
328 playlist items.
329 wait_for_video: If given, wait for scheduled streams to become available.
330 The value should be a tuple containing the range
331 (min_secs, max_secs) to wait between retries
332 postprocessors: A list of dictionaries, each with an entry
333 * key: The name of the postprocessor. See
334 yt_dlp/postprocessor/__init__.py for a list.
335 * when: When to run the postprocessor. Allowed values are
336 the entries of utils.POSTPROCESS_WHEN
337 Assumed to be 'post_process' if not given
338 progress_hooks: A list of functions that get called on download
339 progress, with a dictionary with the entries
340 * status: One of "downloading", "error", or "finished".
341 Check this first and ignore unknown values.
342 * info_dict: The extracted info_dict
343
344 If status is one of "downloading", or "finished", the
345 following properties may also be present:
346 * filename: The final filename (always present)
347 * tmpfilename: The filename we're currently writing to
348 * downloaded_bytes: Bytes on disk
349 * total_bytes: Size of the whole file, None if unknown
350 * total_bytes_estimate: Guess of the eventual file size,
351 None if unavailable.
352 * elapsed: The number of seconds since download started.
353 * eta: The estimated time in seconds, None if unknown
354 * speed: The download speed in bytes/second, None if
355 unknown
356 * fragment_index: The counter of the currently
357 downloaded video fragment.
358 * fragment_count: The number of fragments (= individual
359 files that will be merged)
360
361 Progress hooks are guaranteed to be called at least once
362 (with status "finished") if the download is successful.
363 postprocessor_hooks: A list of functions that get called on postprocessing
364 progress, with a dictionary with the entries
365 * status: One of "started", "processing", or "finished".
366 Check this first and ignore unknown values.
367 * postprocessor: Name of the postprocessor
368 * info_dict: The extracted info_dict
369
370 Progress hooks are guaranteed to be called at least twice
371 (with status "started" and "finished") if the processing is successful.
372 merge_output_format: Extension to use when merging formats.
373 final_ext: Expected final extension; used to detect when the file was
374 already downloaded and converted
375 fixup: Automatically correct known faults of the file.
376 One of:
377 - "never": do nothing
378 - "warn": only emit a warning
379 - "detect_or_warn": check whether we can do anything
380 about it, warn otherwise (default)
381 source_address: Client-side IP address to bind to.
382 sleep_interval_requests: Number of seconds to sleep between requests
383 during extraction
384 sleep_interval: Number of seconds to sleep before each download when
385 used alone or a lower bound of a range for randomized
386 sleep before each download (minimum possible number
387 of seconds to sleep) when used along with
388 max_sleep_interval.
389 max_sleep_interval:Upper bound of a range for randomized sleep before each
390 download (maximum possible number of seconds to sleep).
391 Must only be used along with sleep_interval.
392 Actual sleep time will be a random float from range
393 [sleep_interval; max_sleep_interval].
394 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
395 listformats: Print an overview of available video formats and exit.
396 list_thumbnails: Print a table of all thumbnails and exit.
397 match_filter: A function that gets called for every video with the signature
398 (info_dict, *, incomplete: bool) -> Optional[str]
399 For backward compatibility with youtube-dl, the signature
400 (info_dict) -> Optional[str] is also allowed.
401 - If it returns a message, the video is ignored.
402 - If it returns None, the video is downloaded.
403 - If it returns utils.NO_DEFAULT, the user is interactively
404 asked whether to download the video.
405 match_filter_func in utils.py is one example for this.
406 no_color: Do not emit color codes in output.
407 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
408 HTTP header
409 geo_bypass_country:
410 Two-letter ISO 3166-2 country code that will be used for
411 explicit geographic restriction bypassing via faking
412 X-Forwarded-For HTTP header
413 geo_bypass_ip_block:
414 IP range in CIDR notation that will be used similarly to
415 geo_bypass_country
416 external_downloader: A dictionary of protocol keys and the executable of the
417 external downloader to use for it. The allowed protocols
418 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
419 Set the value to 'native' to use the native downloader
420 compat_opts: Compatibility options. See "Differences in default behavior".
421 The following options do not work when used through the API:
422 filename, abort-on-error, multistreams, no-live-chat, format-sort
423 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
424 Refer __init__.py for their implementation
425 progress_template: Dictionary of templates for progress outputs.
426 Allowed keys are 'download', 'postprocess',
427 'download-title' (console title) and 'postprocess-title'.
428 The template is mapped on a dictionary with keys 'progress' and 'info'
429 retry_sleep_functions: Dictionary of functions that takes the number of attempts
430 as argument and returns the time to sleep in seconds.
431 Allowed keys are 'http', 'fragment', 'file_access'
432 download_ranges: A function that gets called for every video with the signature
433 (info_dict, *, ydl) -> Iterable[Section].
434 Only the returned sections will be downloaded. Each Section contains:
435 * start_time: Start time of the section in seconds
436 * end_time: End time of the section in seconds
437 * title: Section title (Optional)
438 * index: Section number (Optional)
439
440 The following parameters are not used by YoutubeDL itself, they are used by
441 the downloader (see yt_dlp/downloader/common.py):
442 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
443 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
444 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
445 external_downloader_args, concurrent_fragment_downloads.
446
447 The following options are used by the post processors:
448 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
449 to the binary or its containing directory.
450 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
451 and a list of additional command-line arguments for the
452 postprocessor/executable. The dict can also have "PP+EXE" keys
453 which are used when the given exe is used by the given PP.
454 Use 'default' as the name for arguments to passed to all PP
455 For compatibility with youtube-dl, a single list of args
456 can also be used
457
458 The following options are used by the extractors:
459 extractor_retries: Number of times to retry for known errors
460 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
461 hls_split_discontinuity: Split HLS playlists to different formats at
462 discontinuities such as ad breaks (default: False)
463 extractor_args: A dictionary of arguments to be passed to the extractors.
464 See "EXTRACTOR ARGUMENTS" for details.
465 Eg: {'youtube': {'skip': ['dash', 'hls']}}
466 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
467
468 The following options are deprecated and may be removed in the future:
469
470 playliststart: - Use playlist_items
471 Playlist item to start at.
472 playlistend: - Use playlist_items
473 Playlist item to end at.
474 playlistreverse: - Use playlist_items
475 Download playlist items in reverse order.
476 forceurl: - Use forceprint
477 Force printing final URL.
478 forcetitle: - Use forceprint
479 Force printing title.
480 forceid: - Use forceprint
481 Force printing ID.
482 forcethumbnail: - Use forceprint
483 Force printing thumbnail URL.
484 forcedescription: - Use forceprint
485 Force printing description.
486 forcefilename: - Use forceprint
487 Force printing final filename.
488 forceduration: - Use forceprint
489 Force printing duration.
490 allsubtitles: - Use subtitleslangs = ['all']
491 Downloads all the subtitles of the video
492 (requires writesubtitles or writeautomaticsub)
493 include_ads: - Doesn't work
494 Download ads as well
495 call_home: - Not implemented
496 Boolean, true iff we are allowed to contact the
497 yt-dlp servers for debugging.
498 post_hooks: - Register a custom postprocessor
499 A list of functions that get called as the final step
500 for each video file, after all postprocessors have been
501 called. The filename will be passed as the only argument.
502 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
503 Use the native HLS downloader instead of ffmpeg/avconv
504 if True, otherwise use ffmpeg/avconv if False, otherwise
505 use downloader suggested by extractor if None.
506 prefer_ffmpeg: - avconv support is deprecated
507 If False, use avconv instead of ffmpeg if both are available,
508 otherwise prefer ffmpeg.
509 youtube_include_dash_manifest: - Use extractor_args
510 If True (default), DASH manifests and related
511 data will be downloaded and processed by extractor.
512 You can reduce network I/O by disabling it if you don't
513 care about DASH. (only for youtube)
514 youtube_include_hls_manifest: - Use extractor_args
515 If True (default), HLS manifests and related
516 data will be downloaded and processed by extractor.
517 You can reduce network I/O by disabling it if you don't
518 care about HLS. (only for youtube)
519 """
520
521 _NUMERIC_FIELDS = {
522 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
523 'timestamp', 'release_timestamp',
524 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
525 'average_rating', 'comment_count', 'age_limit',
526 'start_time', 'end_time',
527 'chapter_number', 'season_number', 'episode_number',
528 'track_number', 'disc_number', 'release_year',
529 }
530
531 _format_fields = {
532 # NB: Keep in sync with the docstring of extractor/common.py
533 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
534 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
535 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
536 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
537 'preference', 'language', 'language_preference', 'quality', 'source_preference',
538 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
539 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
540 }
541 _format_selection_exts = {
542 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
543 'video': {'mp4', 'flv', 'webm', '3gp'},
544 'storyboards': {'mhtml'},
545 }
546
547 def __init__(self, params=None, auto_init=True):
548 """Create a FileDownloader object with the given options.
549 @param auto_init Whether to load the default extractors and print header (if verbose).
550 Set to 'no_verbose_header' to not print the header
551 """
552 if params is None:
553 params = {}
554 self.params = params
555 self._ies = {}
556 self._ies_instances = {}
557 self._pps = {k: [] for k in POSTPROCESS_WHEN}
558 self._printed_messages = set()
559 self._first_webpage_request = True
560 self._post_hooks = []
561 self._progress_hooks = []
562 self._postprocessor_hooks = []
563 self._download_retcode = 0
564 self._num_downloads = 0
565 self._num_videos = 0
566 self._playlist_level = 0
567 self._playlist_urls = set()
568 self.cache = Cache(self)
569
570 windows_enable_vt_mode()
571 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
572 self._out_files = Namespace(
573 out=stdout,
574 error=sys.stderr,
575 screen=sys.stderr if self.params.get('quiet') else stdout,
576 console=None if compat_os_name == 'nt' else next(
577 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
578 )
579 self._allow_colors = Namespace(**{
580 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
581 for type_, stream in self._out_files.items_ if type_ != 'console'
582 })
583
584 if sys.version_info < (3, 6):
585 self.report_warning(
586 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
587
588 if self.params.get('allow_unplayable_formats'):
589 self.report_warning(
590 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
591 'This is a developer option intended for debugging. \n'
592 ' If you experience any issues while using this option, '
593 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
594
595 def check_deprecated(param, option, suggestion):
596 if self.params.get(param) is not None:
597 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
598 return True
599 return False
600
601 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
602 if self.params.get('geo_verification_proxy') is None:
603 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
604
605 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
606 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
607 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
608
609 for msg in self.params.get('_warnings', []):
610 self.report_warning(msg)
611 for msg in self.params.get('_deprecation_warnings', []):
612 self.deprecation_warning(msg)
613
614 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
615 if not compat_has_legacy:
616 self.params['compat_opts'].add('no-compat-legacy')
617 if 'list-formats' in self.params['compat_opts']:
618 self.params['listformats_table'] = False
619
620 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
621 # nooverwrites was unnecessarily changed to overwrites
622 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
623 # This ensures compatibility with both keys
624 self.params['overwrites'] = not self.params['nooverwrites']
625 elif self.params.get('overwrites') is None:
626 self.params.pop('overwrites', None)
627 else:
628 self.params['nooverwrites'] = not self.params['overwrites']
629
630 self.params.setdefault('forceprint', {})
631 self.params.setdefault('print_to_file', {})
632
633 # Compatibility with older syntax
634 if not isinstance(params['forceprint'], dict):
635 self.params['forceprint'] = {'video': params['forceprint']}
636
637 if self.params.get('bidi_workaround', False):
638 try:
639 import pty
640 master, slave = pty.openpty()
641 width = compat_get_terminal_size().columns
642 width_args = [] if width is None else ['-w', str(width)]
643 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
644 try:
645 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
646 except OSError:
647 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
648 self._output_channel = os.fdopen(master, 'rb')
649 except OSError as ose:
650 if ose.errno == errno.ENOENT:
651 self.report_warning(
652 'Could not find fribidi executable, ignoring --bidi-workaround. '
653 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
654 else:
655 raise
656
657 if auto_init:
658 if auto_init != 'no_verbose_header':
659 self.print_debug_header()
660 self.add_default_info_extractors()
661
662 if (sys.platform != 'win32'
663 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
664 and not self.params.get('restrictfilenames', False)):
665 # Unicode filesystem API will throw errors (#1474, #13027)
666 self.report_warning(
667 'Assuming --restrict-filenames since file system encoding '
668 'cannot encode all characters. '
669 'Set the LC_ALL environment variable to fix this.')
670 self.params['restrictfilenames'] = True
671
672 self._parse_outtmpl()
673
674 # Creating format selector here allows us to catch syntax errors before the extraction
675 self.format_selector = (
676 self.params.get('format') if self.params.get('format') in (None, '-')
677 else self.params['format'] if callable(self.params['format'])
678 else self.build_format_selector(self.params['format']))
679
680 # Set http_headers defaults according to std_headers
681 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
682
683 hooks = {
684 'post_hooks': self.add_post_hook,
685 'progress_hooks': self.add_progress_hook,
686 'postprocessor_hooks': self.add_postprocessor_hook,
687 }
688 for opt, fn in hooks.items():
689 for ph in self.params.get(opt, []):
690 fn(ph)
691
692 for pp_def_raw in self.params.get('postprocessors', []):
693 pp_def = dict(pp_def_raw)
694 when = pp_def.pop('when', 'post_process')
695 self.add_post_processor(
696 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
697 when=when)
698
699 self._setup_opener()
700 register_socks_protocols()
701
702 def preload_download_archive(fn):
703 """Preload the archive, if any is specified"""
704 if fn is None:
705 return False
706 self.write_debug(f'Loading archive file {fn!r}')
707 try:
708 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
709 for line in archive_file:
710 self.archive.add(line.strip())
711 except OSError as ioe:
712 if ioe.errno != errno.ENOENT:
713 raise
714 return False
715 return True
716
717 self.archive = set()
718 preload_download_archive(self.params.get('download_archive'))
719
720 def warn_if_short_id(self, argv):
721 # short YouTube ID starting with dash?
722 idxs = [
723 i for i, a in enumerate(argv)
724 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
725 if idxs:
726 correct_argv = (
727 ['yt-dlp']
728 + [a for i, a in enumerate(argv) if i not in idxs]
729 + ['--'] + [argv[i] for i in idxs]
730 )
731 self.report_warning(
732 'Long argument string detected. '
733 'Use -- to separate parameters and URLs, like this:\n%s' %
734 args_to_str(correct_argv))
735
736 def add_info_extractor(self, ie):
737 """Add an InfoExtractor object to the end of the list."""
738 ie_key = ie.ie_key()
739 self._ies[ie_key] = ie
740 if not isinstance(ie, type):
741 self._ies_instances[ie_key] = ie
742 ie.set_downloader(self)
743
744 def _get_info_extractor_class(self, ie_key):
745 ie = self._ies.get(ie_key)
746 if ie is None:
747 ie = get_info_extractor(ie_key)
748 self.add_info_extractor(ie)
749 return ie
750
751 def get_info_extractor(self, ie_key):
752 """
753 Get an instance of an IE with name ie_key, it will try to get one from
754 the _ies list, if there's no instance it will create a new one and add
755 it to the extractor list.
756 """
757 ie = self._ies_instances.get(ie_key)
758 if ie is None:
759 ie = get_info_extractor(ie_key)()
760 self.add_info_extractor(ie)
761 return ie
762
763 def add_default_info_extractors(self):
764 """
765 Add the InfoExtractors returned by gen_extractors to the end of the list
766 """
767 for ie in gen_extractor_classes():
768 self.add_info_extractor(ie)
769
770 def add_post_processor(self, pp, when='post_process'):
771 """Add a PostProcessor object to the end of the chain."""
772 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
773 self._pps[when].append(pp)
774 pp.set_downloader(self)
775
776 def add_post_hook(self, ph):
777 """Add the post hook"""
778 self._post_hooks.append(ph)
779
780 def add_progress_hook(self, ph):
781 """Add the download progress hook"""
782 self._progress_hooks.append(ph)
783
784 def add_postprocessor_hook(self, ph):
785 """Add the postprocessing progress hook"""
786 self._postprocessor_hooks.append(ph)
787 for pps in self._pps.values():
788 for pp in pps:
789 pp.add_progress_hook(ph)
790
791 def _bidi_workaround(self, message):
792 if not hasattr(self, '_output_channel'):
793 return message
794
795 assert hasattr(self, '_output_process')
796 assert isinstance(message, compat_str)
797 line_count = message.count('\n') + 1
798 self._output_process.stdin.write((message + '\n').encode())
799 self._output_process.stdin.flush()
800 res = ''.join(self._output_channel.readline().decode()
801 for _ in range(line_count))
802 return res[:-len('\n')]
803
804 def _write_string(self, message, out=None, only_once=False):
805 if only_once:
806 if message in self._printed_messages:
807 return
808 self._printed_messages.add(message)
809 write_string(message, out=out, encoding=self.params.get('encoding'))
810
811 def to_stdout(self, message, skip_eol=False, quiet=None):
812 """Print message to stdout"""
813 if quiet is not None:
814 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
815 if skip_eol is not False:
816 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. Use "YoutubeDL.to_screen" instead')
817 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
818
819 def to_screen(self, message, skip_eol=False, quiet=None):
820 """Print message to screen if not in quiet mode"""
821 if self.params.get('logger'):
822 self.params['logger'].debug(message)
823 return
824 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
825 return
826 self._write_string(
827 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
828 self._out_files.screen)
829
830 def to_stderr(self, message, only_once=False):
831 """Print message to stderr"""
832 assert isinstance(message, compat_str)
833 if self.params.get('logger'):
834 self.params['logger'].error(message)
835 else:
836 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
837
838 def _send_console_code(self, code):
839 if compat_os_name == 'nt' or not self._out_files.console:
840 return
841 self._write_string(code, self._out_files.console)
842
843 def to_console_title(self, message):
844 if not self.params.get('consoletitle', False):
845 return
846 message = remove_terminal_sequences(message)
847 if compat_os_name == 'nt':
848 if ctypes.windll.kernel32.GetConsoleWindow():
849 # c_wchar_p() might not be necessary if `message` is
850 # already of type unicode()
851 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
852 else:
853 self._send_console_code(f'\033]0;{message}\007')
854
855 def save_console_title(self):
856 if not self.params.get('consoletitle') or self.params.get('simulate'):
857 return
858 self._send_console_code('\033[22;0t') # Save the title on stack
859
860 def restore_console_title(self):
861 if not self.params.get('consoletitle') or self.params.get('simulate'):
862 return
863 self._send_console_code('\033[23;0t') # Restore the title from stack
864
865 def __enter__(self):
866 self.save_console_title()
867 return self
868
869 def __exit__(self, *args):
870 self.restore_console_title()
871
872 if self.params.get('cookiefile') is not None:
873 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
874
875 def trouble(self, message=None, tb=None, is_error=True):
876 """Determine action to take when a download problem appears.
877
878 Depending on if the downloader has been configured to ignore
879 download errors or not, this method may throw an exception or
880 not when errors are found, after printing the message.
881
882 @param tb If given, is additional traceback information
883 @param is_error Whether to raise error according to ignorerrors
884 """
885 if message is not None:
886 self.to_stderr(message)
887 if self.params.get('verbose'):
888 if tb is None:
889 if sys.exc_info()[0]: # if .trouble has been called from an except block
890 tb = ''
891 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
892 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
893 tb += encode_compat_str(traceback.format_exc())
894 else:
895 tb_data = traceback.format_list(traceback.extract_stack())
896 tb = ''.join(tb_data)
897 if tb:
898 self.to_stderr(tb)
899 if not is_error:
900 return
901 if not self.params.get('ignoreerrors'):
902 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
903 exc_info = sys.exc_info()[1].exc_info
904 else:
905 exc_info = sys.exc_info()
906 raise DownloadError(message, exc_info)
907 self._download_retcode = 1
908
909 Styles = Namespace(
910 HEADERS='yellow',
911 EMPHASIS='light blue',
912 FILENAME='green',
913 ID='green',
914 DELIM='blue',
915 ERROR='red',
916 WARNING='yellow',
917 SUPPRESS='light black',
918 )
919
920 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
921 text = str(text)
922 if test_encoding:
923 original_text = text
924 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
925 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
926 text = text.encode(encoding, 'ignore').decode(encoding)
927 if fallback is not None and text != original_text:
928 text = fallback
929 return format_text(text, f) if allow_colors else text if fallback is None else fallback
930
931 def _format_out(self, *args, **kwargs):
932 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
933
934 def _format_screen(self, *args, **kwargs):
935 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
936
937 def _format_err(self, *args, **kwargs):
938 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
939
940 def report_warning(self, message, only_once=False):
941 '''
942 Print the message to stderr, it will be prefixed with 'WARNING:'
943 If stderr is a tty file the 'WARNING:' will be colored
944 '''
945 if self.params.get('logger') is not None:
946 self.params['logger'].warning(message)
947 else:
948 if self.params.get('no_warnings'):
949 return
950 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
951
952 def deprecation_warning(self, message):
953 if self.params.get('logger') is not None:
954 self.params['logger'].warning(f'DeprecationWarning: {message}')
955 else:
956 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
957
958 def report_error(self, message, *args, **kwargs):
959 '''
960 Do the same as trouble, but prefixes the message with 'ERROR:', colored
961 in red if stderr is a tty file.
962 '''
963 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
964
965 def write_debug(self, message, only_once=False):
966 '''Log debug message or Print message to stderr'''
967 if not self.params.get('verbose', False):
968 return
969 message = f'[debug] {message}'
970 if self.params.get('logger'):
971 self.params['logger'].debug(message)
972 else:
973 self.to_stderr(message, only_once)
974
975 def report_file_already_downloaded(self, file_name):
976 """Report file has already been fully downloaded."""
977 try:
978 self.to_screen('[download] %s has already been downloaded' % file_name)
979 except UnicodeEncodeError:
980 self.to_screen('[download] The file has already been downloaded')
981
982 def report_file_delete(self, file_name):
983 """Report that existing file will be deleted."""
984 try:
985 self.to_screen('Deleting existing file %s' % file_name)
986 except UnicodeEncodeError:
987 self.to_screen('Deleting existing file')
988
989 def raise_no_formats(self, info, forced=False, *, msg=None):
990 has_drm = info.get('_has_drm')
991 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
992 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
993 if forced or not ignored:
994 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
995 expected=has_drm or ignored or expected)
996 else:
997 self.report_warning(msg)
998
999 def parse_outtmpl(self):
1000 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1001 self._parse_outtmpl()
1002 return self.params['outtmpl']
1003
1004 def _parse_outtmpl(self):
1005 sanitize = lambda x: x
1006 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1007 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1008
1009 outtmpl = self.params.setdefault('outtmpl', {})
1010 if not isinstance(outtmpl, dict):
1011 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1012 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1013
1014 def get_output_path(self, dir_type='', filename=None):
1015 paths = self.params.get('paths', {})
1016 assert isinstance(paths, dict)
1017 path = os.path.join(
1018 expand_path(paths.get('home', '').strip()),
1019 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1020 filename or '')
1021 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1022
1023 @staticmethod
1024 def _outtmpl_expandpath(outtmpl):
1025 # expand_path translates '%%' into '%' and '$$' into '$'
1026 # correspondingly that is not what we want since we need to keep
1027 # '%%' intact for template dict substitution step. Working around
1028 # with boundary-alike separator hack.
1029 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1030 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1031
1032 # outtmpl should be expand_path'ed before template dict substitution
1033 # because meta fields may contain env variables we don't want to
1034 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1035 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1036 return expand_path(outtmpl).replace(sep, '')
1037
1038 @staticmethod
1039 def escape_outtmpl(outtmpl):
1040 ''' Escape any remaining strings like %s, %abc% etc. '''
1041 return re.sub(
1042 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1043 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1044 outtmpl)
1045
1046 @classmethod
1047 def validate_outtmpl(cls, outtmpl):
1048 ''' @return None or Exception object '''
1049 outtmpl = re.sub(
1050 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1051 lambda mobj: f'{mobj.group(0)[:-1]}s',
1052 cls._outtmpl_expandpath(outtmpl))
1053 try:
1054 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1055 return None
1056 except ValueError as err:
1057 return err
1058
1059 @staticmethod
1060 def _copy_infodict(info_dict):
1061 info_dict = dict(info_dict)
1062 info_dict.pop('__postprocessors', None)
1063 info_dict.pop('__pending_error', None)
1064 return info_dict
1065
1066 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1067 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1068 @param sanitize Whether to sanitize the output as a filename.
1069 For backward compatibility, a function can also be passed
1070 """
1071
1072 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1073
1074 info_dict = self._copy_infodict(info_dict)
1075 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1076 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1077 if info_dict.get('duration', None) is not None
1078 else None)
1079 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1080 info_dict['video_autonumber'] = self._num_videos
1081 if info_dict.get('resolution') is None:
1082 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1083
1084 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1085 # of %(field)s to %(field)0Nd for backward compatibility
1086 field_size_compat_map = {
1087 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1088 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1089 'autonumber': self.params.get('autonumber_size') or 5,
1090 }
1091
1092 TMPL_DICT = {}
1093 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1094 MATH_FUNCTIONS = {
1095 '+': float.__add__,
1096 '-': float.__sub__,
1097 }
1098 # Field is of the form key1.key2...
1099 # where keys (except first) can be string, int or slice
1100 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1101 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1102 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1103 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1104 (?P<negate>-)?
1105 (?P<fields>{FIELD_RE})
1106 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1107 (?:>(?P<strf_format>.+?))?
1108 (?P<remaining>
1109 (?P<alternate>(?<!\\),[^|&)]+)?
1110 (?:&(?P<replacement>.*?))?
1111 (?:\|(?P<default>.*?))?
1112 )$''')
1113
1114 def _traverse_infodict(k):
1115 k = k.split('.')
1116 if k[0] == '':
1117 k.pop(0)
1118 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1119
1120 def get_value(mdict):
1121 # Object traversal
1122 value = _traverse_infodict(mdict['fields'])
1123 # Negative
1124 if mdict['negate']:
1125 value = float_or_none(value)
1126 if value is not None:
1127 value *= -1
1128 # Do maths
1129 offset_key = mdict['maths']
1130 if offset_key:
1131 value = float_or_none(value)
1132 operator = None
1133 while offset_key:
1134 item = re.match(
1135 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1136 offset_key).group(0)
1137 offset_key = offset_key[len(item):]
1138 if operator is None:
1139 operator = MATH_FUNCTIONS[item]
1140 continue
1141 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1142 offset = float_or_none(item)
1143 if offset is None:
1144 offset = float_or_none(_traverse_infodict(item))
1145 try:
1146 value = operator(value, multiplier * offset)
1147 except (TypeError, ZeroDivisionError):
1148 return None
1149 operator = None
1150 # Datetime formatting
1151 if mdict['strf_format']:
1152 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1153
1154 return value
1155
1156 na = self.params.get('outtmpl_na_placeholder', 'NA')
1157
1158 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1159 return sanitize_filename(str(value), restricted=restricted, is_id=(
1160 bool(re.search(r'(^|[_.])id(\.|$)', key))
1161 if 'filename-sanitization' in self.params['compat_opts']
1162 else NO_DEFAULT))
1163
1164 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1165 sanitize = bool(sanitize)
1166
1167 def _dumpjson_default(obj):
1168 if isinstance(obj, (set, LazyList)):
1169 return list(obj)
1170 return repr(obj)
1171
1172 def create_key(outer_mobj):
1173 if not outer_mobj.group('has_key'):
1174 return outer_mobj.group(0)
1175 key = outer_mobj.group('key')
1176 mobj = re.match(INTERNAL_FORMAT_RE, key)
1177 initial_field = mobj.group('fields') if mobj else ''
1178 value, replacement, default = None, None, na
1179 while mobj:
1180 mobj = mobj.groupdict()
1181 default = mobj['default'] if mobj['default'] is not None else default
1182 value = get_value(mobj)
1183 replacement = mobj['replacement']
1184 if value is None and mobj['alternate']:
1185 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1186 else:
1187 break
1188
1189 fmt = outer_mobj.group('format')
1190 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1191 fmt = f'0{field_size_compat_map[key]:d}d'
1192
1193 value = default if value is None else value if replacement is None else replacement
1194
1195 flags = outer_mobj.group('conversion') or ''
1196 str_fmt = f'{fmt[:-1]}s'
1197 if fmt[-1] == 'l': # list
1198 delim = '\n' if '#' in flags else ', '
1199 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1200 elif fmt[-1] == 'j': # json
1201 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1202 elif fmt[-1] == 'q': # quoted
1203 value = map(str, variadic(value) if '#' in flags else [value])
1204 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1205 elif fmt[-1] == 'B': # bytes
1206 value = f'%{str_fmt}'.encode() % str(value).encode()
1207 value, fmt = value.decode('utf-8', 'ignore'), 's'
1208 elif fmt[-1] == 'U': # unicode normalized
1209 value, fmt = unicodedata.normalize(
1210 # "+" = compatibility equivalence, "#" = NFD
1211 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1212 value), str_fmt
1213 elif fmt[-1] == 'D': # decimal suffix
1214 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1215 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1216 factor=1024 if '#' in flags else 1000)
1217 elif fmt[-1] == 'S': # filename sanitization
1218 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1219 elif fmt[-1] == 'c':
1220 if value:
1221 value = str(value)[0]
1222 else:
1223 fmt = str_fmt
1224 elif fmt[-1] not in 'rs': # numeric
1225 value = float_or_none(value)
1226 if value is None:
1227 value, fmt = default, 's'
1228
1229 if sanitize:
1230 if fmt[-1] == 'r':
1231 # If value is an object, sanitize might convert it to a string
1232 # So we convert it to repr first
1233 value, fmt = repr(value), str_fmt
1234 if fmt[-1] in 'csr':
1235 value = sanitizer(initial_field, value)
1236
1237 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1238 TMPL_DICT[key] = value
1239 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1240
1241 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1242
1243 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1244 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1245 return self.escape_outtmpl(outtmpl) % info_dict
1246
1247 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1248 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1249 if outtmpl is None:
1250 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1251 try:
1252 outtmpl = self._outtmpl_expandpath(outtmpl)
1253 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1254 if not filename:
1255 return None
1256
1257 if tmpl_type in ('', 'temp'):
1258 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1259 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1260 filename = replace_extension(filename, ext, final_ext)
1261 elif tmpl_type:
1262 force_ext = OUTTMPL_TYPES[tmpl_type]
1263 if force_ext:
1264 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1265
1266 # https://github.com/blackjack4494/youtube-dlc/issues/85
1267 trim_file_name = self.params.get('trim_file_name', False)
1268 if trim_file_name:
1269 no_ext, *ext = filename.rsplit('.', 2)
1270 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1271
1272 return filename
1273 except ValueError as err:
1274 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1275 return None
1276
1277 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1278 """Generate the output filename"""
1279 if outtmpl:
1280 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1281 dir_type = None
1282 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1283 if not filename and dir_type not in ('', 'temp'):
1284 return ''
1285
1286 if warn:
1287 if not self.params.get('paths'):
1288 pass
1289 elif filename == '-':
1290 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1291 elif os.path.isabs(filename):
1292 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1293 if filename == '-' or not filename:
1294 return filename
1295
1296 return self.get_output_path(dir_type, filename)
1297
1298 def _match_entry(self, info_dict, incomplete=False, silent=False):
1299 """ Returns None if the file should be downloaded """
1300
1301 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1302
1303 def check_filter():
1304 if 'title' in info_dict:
1305 # This can happen when we're just evaluating the playlist
1306 title = info_dict['title']
1307 matchtitle = self.params.get('matchtitle', False)
1308 if matchtitle:
1309 if not re.search(matchtitle, title, re.IGNORECASE):
1310 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1311 rejecttitle = self.params.get('rejecttitle', False)
1312 if rejecttitle:
1313 if re.search(rejecttitle, title, re.IGNORECASE):
1314 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1315 date = info_dict.get('upload_date')
1316 if date is not None:
1317 dateRange = self.params.get('daterange', DateRange())
1318 if date not in dateRange:
1319 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1320 view_count = info_dict.get('view_count')
1321 if view_count is not None:
1322 min_views = self.params.get('min_views')
1323 if min_views is not None and view_count < min_views:
1324 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1325 max_views = self.params.get('max_views')
1326 if max_views is not None and view_count > max_views:
1327 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1328 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1329 return 'Skipping "%s" because it is age restricted' % video_title
1330
1331 match_filter = self.params.get('match_filter')
1332 if match_filter is not None:
1333 try:
1334 ret = match_filter(info_dict, incomplete=incomplete)
1335 except TypeError:
1336 # For backward compatibility
1337 ret = None if incomplete else match_filter(info_dict)
1338 if ret is NO_DEFAULT:
1339 while True:
1340 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1341 reply = input(self._format_screen(
1342 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1343 if reply in {'y', ''}:
1344 return None
1345 elif reply == 'n':
1346 return f'Skipping {video_title}'
1347 elif ret is not None:
1348 return ret
1349 return None
1350
1351 if self.in_download_archive(info_dict):
1352 reason = '%s has already been recorded in the archive' % video_title
1353 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1354 else:
1355 reason = check_filter()
1356 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1357 if reason is not None:
1358 if not silent:
1359 self.to_screen('[download] ' + reason)
1360 if self.params.get(break_opt, False):
1361 raise break_err()
1362 return reason
1363
1364 @staticmethod
1365 def add_extra_info(info_dict, extra_info):
1366 '''Set the keys from extra_info in info dict if they are missing'''
1367 for key, value in extra_info.items():
1368 info_dict.setdefault(key, value)
1369
1370 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1371 process=True, force_generic_extractor=False):
1372 """
1373 Return a list with a dictionary for each video extracted.
1374
1375 Arguments:
1376 url -- URL to extract
1377
1378 Keyword arguments:
1379 download -- whether to download videos during extraction
1380 ie_key -- extractor key hint
1381 extra_info -- dictionary containing the extra values to add to each result
1382 process -- whether to resolve all unresolved references (URLs, playlist items),
1383 must be True for download to work.
1384 force_generic_extractor -- force using the generic extractor
1385 """
1386
1387 if extra_info is None:
1388 extra_info = {}
1389
1390 if not ie_key and force_generic_extractor:
1391 ie_key = 'Generic'
1392
1393 if ie_key:
1394 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1395 else:
1396 ies = self._ies
1397
1398 for ie_key, ie in ies.items():
1399 if not ie.suitable(url):
1400 continue
1401
1402 if not ie.working():
1403 self.report_warning('The program functionality for this site has been marked as broken, '
1404 'and will probably not work.')
1405
1406 temp_id = ie.get_temp_id(url)
1407 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1408 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1409 if self.params.get('break_on_existing', False):
1410 raise ExistingVideoReached()
1411 break
1412 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1413 else:
1414 self.report_error('no suitable InfoExtractor for URL %s' % url)
1415
1416 def _handle_extraction_exceptions(func):
1417 @functools.wraps(func)
1418 def wrapper(self, *args, **kwargs):
1419 while True:
1420 try:
1421 return func(self, *args, **kwargs)
1422 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1423 raise
1424 except ReExtractInfo as e:
1425 if e.expected:
1426 self.to_screen(f'{e}; Re-extracting data')
1427 else:
1428 self.to_stderr('\r')
1429 self.report_warning(f'{e}; Re-extracting data')
1430 continue
1431 except GeoRestrictedError as e:
1432 msg = e.msg
1433 if e.countries:
1434 msg += '\nThis video is available in %s.' % ', '.join(
1435 map(ISO3166Utils.short2full, e.countries))
1436 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1437 self.report_error(msg)
1438 except ExtractorError as e: # An error we somewhat expected
1439 self.report_error(str(e), e.format_traceback())
1440 except Exception as e:
1441 if self.params.get('ignoreerrors'):
1442 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1443 else:
1444 raise
1445 break
1446 return wrapper
1447
1448 def _wait_for_video(self, ie_result):
1449 if (not self.params.get('wait_for_video')
1450 or ie_result.get('_type', 'video') != 'video'
1451 or ie_result.get('formats') or ie_result.get('url')):
1452 return
1453
1454 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1455 last_msg = ''
1456
1457 def progress(msg):
1458 nonlocal last_msg
1459 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1460 last_msg = msg
1461
1462 min_wait, max_wait = self.params.get('wait_for_video')
1463 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1464 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1465 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1466 self.report_warning('Release time of video is not known')
1467 elif (diff or 0) <= 0:
1468 self.report_warning('Video should already be available according to extracted info')
1469 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1470 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1471
1472 wait_till = time.time() + diff
1473 try:
1474 while True:
1475 diff = wait_till - time.time()
1476 if diff <= 0:
1477 progress('')
1478 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1479 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1480 time.sleep(1)
1481 except KeyboardInterrupt:
1482 progress('')
1483 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1484 except BaseException as e:
1485 if not isinstance(e, ReExtractInfo):
1486 self.to_screen('')
1487 raise
1488
1489 @_handle_extraction_exceptions
1490 def __extract_info(self, url, ie, download, extra_info, process):
1491 ie_result = ie.extract(url)
1492 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1493 return
1494 if isinstance(ie_result, list):
1495 # Backwards compatibility: old IE result format
1496 ie_result = {
1497 '_type': 'compat_list',
1498 'entries': ie_result,
1499 }
1500 if extra_info.get('original_url'):
1501 ie_result.setdefault('original_url', extra_info['original_url'])
1502 self.add_default_extra_info(ie_result, ie, url)
1503 if process:
1504 self._wait_for_video(ie_result)
1505 return self.process_ie_result(ie_result, download, extra_info)
1506 else:
1507 return ie_result
1508
1509 def add_default_extra_info(self, ie_result, ie, url):
1510 if url is not None:
1511 self.add_extra_info(ie_result, {
1512 'webpage_url': url,
1513 'original_url': url,
1514 })
1515 webpage_url = ie_result.get('webpage_url')
1516 if webpage_url:
1517 self.add_extra_info(ie_result, {
1518 'webpage_url_basename': url_basename(webpage_url),
1519 'webpage_url_domain': get_domain(webpage_url),
1520 })
1521 if ie is not None:
1522 self.add_extra_info(ie_result, {
1523 'extractor': ie.IE_NAME,
1524 'extractor_key': ie.ie_key(),
1525 })
1526
1527 def process_ie_result(self, ie_result, download=True, extra_info=None):
1528 """
1529 Take the result of the ie(may be modified) and resolve all unresolved
1530 references (URLs, playlist items).
1531
1532 It will also download the videos if 'download'.
1533 Returns the resolved ie_result.
1534 """
1535 if extra_info is None:
1536 extra_info = {}
1537 result_type = ie_result.get('_type', 'video')
1538
1539 if result_type in ('url', 'url_transparent'):
1540 ie_result['url'] = sanitize_url(ie_result['url'])
1541 if ie_result.get('original_url'):
1542 extra_info.setdefault('original_url', ie_result['original_url'])
1543
1544 extract_flat = self.params.get('extract_flat', False)
1545 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1546 or extract_flat is True):
1547 info_copy = ie_result.copy()
1548 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1549 if ie and not ie_result.get('id'):
1550 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1551 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1552 self.add_extra_info(info_copy, extra_info)
1553 info_copy, _ = self.pre_process(info_copy)
1554 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1555 self._raise_pending_errors(info_copy)
1556 if self.params.get('force_write_download_archive', False):
1557 self.record_download_archive(info_copy)
1558 return ie_result
1559
1560 if result_type == 'video':
1561 self.add_extra_info(ie_result, extra_info)
1562 ie_result = self.process_video_result(ie_result, download=download)
1563 self._raise_pending_errors(ie_result)
1564 additional_urls = (ie_result or {}).get('additional_urls')
1565 if additional_urls:
1566 # TODO: Improve MetadataParserPP to allow setting a list
1567 if isinstance(additional_urls, compat_str):
1568 additional_urls = [additional_urls]
1569 self.to_screen(
1570 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1571 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1572 ie_result['additional_entries'] = [
1573 self.extract_info(
1574 url, download, extra_info=extra_info,
1575 force_generic_extractor=self.params.get('force_generic_extractor'))
1576 for url in additional_urls
1577 ]
1578 return ie_result
1579 elif result_type == 'url':
1580 # We have to add extra_info to the results because it may be
1581 # contained in a playlist
1582 return self.extract_info(
1583 ie_result['url'], download,
1584 ie_key=ie_result.get('ie_key'),
1585 extra_info=extra_info)
1586 elif result_type == 'url_transparent':
1587 # Use the information from the embedding page
1588 info = self.extract_info(
1589 ie_result['url'], ie_key=ie_result.get('ie_key'),
1590 extra_info=extra_info, download=False, process=False)
1591
1592 # extract_info may return None when ignoreerrors is enabled and
1593 # extraction failed with an error, don't crash and return early
1594 # in this case
1595 if not info:
1596 return info
1597
1598 new_result = info.copy()
1599 new_result.update(filter_dict(ie_result, lambda k, v: (
1600 v is not None and k not in {'_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'})))
1601
1602 # Extracted info may not be a video result (i.e.
1603 # info.get('_type', 'video') != video) but rather an url or
1604 # url_transparent. In such cases outer metadata (from ie_result)
1605 # should be propagated to inner one (info). For this to happen
1606 # _type of info should be overridden with url_transparent. This
1607 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1608 if new_result.get('_type') == 'url':
1609 new_result['_type'] = 'url_transparent'
1610
1611 return self.process_ie_result(
1612 new_result, download=download, extra_info=extra_info)
1613 elif result_type in ('playlist', 'multi_video'):
1614 # Protect from infinite recursion due to recursively nested playlists
1615 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1616 webpage_url = ie_result['webpage_url']
1617 if webpage_url in self._playlist_urls:
1618 self.to_screen(
1619 '[download] Skipping already downloaded playlist: %s'
1620 % ie_result.get('title') or ie_result.get('id'))
1621 return
1622
1623 self._playlist_level += 1
1624 self._playlist_urls.add(webpage_url)
1625 self._fill_common_fields(ie_result, False)
1626 self._sanitize_thumbnails(ie_result)
1627 try:
1628 return self.__process_playlist(ie_result, download)
1629 finally:
1630 self._playlist_level -= 1
1631 if not self._playlist_level:
1632 self._playlist_urls.clear()
1633 elif result_type == 'compat_list':
1634 self.report_warning(
1635 'Extractor %s returned a compat_list result. '
1636 'It needs to be updated.' % ie_result.get('extractor'))
1637
1638 def _fixup(r):
1639 self.add_extra_info(r, {
1640 'extractor': ie_result['extractor'],
1641 'webpage_url': ie_result['webpage_url'],
1642 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1643 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1644 'extractor_key': ie_result['extractor_key'],
1645 })
1646 return r
1647 ie_result['entries'] = [
1648 self.process_ie_result(_fixup(r), download, extra_info)
1649 for r in ie_result['entries']
1650 ]
1651 return ie_result
1652 else:
1653 raise Exception('Invalid result type: %s' % result_type)
1654
1655 def _ensure_dir_exists(self, path):
1656 return make_dir(path, self.report_error)
1657
1658 @staticmethod
1659 def _playlist_infodict(ie_result, **kwargs):
1660 return {
1661 **ie_result,
1662 'playlist': ie_result.get('title') or ie_result.get('id'),
1663 'playlist_id': ie_result.get('id'),
1664 'playlist_title': ie_result.get('title'),
1665 'playlist_uploader': ie_result.get('uploader'),
1666 'playlist_uploader_id': ie_result.get('uploader_id'),
1667 'playlist_index': 0,
1668 **kwargs,
1669 }
1670
1671 def __process_playlist(self, ie_result, download):
1672 """Process each entry in the playlist"""
1673 title = ie_result.get('title') or ie_result.get('id') or '<Untitled>'
1674 self.to_screen(f'[download] Downloading playlist: {title}')
1675
1676 all_entries = PlaylistEntries(self, ie_result)
1677 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1678
1679 lazy = self.params.get('lazy_playlist')
1680 if lazy:
1681 resolved_entries, n_entries = [], 'N/A'
1682 ie_result['requested_entries'], ie_result['entries'] = None, None
1683 else:
1684 entries = resolved_entries = list(entries)
1685 n_entries = len(resolved_entries)
1686 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1687 if not ie_result.get('playlist_count'):
1688 # Better to do this after potentially exhausting entries
1689 ie_result['playlist_count'] = all_entries.get_full_count()
1690
1691 _infojson_written = False
1692 write_playlist_files = self.params.get('allow_playlist_files', True)
1693 if write_playlist_files and self.params.get('list_thumbnails'):
1694 self.list_thumbnails(ie_result)
1695 if write_playlist_files and not self.params.get('simulate'):
1696 ie_copy = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1697 _infojson_written = self._write_info_json(
1698 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1699 if _infojson_written is None:
1700 return
1701 if self._write_description('playlist', ie_result,
1702 self.prepare_filename(ie_copy, 'pl_description')) is None:
1703 return
1704 # TODO: This should be passed to ThumbnailsConvertor if necessary
1705 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1706
1707 if lazy:
1708 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1709 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1710 elif self.params.get('playlistreverse'):
1711 entries.reverse()
1712 elif self.params.get('playlistrandom'):
1713 random.shuffle(entries)
1714
1715 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} videos'
1716 f'{format_field(ie_result, "playlist_count", " of %s")}')
1717
1718 failures = 0
1719 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1720 for i, (playlist_index, entry) in enumerate(entries):
1721 if lazy:
1722 resolved_entries.append((playlist_index, entry))
1723
1724 # TODO: Add auto-generated fields
1725 if self._match_entry(entry, incomplete=True) is not None:
1726 continue
1727
1728 self.to_screen('[download] Downloading video %s of %s' % (
1729 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1730
1731 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1732 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1733 playlist_index = ie_result['requested_entries'][i]
1734
1735 entry_result = self.__process_iterable_entry(entry, download, {
1736 'n_entries': int_or_none(n_entries),
1737 '__last_playlist_index': max(ie_result['requested_entries'] or (0, 0)),
1738 'playlist_count': ie_result.get('playlist_count'),
1739 'playlist_index': playlist_index,
1740 'playlist_autonumber': i + 1,
1741 'playlist': title,
1742 'playlist_id': ie_result.get('id'),
1743 'playlist_title': ie_result.get('title'),
1744 'playlist_uploader': ie_result.get('uploader'),
1745 'playlist_uploader_id': ie_result.get('uploader_id'),
1746 'extractor': ie_result['extractor'],
1747 'webpage_url': ie_result['webpage_url'],
1748 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1749 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1750 'extractor_key': ie_result['extractor_key'],
1751 })
1752 if not entry_result:
1753 failures += 1
1754 if failures >= max_failures:
1755 self.report_error(
1756 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1757 break
1758 resolved_entries[i] = (playlist_index, entry_result)
1759
1760 # Update with processed data
1761 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1762
1763 # Write the updated info to json
1764 if _infojson_written is True and self._write_info_json(
1765 'updated playlist', ie_result,
1766 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1767 return
1768
1769 ie_result = self.run_all_pps('playlist', ie_result)
1770 self.to_screen(f'[download] Finished downloading playlist: {title}')
1771 return ie_result
1772
1773 @_handle_extraction_exceptions
1774 def __process_iterable_entry(self, entry, download, extra_info):
1775 return self.process_ie_result(
1776 entry, download=download, extra_info=extra_info)
1777
1778 def _build_format_filter(self, filter_spec):
1779 " Returns a function to filter the formats according to the filter_spec "
1780
1781 OPERATORS = {
1782 '<': operator.lt,
1783 '<=': operator.le,
1784 '>': operator.gt,
1785 '>=': operator.ge,
1786 '=': operator.eq,
1787 '!=': operator.ne,
1788 }
1789 operator_rex = re.compile(r'''(?x)\s*
1790 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1791 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1792 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1793 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1794 m = operator_rex.fullmatch(filter_spec)
1795 if m:
1796 try:
1797 comparison_value = int(m.group('value'))
1798 except ValueError:
1799 comparison_value = parse_filesize(m.group('value'))
1800 if comparison_value is None:
1801 comparison_value = parse_filesize(m.group('value') + 'B')
1802 if comparison_value is None:
1803 raise ValueError(
1804 'Invalid value %r in format specification %r' % (
1805 m.group('value'), filter_spec))
1806 op = OPERATORS[m.group('op')]
1807
1808 if not m:
1809 STR_OPERATORS = {
1810 '=': operator.eq,
1811 '^=': lambda attr, value: attr.startswith(value),
1812 '$=': lambda attr, value: attr.endswith(value),
1813 '*=': lambda attr, value: value in attr,
1814 '~=': lambda attr, value: value.search(attr) is not None
1815 }
1816 str_operator_rex = re.compile(r'''(?x)\s*
1817 (?P<key>[a-zA-Z0-9._-]+)\s*
1818 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1819 (?P<quote>["'])?
1820 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1821 (?(quote)(?P=quote))\s*
1822 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1823 m = str_operator_rex.fullmatch(filter_spec)
1824 if m:
1825 if m.group('op') == '~=':
1826 comparison_value = re.compile(m.group('value'))
1827 else:
1828 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1829 str_op = STR_OPERATORS[m.group('op')]
1830 if m.group('negation'):
1831 op = lambda attr, value: not str_op(attr, value)
1832 else:
1833 op = str_op
1834
1835 if not m:
1836 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1837
1838 def _filter(f):
1839 actual_value = f.get(m.group('key'))
1840 if actual_value is None:
1841 return m.group('none_inclusive')
1842 return op(actual_value, comparison_value)
1843 return _filter
1844
1845 def _check_formats(self, formats):
1846 for f in formats:
1847 self.to_screen('[info] Testing format %s' % f['format_id'])
1848 path = self.get_output_path('temp')
1849 if not self._ensure_dir_exists(f'{path}/'):
1850 continue
1851 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1852 temp_file.close()
1853 try:
1854 success, _ = self.dl(temp_file.name, f, test=True)
1855 except (DownloadError, OSError, ValueError) + network_exceptions:
1856 success = False
1857 finally:
1858 if os.path.exists(temp_file.name):
1859 try:
1860 os.remove(temp_file.name)
1861 except OSError:
1862 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1863 if success:
1864 yield f
1865 else:
1866 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1867
1868 def _default_format_spec(self, info_dict, download=True):
1869
1870 def can_merge():
1871 merger = FFmpegMergerPP(self)
1872 return merger.available and merger.can_merge()
1873
1874 prefer_best = (
1875 not self.params.get('simulate')
1876 and download
1877 and (
1878 not can_merge()
1879 or info_dict.get('is_live') and not self.params.get('live_from_start')
1880 or self.params['outtmpl']['default'] == '-'))
1881 compat = (
1882 prefer_best
1883 or self.params.get('allow_multiple_audio_streams', False)
1884 or 'format-spec' in self.params['compat_opts'])
1885
1886 return (
1887 'best/bestvideo+bestaudio' if prefer_best
1888 else 'bestvideo*+bestaudio/best' if not compat
1889 else 'bestvideo+bestaudio/best')
1890
1891 def build_format_selector(self, format_spec):
1892 def syntax_error(note, start):
1893 message = (
1894 'Invalid format specification: '
1895 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1896 return SyntaxError(message)
1897
1898 PICKFIRST = 'PICKFIRST'
1899 MERGE = 'MERGE'
1900 SINGLE = 'SINGLE'
1901 GROUP = 'GROUP'
1902 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1903
1904 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1905 'video': self.params.get('allow_multiple_video_streams', False)}
1906
1907 check_formats = self.params.get('check_formats') == 'selected'
1908
1909 def _parse_filter(tokens):
1910 filter_parts = []
1911 for type, string, start, _, _ in tokens:
1912 if type == tokenize.OP and string == ']':
1913 return ''.join(filter_parts)
1914 else:
1915 filter_parts.append(string)
1916
1917 def _remove_unused_ops(tokens):
1918 # Remove operators that we don't use and join them with the surrounding strings
1919 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1920 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1921 last_string, last_start, last_end, last_line = None, None, None, None
1922 for type, string, start, end, line in tokens:
1923 if type == tokenize.OP and string == '[':
1924 if last_string:
1925 yield tokenize.NAME, last_string, last_start, last_end, last_line
1926 last_string = None
1927 yield type, string, start, end, line
1928 # everything inside brackets will be handled by _parse_filter
1929 for type, string, start, end, line in tokens:
1930 yield type, string, start, end, line
1931 if type == tokenize.OP and string == ']':
1932 break
1933 elif type == tokenize.OP and string in ALLOWED_OPS:
1934 if last_string:
1935 yield tokenize.NAME, last_string, last_start, last_end, last_line
1936 last_string = None
1937 yield type, string, start, end, line
1938 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1939 if not last_string:
1940 last_string = string
1941 last_start = start
1942 last_end = end
1943 else:
1944 last_string += string
1945 if last_string:
1946 yield tokenize.NAME, last_string, last_start, last_end, last_line
1947
1948 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1949 selectors = []
1950 current_selector = None
1951 for type, string, start, _, _ in tokens:
1952 # ENCODING is only defined in python 3.x
1953 if type == getattr(tokenize, 'ENCODING', None):
1954 continue
1955 elif type in [tokenize.NAME, tokenize.NUMBER]:
1956 current_selector = FormatSelector(SINGLE, string, [])
1957 elif type == tokenize.OP:
1958 if string == ')':
1959 if not inside_group:
1960 # ')' will be handled by the parentheses group
1961 tokens.restore_last_token()
1962 break
1963 elif inside_merge and string in ['/', ',']:
1964 tokens.restore_last_token()
1965 break
1966 elif inside_choice and string == ',':
1967 tokens.restore_last_token()
1968 break
1969 elif string == ',':
1970 if not current_selector:
1971 raise syntax_error('"," must follow a format selector', start)
1972 selectors.append(current_selector)
1973 current_selector = None
1974 elif string == '/':
1975 if not current_selector:
1976 raise syntax_error('"/" must follow a format selector', start)
1977 first_choice = current_selector
1978 second_choice = _parse_format_selection(tokens, inside_choice=True)
1979 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1980 elif string == '[':
1981 if not current_selector:
1982 current_selector = FormatSelector(SINGLE, 'best', [])
1983 format_filter = _parse_filter(tokens)
1984 current_selector.filters.append(format_filter)
1985 elif string == '(':
1986 if current_selector:
1987 raise syntax_error('Unexpected "("', start)
1988 group = _parse_format_selection(tokens, inside_group=True)
1989 current_selector = FormatSelector(GROUP, group, [])
1990 elif string == '+':
1991 if not current_selector:
1992 raise syntax_error('Unexpected "+"', start)
1993 selector_1 = current_selector
1994 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1995 if not selector_2:
1996 raise syntax_error('Expected a selector', start)
1997 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1998 else:
1999 raise syntax_error(f'Operator not recognized: "{string}"', start)
2000 elif type == tokenize.ENDMARKER:
2001 break
2002 if current_selector:
2003 selectors.append(current_selector)
2004 return selectors
2005
2006 def _merge(formats_pair):
2007 format_1, format_2 = formats_pair
2008
2009 formats_info = []
2010 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2011 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2012
2013 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2014 get_no_more = {'video': False, 'audio': False}
2015 for (i, fmt_info) in enumerate(formats_info):
2016 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2017 formats_info.pop(i)
2018 continue
2019 for aud_vid in ['audio', 'video']:
2020 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2021 if get_no_more[aud_vid]:
2022 formats_info.pop(i)
2023 break
2024 get_no_more[aud_vid] = True
2025
2026 if len(formats_info) == 1:
2027 return formats_info[0]
2028
2029 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2030 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2031
2032 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2033 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2034
2035 output_ext = self.params.get('merge_output_format')
2036 if not output_ext:
2037 if the_only_video:
2038 output_ext = the_only_video['ext']
2039 elif the_only_audio and not video_fmts:
2040 output_ext = the_only_audio['ext']
2041 else:
2042 output_ext = 'mkv'
2043
2044 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2045
2046 new_dict = {
2047 'requested_formats': formats_info,
2048 'format': '+'.join(filtered('format')),
2049 'format_id': '+'.join(filtered('format_id')),
2050 'ext': output_ext,
2051 'protocol': '+'.join(map(determine_protocol, formats_info)),
2052 'language': '+'.join(orderedSet(filtered('language'))) or None,
2053 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2054 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2055 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2056 }
2057
2058 if the_only_video:
2059 new_dict.update({
2060 'width': the_only_video.get('width'),
2061 'height': the_only_video.get('height'),
2062 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2063 'fps': the_only_video.get('fps'),
2064 'dynamic_range': the_only_video.get('dynamic_range'),
2065 'vcodec': the_only_video.get('vcodec'),
2066 'vbr': the_only_video.get('vbr'),
2067 'stretched_ratio': the_only_video.get('stretched_ratio'),
2068 })
2069
2070 if the_only_audio:
2071 new_dict.update({
2072 'acodec': the_only_audio.get('acodec'),
2073 'abr': the_only_audio.get('abr'),
2074 'asr': the_only_audio.get('asr'),
2075 })
2076
2077 return new_dict
2078
2079 def _check_formats(formats):
2080 if not check_formats:
2081 yield from formats
2082 return
2083 yield from self._check_formats(formats)
2084
2085 def _build_selector_function(selector):
2086 if isinstance(selector, list): # ,
2087 fs = [_build_selector_function(s) for s in selector]
2088
2089 def selector_function(ctx):
2090 for f in fs:
2091 yield from f(ctx)
2092 return selector_function
2093
2094 elif selector.type == GROUP: # ()
2095 selector_function = _build_selector_function(selector.selector)
2096
2097 elif selector.type == PICKFIRST: # /
2098 fs = [_build_selector_function(s) for s in selector.selector]
2099
2100 def selector_function(ctx):
2101 for f in fs:
2102 picked_formats = list(f(ctx))
2103 if picked_formats:
2104 return picked_formats
2105 return []
2106
2107 elif selector.type == MERGE: # +
2108 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2109
2110 def selector_function(ctx):
2111 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2112 yield _merge(pair)
2113
2114 elif selector.type == SINGLE: # atom
2115 format_spec = selector.selector or 'best'
2116
2117 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2118 if format_spec == 'all':
2119 def selector_function(ctx):
2120 yield from _check_formats(ctx['formats'][::-1])
2121 elif format_spec == 'mergeall':
2122 def selector_function(ctx):
2123 formats = list(_check_formats(
2124 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2125 if not formats:
2126 return
2127 merged_format = formats[-1]
2128 for f in formats[-2::-1]:
2129 merged_format = _merge((merged_format, f))
2130 yield merged_format
2131
2132 else:
2133 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2134 mobj = re.match(
2135 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2136 format_spec)
2137 if mobj is not None:
2138 format_idx = int_or_none(mobj.group('n'), default=1)
2139 format_reverse = mobj.group('bw')[0] == 'b'
2140 format_type = (mobj.group('type') or [None])[0]
2141 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2142 format_modified = mobj.group('mod') is not None
2143
2144 format_fallback = not format_type and not format_modified # for b, w
2145 _filter_f = (
2146 (lambda f: f.get('%scodec' % format_type) != 'none')
2147 if format_type and format_modified # bv*, ba*, wv*, wa*
2148 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2149 if format_type # bv, ba, wv, wa
2150 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2151 if not format_modified # b, w
2152 else lambda f: True) # b*, w*
2153 filter_f = lambda f: _filter_f(f) and (
2154 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2155 else:
2156 if format_spec in self._format_selection_exts['audio']:
2157 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2158 elif format_spec in self._format_selection_exts['video']:
2159 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2160 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2161 elif format_spec in self._format_selection_exts['storyboards']:
2162 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2163 else:
2164 filter_f = lambda f: f.get('format_id') == format_spec # id
2165
2166 def selector_function(ctx):
2167 formats = list(ctx['formats'])
2168 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2169 if not matches:
2170 if format_fallback and ctx['incomplete_formats']:
2171 # for extractors with incomplete formats (audio only (soundcloud)
2172 # or video only (imgur)) best/worst will fallback to
2173 # best/worst {video,audio}-only format
2174 matches = formats
2175 elif seperate_fallback and not ctx['has_merged_format']:
2176 # for compatibility with youtube-dl when there is no pre-merged format
2177 matches = list(filter(seperate_fallback, formats))
2178 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2179 try:
2180 yield matches[format_idx - 1]
2181 except LazyList.IndexError:
2182 return
2183
2184 filters = [self._build_format_filter(f) for f in selector.filters]
2185
2186 def final_selector(ctx):
2187 ctx_copy = dict(ctx)
2188 for _filter in filters:
2189 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2190 return selector_function(ctx_copy)
2191 return final_selector
2192
2193 stream = io.BytesIO(format_spec.encode())
2194 try:
2195 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2196 except tokenize.TokenError:
2197 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2198
2199 class TokenIterator:
2200 def __init__(self, tokens):
2201 self.tokens = tokens
2202 self.counter = 0
2203
2204 def __iter__(self):
2205 return self
2206
2207 def __next__(self):
2208 if self.counter >= len(self.tokens):
2209 raise StopIteration()
2210 value = self.tokens[self.counter]
2211 self.counter += 1
2212 return value
2213
2214 next = __next__
2215
2216 def restore_last_token(self):
2217 self.counter -= 1
2218
2219 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2220 return _build_selector_function(parsed_selector)
2221
2222 def _calc_headers(self, info_dict):
2223 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2224
2225 cookies = self._calc_cookies(info_dict['url'])
2226 if cookies:
2227 res['Cookie'] = cookies
2228
2229 if 'X-Forwarded-For' not in res:
2230 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2231 if x_forwarded_for_ip:
2232 res['X-Forwarded-For'] = x_forwarded_for_ip
2233
2234 return res
2235
2236 def _calc_cookies(self, url):
2237 pr = sanitized_Request(url)
2238 self.cookiejar.add_cookie_header(pr)
2239 return pr.get_header('Cookie')
2240
2241 def _sort_thumbnails(self, thumbnails):
2242 thumbnails.sort(key=lambda t: (
2243 t.get('preference') if t.get('preference') is not None else -1,
2244 t.get('width') if t.get('width') is not None else -1,
2245 t.get('height') if t.get('height') is not None else -1,
2246 t.get('id') if t.get('id') is not None else '',
2247 t.get('url')))
2248
2249 def _sanitize_thumbnails(self, info_dict):
2250 thumbnails = info_dict.get('thumbnails')
2251 if thumbnails is None:
2252 thumbnail = info_dict.get('thumbnail')
2253 if thumbnail:
2254 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2255 if not thumbnails:
2256 return
2257
2258 def check_thumbnails(thumbnails):
2259 for t in thumbnails:
2260 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2261 try:
2262 self.urlopen(HEADRequest(t['url']))
2263 except network_exceptions as err:
2264 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2265 continue
2266 yield t
2267
2268 self._sort_thumbnails(thumbnails)
2269 for i, t in enumerate(thumbnails):
2270 if t.get('id') is None:
2271 t['id'] = '%d' % i
2272 if t.get('width') and t.get('height'):
2273 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2274 t['url'] = sanitize_url(t['url'])
2275
2276 if self.params.get('check_formats') is True:
2277 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2278 else:
2279 info_dict['thumbnails'] = thumbnails
2280
2281 def _fill_common_fields(self, info_dict, is_video=True):
2282 # TODO: move sanitization here
2283 if is_video:
2284 # playlists are allowed to lack "title"
2285 title = info_dict.get('title', NO_DEFAULT)
2286 if title is NO_DEFAULT:
2287 raise ExtractorError('Missing "title" field in extractor result',
2288 video_id=info_dict['id'], ie=info_dict['extractor'])
2289 info_dict['fulltitle'] = title
2290 if not title:
2291 if title == '':
2292 self.write_debug('Extractor gave empty title. Creating a generic title')
2293 else:
2294 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2295 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2296
2297 if info_dict.get('duration') is not None:
2298 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2299
2300 for ts_key, date_key in (
2301 ('timestamp', 'upload_date'),
2302 ('release_timestamp', 'release_date'),
2303 ('modified_timestamp', 'modified_date'),
2304 ):
2305 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2306 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2307 # see http://bugs.python.org/issue1646728)
2308 with contextlib.suppress(ValueError, OverflowError, OSError):
2309 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2310 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2311
2312 live_keys = ('is_live', 'was_live')
2313 live_status = info_dict.get('live_status')
2314 if live_status is None:
2315 for key in live_keys:
2316 if info_dict.get(key) is False:
2317 continue
2318 if info_dict.get(key):
2319 live_status = key
2320 break
2321 if all(info_dict.get(key) is False for key in live_keys):
2322 live_status = 'not_live'
2323 if live_status:
2324 info_dict['live_status'] = live_status
2325 for key in live_keys:
2326 if info_dict.get(key) is None:
2327 info_dict[key] = (live_status == key)
2328
2329 # Auto generate title fields corresponding to the *_number fields when missing
2330 # in order to always have clean titles. This is very common for TV series.
2331 for field in ('chapter', 'season', 'episode'):
2332 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2333 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2334
2335 def _raise_pending_errors(self, info):
2336 err = info.pop('__pending_error', None)
2337 if err:
2338 self.report_error(err, tb=False)
2339
2340 def process_video_result(self, info_dict, download=True):
2341 assert info_dict.get('_type', 'video') == 'video'
2342 self._num_videos += 1
2343
2344 if 'id' not in info_dict:
2345 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2346 elif not info_dict.get('id'):
2347 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2348
2349 def report_force_conversion(field, field_not, conversion):
2350 self.report_warning(
2351 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2352 % (field, field_not, conversion))
2353
2354 def sanitize_string_field(info, string_field):
2355 field = info.get(string_field)
2356 if field is None or isinstance(field, compat_str):
2357 return
2358 report_force_conversion(string_field, 'a string', 'string')
2359 info[string_field] = compat_str(field)
2360
2361 def sanitize_numeric_fields(info):
2362 for numeric_field in self._NUMERIC_FIELDS:
2363 field = info.get(numeric_field)
2364 if field is None or isinstance(field, (int, float)):
2365 continue
2366 report_force_conversion(numeric_field, 'numeric', 'int')
2367 info[numeric_field] = int_or_none(field)
2368
2369 sanitize_string_field(info_dict, 'id')
2370 sanitize_numeric_fields(info_dict)
2371 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2372 self.report_warning('"duration" field is negative, there is an error in extractor')
2373
2374 if 'playlist' not in info_dict:
2375 # It isn't part of a playlist
2376 info_dict['playlist'] = None
2377 info_dict['playlist_index'] = None
2378
2379 self._sanitize_thumbnails(info_dict)
2380
2381 thumbnail = info_dict.get('thumbnail')
2382 thumbnails = info_dict.get('thumbnails')
2383 if thumbnail:
2384 info_dict['thumbnail'] = sanitize_url(thumbnail)
2385 elif thumbnails:
2386 info_dict['thumbnail'] = thumbnails[-1]['url']
2387
2388 if info_dict.get('display_id') is None and 'id' in info_dict:
2389 info_dict['display_id'] = info_dict['id']
2390
2391 self._fill_common_fields(info_dict)
2392
2393 for cc_kind in ('subtitles', 'automatic_captions'):
2394 cc = info_dict.get(cc_kind)
2395 if cc:
2396 for _, subtitle in cc.items():
2397 for subtitle_format in subtitle:
2398 if subtitle_format.get('url'):
2399 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2400 if subtitle_format.get('ext') is None:
2401 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2402
2403 automatic_captions = info_dict.get('automatic_captions')
2404 subtitles = info_dict.get('subtitles')
2405
2406 info_dict['requested_subtitles'] = self.process_subtitles(
2407 info_dict['id'], subtitles, automatic_captions)
2408
2409 if info_dict.get('formats') is None:
2410 # There's only one format available
2411 formats = [info_dict]
2412 else:
2413 formats = info_dict['formats']
2414
2415 # or None ensures --clean-infojson removes it
2416 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2417 if not self.params.get('allow_unplayable_formats'):
2418 formats = [f for f in formats if not f.get('has_drm')]
2419 if info_dict['_has_drm'] and all(
2420 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2421 self.report_warning(
2422 'This video is DRM protected and only images are available for download. '
2423 'Use --list-formats to see them')
2424
2425 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2426 if not get_from_start:
2427 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2428 if info_dict.get('is_live') and formats:
2429 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2430 if get_from_start and not formats:
2431 self.raise_no_formats(info_dict, msg=(
2432 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2433 'If you want to download from the current time, use --no-live-from-start'))
2434
2435 if not formats:
2436 self.raise_no_formats(info_dict)
2437
2438 def is_wellformed(f):
2439 url = f.get('url')
2440 if not url:
2441 self.report_warning(
2442 '"url" field is missing or empty - skipping format, '
2443 'there is an error in extractor')
2444 return False
2445 if isinstance(url, bytes):
2446 sanitize_string_field(f, 'url')
2447 return True
2448
2449 # Filter out malformed formats for better extraction robustness
2450 formats = list(filter(is_wellformed, formats))
2451
2452 formats_dict = {}
2453
2454 # We check that all the formats have the format and format_id fields
2455 for i, format in enumerate(formats):
2456 sanitize_string_field(format, 'format_id')
2457 sanitize_numeric_fields(format)
2458 format['url'] = sanitize_url(format['url'])
2459 if not format.get('format_id'):
2460 format['format_id'] = compat_str(i)
2461 else:
2462 # Sanitize format_id from characters used in format selector expression
2463 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2464 format_id = format['format_id']
2465 if format_id not in formats_dict:
2466 formats_dict[format_id] = []
2467 formats_dict[format_id].append(format)
2468
2469 # Make sure all formats have unique format_id
2470 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2471 for format_id, ambiguous_formats in formats_dict.items():
2472 ambigious_id = len(ambiguous_formats) > 1
2473 for i, format in enumerate(ambiguous_formats):
2474 if ambigious_id:
2475 format['format_id'] = '%s-%d' % (format_id, i)
2476 if format.get('ext') is None:
2477 format['ext'] = determine_ext(format['url']).lower()
2478 # Ensure there is no conflict between id and ext in format selection
2479 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2480 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2481 format['format_id'] = 'f%s' % format['format_id']
2482
2483 for i, format in enumerate(formats):
2484 if format.get('format') is None:
2485 format['format'] = '{id} - {res}{note}'.format(
2486 id=format['format_id'],
2487 res=self.format_resolution(format),
2488 note=format_field(format, 'format_note', ' (%s)'),
2489 )
2490 if format.get('protocol') is None:
2491 format['protocol'] = determine_protocol(format)
2492 if format.get('resolution') is None:
2493 format['resolution'] = self.format_resolution(format, default=None)
2494 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2495 format['dynamic_range'] = 'SDR'
2496 if (info_dict.get('duration') and format.get('tbr')
2497 and not format.get('filesize') and not format.get('filesize_approx')):
2498 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2499
2500 # Add HTTP headers, so that external programs can use them from the
2501 # json output
2502 full_format_info = info_dict.copy()
2503 full_format_info.update(format)
2504 format['http_headers'] = self._calc_headers(full_format_info)
2505 # Remove private housekeeping stuff
2506 if '__x_forwarded_for_ip' in info_dict:
2507 del info_dict['__x_forwarded_for_ip']
2508
2509 if self.params.get('check_formats') is True:
2510 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2511
2512 if not formats or formats[0] is not info_dict:
2513 # only set the 'formats' fields if the original info_dict list them
2514 # otherwise we end up with a circular reference, the first (and unique)
2515 # element in the 'formats' field in info_dict is info_dict itself,
2516 # which can't be exported to json
2517 info_dict['formats'] = formats
2518
2519 info_dict, _ = self.pre_process(info_dict)
2520
2521 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2522 return info_dict
2523
2524 self.post_extract(info_dict)
2525 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2526
2527 # The pre-processors may have modified the formats
2528 formats = info_dict.get('formats', [info_dict])
2529
2530 list_only = self.params.get('simulate') is None and (
2531 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2532 interactive_format_selection = not list_only and self.format_selector == '-'
2533 if self.params.get('list_thumbnails'):
2534 self.list_thumbnails(info_dict)
2535 if self.params.get('listsubtitles'):
2536 if 'automatic_captions' in info_dict:
2537 self.list_subtitles(
2538 info_dict['id'], automatic_captions, 'automatic captions')
2539 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2540 if self.params.get('listformats') or interactive_format_selection:
2541 self.list_formats(info_dict)
2542 if list_only:
2543 # Without this printing, -F --print-json will not work
2544 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2545 return info_dict
2546
2547 format_selector = self.format_selector
2548 if format_selector is None:
2549 req_format = self._default_format_spec(info_dict, download=download)
2550 self.write_debug('Default format spec: %s' % req_format)
2551 format_selector = self.build_format_selector(req_format)
2552
2553 while True:
2554 if interactive_format_selection:
2555 req_format = input(
2556 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2557 try:
2558 format_selector = self.build_format_selector(req_format)
2559 except SyntaxError as err:
2560 self.report_error(err, tb=False, is_error=False)
2561 continue
2562
2563 formats_to_download = list(format_selector({
2564 'formats': formats,
2565 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2566 'incomplete_formats': (
2567 # All formats are video-only or
2568 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2569 # all formats are audio-only
2570 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2571 }))
2572 if interactive_format_selection and not formats_to_download:
2573 self.report_error('Requested format is not available', tb=False, is_error=False)
2574 continue
2575 break
2576
2577 if not formats_to_download:
2578 if not self.params.get('ignore_no_formats_error'):
2579 raise ExtractorError(
2580 'Requested format is not available. Use --list-formats for a list of available formats',
2581 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2582 self.report_warning('Requested format is not available')
2583 # Process what we can, even without any available formats.
2584 formats_to_download = [{}]
2585
2586 requested_ranges = self.params.get('download_ranges')
2587 if requested_ranges:
2588 requested_ranges = tuple(requested_ranges(info_dict, self))
2589
2590 best_format, downloaded_formats = formats_to_download[-1], []
2591 if download:
2592 if best_format:
2593 def to_screen(*msg):
2594 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2595
2596 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2597 (f['format_id'] for f in formats_to_download))
2598 if requested_ranges:
2599 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2600 (f'{int(c["start_time"])}-{int(c["end_time"])}' for c in requested_ranges))
2601 max_downloads_reached = False
2602
2603 for fmt, chapter in itertools.product(formats_to_download, requested_ranges or [{}]):
2604 new_info = self._copy_infodict(info_dict)
2605 new_info.update(fmt)
2606 if chapter:
2607 new_info.update({
2608 'section_start': chapter.get('start_time'),
2609 'section_end': chapter.get('end_time', 0),
2610 'section_title': chapter.get('title'),
2611 'section_number': chapter.get('index'),
2612 })
2613 downloaded_formats.append(new_info)
2614 try:
2615 self.process_info(new_info)
2616 except MaxDownloadsReached:
2617 max_downloads_reached = True
2618 self._raise_pending_errors(new_info)
2619 # Remove copied info
2620 for key, val in tuple(new_info.items()):
2621 if info_dict.get(key) == val:
2622 new_info.pop(key)
2623 if max_downloads_reached:
2624 break
2625
2626 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2627 assert write_archive.issubset({True, False, 'ignore'})
2628 if True in write_archive and False not in write_archive:
2629 self.record_download_archive(info_dict)
2630
2631 info_dict['requested_downloads'] = downloaded_formats
2632 info_dict = self.run_all_pps('after_video', info_dict)
2633 if max_downloads_reached:
2634 raise MaxDownloadsReached()
2635
2636 # We update the info dict with the selected best quality format (backwards compatibility)
2637 info_dict.update(best_format)
2638 return info_dict
2639
2640 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2641 """Select the requested subtitles and their format"""
2642 available_subs, normal_sub_langs = {}, []
2643 if normal_subtitles and self.params.get('writesubtitles'):
2644 available_subs.update(normal_subtitles)
2645 normal_sub_langs = tuple(normal_subtitles.keys())
2646 if automatic_captions and self.params.get('writeautomaticsub'):
2647 for lang, cap_info in automatic_captions.items():
2648 if lang not in available_subs:
2649 available_subs[lang] = cap_info
2650
2651 if (not self.params.get('writesubtitles') and not
2652 self.params.get('writeautomaticsub') or not
2653 available_subs):
2654 return None
2655
2656 all_sub_langs = tuple(available_subs.keys())
2657 if self.params.get('allsubtitles', False):
2658 requested_langs = all_sub_langs
2659 elif self.params.get('subtitleslangs', False):
2660 # A list is used so that the order of languages will be the same as
2661 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2662 requested_langs = []
2663 for lang_re in self.params.get('subtitleslangs'):
2664 discard = lang_re[0] == '-'
2665 if discard:
2666 lang_re = lang_re[1:]
2667 if lang_re == 'all':
2668 if discard:
2669 requested_langs = []
2670 else:
2671 requested_langs.extend(all_sub_langs)
2672 continue
2673 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2674 if discard:
2675 for lang in current_langs:
2676 while lang in requested_langs:
2677 requested_langs.remove(lang)
2678 else:
2679 requested_langs.extend(current_langs)
2680 requested_langs = orderedSet(requested_langs)
2681 elif normal_sub_langs:
2682 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2683 else:
2684 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2685 if requested_langs:
2686 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2687
2688 formats_query = self.params.get('subtitlesformat', 'best')
2689 formats_preference = formats_query.split('/') if formats_query else []
2690 subs = {}
2691 for lang in requested_langs:
2692 formats = available_subs.get(lang)
2693 if formats is None:
2694 self.report_warning(f'{lang} subtitles not available for {video_id}')
2695 continue
2696 for ext in formats_preference:
2697 if ext == 'best':
2698 f = formats[-1]
2699 break
2700 matches = list(filter(lambda f: f['ext'] == ext, formats))
2701 if matches:
2702 f = matches[-1]
2703 break
2704 else:
2705 f = formats[-1]
2706 self.report_warning(
2707 'No subtitle format found matching "%s" for language %s, '
2708 'using %s' % (formats_query, lang, f['ext']))
2709 subs[lang] = f
2710 return subs
2711
2712 def _forceprint(self, key, info_dict):
2713 if info_dict is None:
2714 return
2715 info_copy = info_dict.copy()
2716 info_copy['formats_table'] = self.render_formats_table(info_dict)
2717 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2718 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2719 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2720
2721 def format_tmpl(tmpl):
2722 mobj = re.match(r'\w+(=?)$', tmpl)
2723 if mobj and mobj.group(1):
2724 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2725 elif mobj:
2726 return f'%({tmpl})s'
2727 return tmpl
2728
2729 for tmpl in self.params['forceprint'].get(key, []):
2730 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2731
2732 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2733 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2734 tmpl = format_tmpl(tmpl)
2735 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2736 if self._ensure_dir_exists(filename):
2737 with open(filename, 'a', encoding='utf-8') as f:
2738 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2739
2740 def __forced_printings(self, info_dict, filename, incomplete):
2741 def print_mandatory(field, actual_field=None):
2742 if actual_field is None:
2743 actual_field = field
2744 if (self.params.get('force%s' % field, False)
2745 and (not incomplete or info_dict.get(actual_field) is not None)):
2746 self.to_stdout(info_dict[actual_field])
2747
2748 def print_optional(field):
2749 if (self.params.get('force%s' % field, False)
2750 and info_dict.get(field) is not None):
2751 self.to_stdout(info_dict[field])
2752
2753 info_dict = info_dict.copy()
2754 if filename is not None:
2755 info_dict['filename'] = filename
2756 if info_dict.get('requested_formats') is not None:
2757 # For RTMP URLs, also include the playpath
2758 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2759 elif info_dict.get('url'):
2760 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2761
2762 if (self.params.get('forcejson')
2763 or self.params['forceprint'].get('video')
2764 or self.params['print_to_file'].get('video')):
2765 self.post_extract(info_dict)
2766 self._forceprint('video', info_dict)
2767
2768 print_mandatory('title')
2769 print_mandatory('id')
2770 print_mandatory('url', 'urls')
2771 print_optional('thumbnail')
2772 print_optional('description')
2773 print_optional('filename')
2774 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2775 self.to_stdout(formatSeconds(info_dict['duration']))
2776 print_mandatory('format')
2777
2778 if self.params.get('forcejson'):
2779 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2780
2781 def dl(self, name, info, subtitle=False, test=False):
2782 if not info.get('url'):
2783 self.raise_no_formats(info, True)
2784
2785 if test:
2786 verbose = self.params.get('verbose')
2787 params = {
2788 'test': True,
2789 'quiet': self.params.get('quiet') or not verbose,
2790 'verbose': verbose,
2791 'noprogress': not verbose,
2792 'nopart': True,
2793 'skip_unavailable_fragments': False,
2794 'keep_fragments': False,
2795 'overwrites': True,
2796 '_no_ytdl_file': True,
2797 }
2798 else:
2799 params = self.params
2800 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2801 if not test:
2802 for ph in self._progress_hooks:
2803 fd.add_progress_hook(ph)
2804 urls = '", "'.join(
2805 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2806 for f in info.get('requested_formats', []) or [info])
2807 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2808
2809 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2810 # But it may contain objects that are not deep-copyable
2811 new_info = self._copy_infodict(info)
2812 if new_info.get('http_headers') is None:
2813 new_info['http_headers'] = self._calc_headers(new_info)
2814 return fd.download(name, new_info, subtitle)
2815
2816 def existing_file(self, filepaths, *, default_overwrite=True):
2817 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2818 if existing_files and not self.params.get('overwrites', default_overwrite):
2819 return existing_files[0]
2820
2821 for file in existing_files:
2822 self.report_file_delete(file)
2823 os.remove(file)
2824 return None
2825
2826 def process_info(self, info_dict):
2827 """Process a single resolved IE result. (Modifies it in-place)"""
2828
2829 assert info_dict.get('_type', 'video') == 'video'
2830 original_infodict = info_dict
2831
2832 if 'format' not in info_dict and 'ext' in info_dict:
2833 info_dict['format'] = info_dict['ext']
2834
2835 # This is mostly just for backward compatibility of process_info
2836 # As a side-effect, this allows for format-specific filters
2837 if self._match_entry(info_dict) is not None:
2838 info_dict['__write_download_archive'] = 'ignore'
2839 return
2840
2841 # Does nothing under normal operation - for backward compatibility of process_info
2842 self.post_extract(info_dict)
2843 self._num_downloads += 1
2844
2845 # info_dict['_filename'] needs to be set for backward compatibility
2846 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2847 temp_filename = self.prepare_filename(info_dict, 'temp')
2848 files_to_move = {}
2849
2850 # Forced printings
2851 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2852
2853 def check_max_downloads():
2854 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
2855 raise MaxDownloadsReached()
2856
2857 if self.params.get('simulate'):
2858 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2859 check_max_downloads()
2860 return
2861
2862 if full_filename is None:
2863 return
2864 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2865 return
2866 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2867 return
2868
2869 if self._write_description('video', info_dict,
2870 self.prepare_filename(info_dict, 'description')) is None:
2871 return
2872
2873 sub_files = self._write_subtitles(info_dict, temp_filename)
2874 if sub_files is None:
2875 return
2876 files_to_move.update(dict(sub_files))
2877
2878 thumb_files = self._write_thumbnails(
2879 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2880 if thumb_files is None:
2881 return
2882 files_to_move.update(dict(thumb_files))
2883
2884 infofn = self.prepare_filename(info_dict, 'infojson')
2885 _infojson_written = self._write_info_json('video', info_dict, infofn)
2886 if _infojson_written:
2887 info_dict['infojson_filename'] = infofn
2888 # For backward compatibility, even though it was a private field
2889 info_dict['__infojson_filename'] = infofn
2890 elif _infojson_written is None:
2891 return
2892
2893 # Note: Annotations are deprecated
2894 annofn = None
2895 if self.params.get('writeannotations', False):
2896 annofn = self.prepare_filename(info_dict, 'annotation')
2897 if annofn:
2898 if not self._ensure_dir_exists(encodeFilename(annofn)):
2899 return
2900 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2901 self.to_screen('[info] Video annotations are already present')
2902 elif not info_dict.get('annotations'):
2903 self.report_warning('There are no annotations to write.')
2904 else:
2905 try:
2906 self.to_screen('[info] Writing video annotations to: ' + annofn)
2907 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2908 annofile.write(info_dict['annotations'])
2909 except (KeyError, TypeError):
2910 self.report_warning('There are no annotations to write.')
2911 except OSError:
2912 self.report_error('Cannot write annotations file: ' + annofn)
2913 return
2914
2915 # Write internet shortcut files
2916 def _write_link_file(link_type):
2917 url = try_get(info_dict['webpage_url'], iri_to_uri)
2918 if not url:
2919 self.report_warning(
2920 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2921 return True
2922 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2923 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2924 return False
2925 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2926 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2927 return True
2928 try:
2929 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2930 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2931 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2932 template_vars = {'url': url}
2933 if link_type == 'desktop':
2934 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2935 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2936 except OSError:
2937 self.report_error(f'Cannot write internet shortcut {linkfn}')
2938 return False
2939 return True
2940
2941 write_links = {
2942 'url': self.params.get('writeurllink'),
2943 'webloc': self.params.get('writewebloclink'),
2944 'desktop': self.params.get('writedesktoplink'),
2945 }
2946 if self.params.get('writelink'):
2947 link_type = ('webloc' if sys.platform == 'darwin'
2948 else 'desktop' if sys.platform.startswith('linux')
2949 else 'url')
2950 write_links[link_type] = True
2951
2952 if any(should_write and not _write_link_file(link_type)
2953 for link_type, should_write in write_links.items()):
2954 return
2955
2956 def replace_info_dict(new_info):
2957 nonlocal info_dict
2958 if new_info == info_dict:
2959 return
2960 info_dict.clear()
2961 info_dict.update(new_info)
2962
2963 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2964 replace_info_dict(new_info)
2965
2966 if self.params.get('skip_download'):
2967 info_dict['filepath'] = temp_filename
2968 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2969 info_dict['__files_to_move'] = files_to_move
2970 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2971 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2972 else:
2973 # Download
2974 info_dict.setdefault('__postprocessors', [])
2975 try:
2976
2977 def existing_video_file(*filepaths):
2978 ext = info_dict.get('ext')
2979 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
2980 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
2981 default_overwrite=False)
2982 if file:
2983 info_dict['ext'] = os.path.splitext(file)[1][1:]
2984 return file
2985
2986 success = True
2987 merger, fd = FFmpegMergerPP(self), None
2988 if info_dict.get('protocol') or info_dict.get('url'):
2989 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
2990 if fd is not FFmpegFD and (
2991 info_dict.get('section_start') or info_dict.get('section_end')):
2992 msg = ('This format cannot be partially downloaded' if merger.available
2993 else 'You have requested downloading the video partially, but ffmpeg is not installed')
2994 self.report_error(f'{msg}. Aborting')
2995 return
2996
2997 if info_dict.get('requested_formats') is not None:
2998
2999 def compatible_formats(formats):
3000 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3001 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3002 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3003 if len(video_formats) > 2 or len(audio_formats) > 2:
3004 return False
3005
3006 # Check extension
3007 exts = {format.get('ext') for format in formats}
3008 COMPATIBLE_EXTS = (
3009 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3010 {'webm'},
3011 )
3012 for ext_sets in COMPATIBLE_EXTS:
3013 if ext_sets.issuperset(exts):
3014 return True
3015 # TODO: Check acodec/vcodec
3016 return False
3017
3018 requested_formats = info_dict['requested_formats']
3019 old_ext = info_dict['ext']
3020 if self.params.get('merge_output_format') is None:
3021 if not compatible_formats(requested_formats):
3022 info_dict['ext'] = 'mkv'
3023 self.report_warning(
3024 'Requested formats are incompatible for merge and will be merged into mkv')
3025 if (info_dict['ext'] == 'webm'
3026 and info_dict.get('thumbnails')
3027 # check with type instead of pp_key, __name__, or isinstance
3028 # since we dont want any custom PPs to trigger this
3029 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3030 info_dict['ext'] = 'mkv'
3031 self.report_warning(
3032 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3033 new_ext = info_dict['ext']
3034
3035 def correct_ext(filename, ext=new_ext):
3036 if filename == '-':
3037 return filename
3038 filename_real_ext = os.path.splitext(filename)[1][1:]
3039 filename_wo_ext = (
3040 os.path.splitext(filename)[0]
3041 if filename_real_ext in (old_ext, new_ext)
3042 else filename)
3043 return f'{filename_wo_ext}.{ext}'
3044
3045 # Ensure filename always has a correct extension for successful merge
3046 full_filename = correct_ext(full_filename)
3047 temp_filename = correct_ext(temp_filename)
3048 dl_filename = existing_video_file(full_filename, temp_filename)
3049 info_dict['__real_download'] = False
3050
3051 downloaded = []
3052 if dl_filename is not None:
3053 self.report_file_already_downloaded(dl_filename)
3054 elif fd:
3055 for f in requested_formats if fd != FFmpegFD else []:
3056 f['filepath'] = fname = prepend_extension(
3057 correct_ext(temp_filename, info_dict['ext']),
3058 'f%s' % f['format_id'], info_dict['ext'])
3059 downloaded.append(fname)
3060 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3061 success, real_download = self.dl(temp_filename, info_dict)
3062 info_dict['__real_download'] = real_download
3063 else:
3064 if self.params.get('allow_unplayable_formats'):
3065 self.report_warning(
3066 'You have requested merging of multiple formats '
3067 'while also allowing unplayable formats to be downloaded. '
3068 'The formats won\'t be merged to prevent data corruption.')
3069 elif not merger.available:
3070 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3071 if not self.params.get('ignoreerrors'):
3072 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3073 return
3074 self.report_warning(f'{msg}. The formats won\'t be merged')
3075
3076 if temp_filename == '-':
3077 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3078 else 'but the formats are incompatible for simultaneous download' if merger.available
3079 else 'but ffmpeg is not installed')
3080 self.report_warning(
3081 f'You have requested downloading multiple formats to stdout {reason}. '
3082 'The formats will be streamed one after the other')
3083 fname = temp_filename
3084 for f in requested_formats:
3085 new_info = dict(info_dict)
3086 del new_info['requested_formats']
3087 new_info.update(f)
3088 if temp_filename != '-':
3089 fname = prepend_extension(
3090 correct_ext(temp_filename, new_info['ext']),
3091 'f%s' % f['format_id'], new_info['ext'])
3092 if not self._ensure_dir_exists(fname):
3093 return
3094 f['filepath'] = fname
3095 downloaded.append(fname)
3096 partial_success, real_download = self.dl(fname, new_info)
3097 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3098 success = success and partial_success
3099
3100 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3101 info_dict['__postprocessors'].append(merger)
3102 info_dict['__files_to_merge'] = downloaded
3103 # Even if there were no downloads, it is being merged only now
3104 info_dict['__real_download'] = True
3105 else:
3106 for file in downloaded:
3107 files_to_move[file] = None
3108 else:
3109 # Just a single file
3110 dl_filename = existing_video_file(full_filename, temp_filename)
3111 if dl_filename is None or dl_filename == temp_filename:
3112 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3113 # So we should try to resume the download
3114 success, real_download = self.dl(temp_filename, info_dict)
3115 info_dict['__real_download'] = real_download
3116 else:
3117 self.report_file_already_downloaded(dl_filename)
3118
3119 dl_filename = dl_filename or temp_filename
3120 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3121
3122 except network_exceptions as err:
3123 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3124 return
3125 except OSError as err:
3126 raise UnavailableVideoError(err)
3127 except (ContentTooShortError, ) as err:
3128 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3129 return
3130
3131 self._raise_pending_errors(info_dict)
3132 if success and full_filename != '-':
3133
3134 def fixup():
3135 do_fixup = True
3136 fixup_policy = self.params.get('fixup')
3137 vid = info_dict['id']
3138
3139 if fixup_policy in ('ignore', 'never'):
3140 return
3141 elif fixup_policy == 'warn':
3142 do_fixup = 'warn'
3143 elif fixup_policy != 'force':
3144 assert fixup_policy in ('detect_or_warn', None)
3145 if not info_dict.get('__real_download'):
3146 do_fixup = False
3147
3148 def ffmpeg_fixup(cndn, msg, cls):
3149 if not (do_fixup and cndn):
3150 return
3151 elif do_fixup == 'warn':
3152 self.report_warning(f'{vid}: {msg}')
3153 return
3154 pp = cls(self)
3155 if pp.available:
3156 info_dict['__postprocessors'].append(pp)
3157 else:
3158 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3159
3160 stretched_ratio = info_dict.get('stretched_ratio')
3161 ffmpeg_fixup(
3162 stretched_ratio not in (1, None),
3163 f'Non-uniform pixel ratio {stretched_ratio}',
3164 FFmpegFixupStretchedPP)
3165
3166 ffmpeg_fixup(
3167 (info_dict.get('requested_formats') is None
3168 and info_dict.get('container') == 'm4a_dash'
3169 and info_dict.get('ext') == 'm4a'),
3170 'writing DASH m4a. Only some players support this container',
3171 FFmpegFixupM4aPP)
3172
3173 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3174 downloader = downloader.FD_NAME if downloader else None
3175
3176 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3177 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3178 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3179 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3180 FFmpegFixupM3u8PP)
3181 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3182 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3183
3184 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3185 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3186
3187 fixup()
3188 try:
3189 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3190 except PostProcessingError as err:
3191 self.report_error('Postprocessing: %s' % str(err))
3192 return
3193 try:
3194 for ph in self._post_hooks:
3195 ph(info_dict['filepath'])
3196 except Exception as err:
3197 self.report_error('post hooks: %s' % str(err))
3198 return
3199 info_dict['__write_download_archive'] = True
3200
3201 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3202 if self.params.get('force_write_download_archive'):
3203 info_dict['__write_download_archive'] = True
3204 check_max_downloads()
3205
3206 def __download_wrapper(self, func):
3207 @functools.wraps(func)
3208 def wrapper(*args, **kwargs):
3209 try:
3210 res = func(*args, **kwargs)
3211 except UnavailableVideoError as e:
3212 self.report_error(e)
3213 except DownloadCancelled as e:
3214 self.to_screen(f'[info] {e}')
3215 if not self.params.get('break_per_url'):
3216 raise
3217 else:
3218 if self.params.get('dump_single_json', False):
3219 self.post_extract(res)
3220 self.to_stdout(json.dumps(self.sanitize_info(res)))
3221 return wrapper
3222
3223 def download(self, url_list):
3224 """Download a given list of URLs."""
3225 url_list = variadic(url_list) # Passing a single URL is a common mistake
3226 outtmpl = self.params['outtmpl']['default']
3227 if (len(url_list) > 1
3228 and outtmpl != '-'
3229 and '%' not in outtmpl
3230 and self.params.get('max_downloads') != 1):
3231 raise SameFileError(outtmpl)
3232
3233 for url in url_list:
3234 self.__download_wrapper(self.extract_info)(
3235 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3236
3237 return self._download_retcode
3238
3239 def download_with_info_file(self, info_filename):
3240 with contextlib.closing(fileinput.FileInput(
3241 [info_filename], mode='r',
3242 openhook=fileinput.hook_encoded('utf-8'))) as f:
3243 # FileInput doesn't have a read method, we can't call json.load
3244 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3245 try:
3246 self.__download_wrapper(self.process_ie_result)(info, download=True)
3247 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3248 if not isinstance(e, EntryNotInPlaylist):
3249 self.to_stderr('\r')
3250 webpage_url = info.get('webpage_url')
3251 if webpage_url is not None:
3252 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3253 return self.download([webpage_url])
3254 else:
3255 raise
3256 return self._download_retcode
3257
3258 @staticmethod
3259 def sanitize_info(info_dict, remove_private_keys=False):
3260 ''' Sanitize the infodict for converting to json '''
3261 if info_dict is None:
3262 return info_dict
3263 info_dict.setdefault('epoch', int(time.time()))
3264 info_dict.setdefault('_type', 'video')
3265
3266 if remove_private_keys:
3267 reject = lambda k, v: v is None or k.startswith('__') or k in {
3268 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3269 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3270 }
3271 else:
3272 reject = lambda k, v: False
3273
3274 def filter_fn(obj):
3275 if isinstance(obj, dict):
3276 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3277 elif isinstance(obj, (list, tuple, set, LazyList)):
3278 return list(map(filter_fn, obj))
3279 elif obj is None or isinstance(obj, (str, int, float, bool)):
3280 return obj
3281 else:
3282 return repr(obj)
3283
3284 return filter_fn(info_dict)
3285
3286 @staticmethod
3287 def filter_requested_info(info_dict, actually_filter=True):
3288 ''' Alias of sanitize_info for backward compatibility '''
3289 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3290
3291 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3292 for filename in set(filter(None, files_to_delete)):
3293 if msg:
3294 self.to_screen(msg % filename)
3295 try:
3296 os.remove(filename)
3297 except OSError:
3298 self.report_warning(f'Unable to delete file {filename}')
3299 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3300 del info['__files_to_move'][filename]
3301
3302 @staticmethod
3303 def post_extract(info_dict):
3304 def actual_post_extract(info_dict):
3305 if info_dict.get('_type') in ('playlist', 'multi_video'):
3306 for video_dict in info_dict.get('entries', {}):
3307 actual_post_extract(video_dict or {})
3308 return
3309
3310 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3311 info_dict.update(post_extractor())
3312
3313 actual_post_extract(info_dict or {})
3314
3315 def run_pp(self, pp, infodict):
3316 files_to_delete = []
3317 if '__files_to_move' not in infodict:
3318 infodict['__files_to_move'] = {}
3319 try:
3320 files_to_delete, infodict = pp.run(infodict)
3321 except PostProcessingError as e:
3322 # Must be True and not 'only_download'
3323 if self.params.get('ignoreerrors') is True:
3324 self.report_error(e)
3325 return infodict
3326 raise
3327
3328 if not files_to_delete:
3329 return infodict
3330 if self.params.get('keepvideo', False):
3331 for f in files_to_delete:
3332 infodict['__files_to_move'].setdefault(f, '')
3333 else:
3334 self._delete_downloaded_files(
3335 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3336 return infodict
3337
3338 def run_all_pps(self, key, info, *, additional_pps=None):
3339 self._forceprint(key, info)
3340 for pp in (additional_pps or []) + self._pps[key]:
3341 info = self.run_pp(pp, info)
3342 return info
3343
3344 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3345 info = dict(ie_info)
3346 info['__files_to_move'] = files_to_move or {}
3347 try:
3348 info = self.run_all_pps(key, info)
3349 except PostProcessingError as err:
3350 msg = f'Preprocessing: {err}'
3351 info.setdefault('__pending_error', msg)
3352 self.report_error(msg, is_error=False)
3353 return info, info.pop('__files_to_move', None)
3354
3355 def post_process(self, filename, info, files_to_move=None):
3356 """Run all the postprocessors on the given file."""
3357 info['filepath'] = filename
3358 info['__files_to_move'] = files_to_move or {}
3359 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3360 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3361 del info['__files_to_move']
3362 return self.run_all_pps('after_move', info)
3363
3364 def _make_archive_id(self, info_dict):
3365 video_id = info_dict.get('id')
3366 if not video_id:
3367 return
3368 # Future-proof against any change in case
3369 # and backwards compatibility with prior versions
3370 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3371 if extractor is None:
3372 url = str_or_none(info_dict.get('url'))
3373 if not url:
3374 return
3375 # Try to find matching extractor for the URL and take its ie_key
3376 for ie_key, ie in self._ies.items():
3377 if ie.suitable(url):
3378 extractor = ie_key
3379 break
3380 else:
3381 return
3382 return f'{extractor.lower()} {video_id}'
3383
3384 def in_download_archive(self, info_dict):
3385 fn = self.params.get('download_archive')
3386 if fn is None:
3387 return False
3388
3389 vid_id = self._make_archive_id(info_dict)
3390 if not vid_id:
3391 return False # Incomplete video information
3392
3393 return vid_id in self.archive
3394
3395 def record_download_archive(self, info_dict):
3396 fn = self.params.get('download_archive')
3397 if fn is None:
3398 return
3399 vid_id = self._make_archive_id(info_dict)
3400 assert vid_id
3401 self.write_debug(f'Adding to archive: {vid_id}')
3402 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3403 archive_file.write(vid_id + '\n')
3404 self.archive.add(vid_id)
3405
3406 @staticmethod
3407 def format_resolution(format, default='unknown'):
3408 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3409 return 'audio only'
3410 if format.get('resolution') is not None:
3411 return format['resolution']
3412 if format.get('width') and format.get('height'):
3413 return '%dx%d' % (format['width'], format['height'])
3414 elif format.get('height'):
3415 return '%sp' % format['height']
3416 elif format.get('width'):
3417 return '%dx?' % format['width']
3418 return default
3419
3420 def _list_format_headers(self, *headers):
3421 if self.params.get('listformats_table', True) is not False:
3422 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3423 return headers
3424
3425 def _format_note(self, fdict):
3426 res = ''
3427 if fdict.get('ext') in ['f4f', 'f4m']:
3428 res += '(unsupported)'
3429 if fdict.get('language'):
3430 if res:
3431 res += ' '
3432 res += '[%s]' % fdict['language']
3433 if fdict.get('format_note') is not None:
3434 if res:
3435 res += ' '
3436 res += fdict['format_note']
3437 if fdict.get('tbr') is not None:
3438 if res:
3439 res += ', '
3440 res += '%4dk' % fdict['tbr']
3441 if fdict.get('container') is not None:
3442 if res:
3443 res += ', '
3444 res += '%s container' % fdict['container']
3445 if (fdict.get('vcodec') is not None
3446 and fdict.get('vcodec') != 'none'):
3447 if res:
3448 res += ', '
3449 res += fdict['vcodec']
3450 if fdict.get('vbr') is not None:
3451 res += '@'
3452 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3453 res += 'video@'
3454 if fdict.get('vbr') is not None:
3455 res += '%4dk' % fdict['vbr']
3456 if fdict.get('fps') is not None:
3457 if res:
3458 res += ', '
3459 res += '%sfps' % fdict['fps']
3460 if fdict.get('acodec') is not None:
3461 if res:
3462 res += ', '
3463 if fdict['acodec'] == 'none':
3464 res += 'video only'
3465 else:
3466 res += '%-5s' % fdict['acodec']
3467 elif fdict.get('abr') is not None:
3468 if res:
3469 res += ', '
3470 res += 'audio'
3471 if fdict.get('abr') is not None:
3472 res += '@%3dk' % fdict['abr']
3473 if fdict.get('asr') is not None:
3474 res += ' (%5dHz)' % fdict['asr']
3475 if fdict.get('filesize') is not None:
3476 if res:
3477 res += ', '
3478 res += format_bytes(fdict['filesize'])
3479 elif fdict.get('filesize_approx') is not None:
3480 if res:
3481 res += ', '
3482 res += '~' + format_bytes(fdict['filesize_approx'])
3483 return res
3484
3485 def render_formats_table(self, info_dict):
3486 if not info_dict.get('formats') and not info_dict.get('url'):
3487 return None
3488
3489 formats = info_dict.get('formats', [info_dict])
3490 if not self.params.get('listformats_table', True) is not False:
3491 table = [
3492 [
3493 format_field(f, 'format_id'),
3494 format_field(f, 'ext'),
3495 self.format_resolution(f),
3496 self._format_note(f)
3497 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3498 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3499
3500 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3501 table = [
3502 [
3503 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3504 format_field(f, 'ext'),
3505 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3506 format_field(f, 'fps', '\t%d'),
3507 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3508 delim,
3509 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3510 format_field(f, 'tbr', '\t%dk'),
3511 shorten_protocol_name(f.get('protocol', '')),
3512 delim,
3513 format_field(f, 'vcodec', default='unknown').replace(
3514 'none', 'images' if f.get('acodec') == 'none'
3515 else self._format_out('audio only', self.Styles.SUPPRESS)),
3516 format_field(f, 'vbr', '\t%dk'),
3517 format_field(f, 'acodec', default='unknown').replace(
3518 'none', '' if f.get('vcodec') == 'none'
3519 else self._format_out('video only', self.Styles.SUPPRESS)),
3520 format_field(f, 'abr', '\t%dk'),
3521 format_field(f, 'asr', '\t%dHz'),
3522 join_nonempty(
3523 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3524 format_field(f, 'language', '[%s]'),
3525 join_nonempty(format_field(f, 'format_note'),
3526 format_field(f, 'container', ignore=(None, f.get('ext'))),
3527 delim=', '),
3528 delim=' '),
3529 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3530 header_line = self._list_format_headers(
3531 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3532 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3533
3534 return render_table(
3535 header_line, table, hide_empty=True,
3536 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3537
3538 def render_thumbnails_table(self, info_dict):
3539 thumbnails = list(info_dict.get('thumbnails') or [])
3540 if not thumbnails:
3541 return None
3542 return render_table(
3543 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3544 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3545
3546 def render_subtitles_table(self, video_id, subtitles):
3547 def _row(lang, formats):
3548 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3549 if len(set(names)) == 1:
3550 names = [] if names[0] == 'unknown' else names[:1]
3551 return [lang, ', '.join(names), ', '.join(exts)]
3552
3553 if not subtitles:
3554 return None
3555 return render_table(
3556 self._list_format_headers('Language', 'Name', 'Formats'),
3557 [_row(lang, formats) for lang, formats in subtitles.items()],
3558 hide_empty=True)
3559
3560 def __list_table(self, video_id, name, func, *args):
3561 table = func(*args)
3562 if not table:
3563 self.to_screen(f'{video_id} has no {name}')
3564 return
3565 self.to_screen(f'[info] Available {name} for {video_id}:')
3566 self.to_stdout(table)
3567
3568 def list_formats(self, info_dict):
3569 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3570
3571 def list_thumbnails(self, info_dict):
3572 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3573
3574 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3575 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3576
3577 def urlopen(self, req):
3578 """ Start an HTTP download """
3579 if isinstance(req, str):
3580 req = sanitized_Request(req)
3581 return self._opener.open(req, timeout=self._socket_timeout)
3582
3583 def print_debug_header(self):
3584 if not self.params.get('verbose'):
3585 return
3586
3587 # These imports can be slow. So import them only as needed
3588 from .extractor.extractors import _LAZY_LOADER
3589 from .extractor.extractors import _PLUGIN_CLASSES as plugin_extractors
3590
3591 def get_encoding(stream):
3592 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3593 if not supports_terminal_sequences(stream):
3594 from .utils import WINDOWS_VT_MODE # Must be imported locally
3595 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3596 return ret
3597
3598 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3599 locale.getpreferredencoding(),
3600 sys.getfilesystemencoding(),
3601 self.get_encoding(),
3602 ', '.join(
3603 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3604 if stream is not None and key != 'console')
3605 )
3606
3607 logger = self.params.get('logger')
3608 if logger:
3609 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3610 write_debug(encoding_str)
3611 else:
3612 write_string(f'[debug] {encoding_str}\n', encoding=None)
3613 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3614
3615 source = detect_variant()
3616 write_debug(join_nonempty(
3617 'yt-dlp version', __version__,
3618 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3619 '' if source == 'unknown' else f'({source})',
3620 delim=' '))
3621 if not _LAZY_LOADER:
3622 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3623 write_debug('Lazy loading extractors is forcibly disabled')
3624 else:
3625 write_debug('Lazy loading extractors is disabled')
3626 if plugin_extractors or plugin_postprocessors:
3627 write_debug('Plugins: %s' % [
3628 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3629 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3630 if self.params['compat_opts']:
3631 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3632
3633 if source == 'source':
3634 try:
3635 stdout, _, _ = Popen.run(
3636 ['git', 'rev-parse', '--short', 'HEAD'],
3637 text=True, cwd=os.path.dirname(os.path.abspath(__file__)),
3638 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
3639 if re.fullmatch('[0-9a-f]+', stdout.strip()):
3640 write_debug(f'Git HEAD: {stdout.strip()}')
3641 except Exception:
3642 with contextlib.suppress(Exception):
3643 sys.exc_clear()
3644
3645 def python_implementation():
3646 impl_name = platform.python_implementation()
3647 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3648 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3649 return impl_name
3650
3651 write_debug('Python version %s (%s %s) - %s' % (
3652 platform.python_version(),
3653 python_implementation(),
3654 platform.architecture()[0],
3655 platform_name()))
3656
3657 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3658 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3659 if ffmpeg_features:
3660 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3661
3662 exe_versions['rtmpdump'] = rtmpdump_version()
3663 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3664 exe_str = ', '.join(
3665 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3666 ) or 'none'
3667 write_debug('exe versions: %s' % exe_str)
3668
3669 from .compat.compat_utils import get_package_info
3670 from .dependencies import available_dependencies
3671
3672 write_debug('Optional libraries: %s' % (', '.join(sorted({
3673 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3674 })) or 'none'))
3675
3676 self._setup_opener()
3677 proxy_map = {}
3678 for handler in self._opener.handlers:
3679 if hasattr(handler, 'proxies'):
3680 proxy_map.update(handler.proxies)
3681 write_debug(f'Proxy map: {proxy_map}')
3682
3683 # Not implemented
3684 if False and self.params.get('call_home'):
3685 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3686 write_debug('Public IP address: %s' % ipaddr)
3687 latest_version = self.urlopen(
3688 'https://yt-dl.org/latest/version').read().decode()
3689 if version_tuple(latest_version) > version_tuple(__version__):
3690 self.report_warning(
3691 'You are using an outdated version (newest version: %s)! '
3692 'See https://yt-dl.org/update if you need help updating.' %
3693 latest_version)
3694
3695 def _setup_opener(self):
3696 if hasattr(self, '_opener'):
3697 return
3698 timeout_val = self.params.get('socket_timeout')
3699 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3700
3701 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3702 opts_cookiefile = self.params.get('cookiefile')
3703 opts_proxy = self.params.get('proxy')
3704
3705 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3706
3707 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3708 if opts_proxy is not None:
3709 if opts_proxy == '':
3710 proxies = {}
3711 else:
3712 proxies = {'http': opts_proxy, 'https': opts_proxy}
3713 else:
3714 proxies = compat_urllib_request.getproxies()
3715 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3716 if 'http' in proxies and 'https' not in proxies:
3717 proxies['https'] = proxies['http']
3718 proxy_handler = PerRequestProxyHandler(proxies)
3719
3720 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3721 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3722 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3723 redirect_handler = YoutubeDLRedirectHandler()
3724 data_handler = urllib.request.DataHandler()
3725
3726 # When passing our own FileHandler instance, build_opener won't add the
3727 # default FileHandler and allows us to disable the file protocol, which
3728 # can be used for malicious purposes (see
3729 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3730 file_handler = compat_urllib_request.FileHandler()
3731
3732 def file_open(*args, **kwargs):
3733 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3734 file_handler.file_open = file_open
3735
3736 opener = compat_urllib_request.build_opener(
3737 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3738
3739 # Delete the default user-agent header, which would otherwise apply in
3740 # cases where our custom HTTP handler doesn't come into play
3741 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3742 opener.addheaders = []
3743 self._opener = opener
3744
3745 def encode(self, s):
3746 if isinstance(s, bytes):
3747 return s # Already encoded
3748
3749 try:
3750 return s.encode(self.get_encoding())
3751 except UnicodeEncodeError as err:
3752 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3753 raise
3754
3755 def get_encoding(self):
3756 encoding = self.params.get('encoding')
3757 if encoding is None:
3758 encoding = preferredencoding()
3759 return encoding
3760
3761 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3762 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3763 if overwrite is None:
3764 overwrite = self.params.get('overwrites', True)
3765 if not self.params.get('writeinfojson'):
3766 return False
3767 elif not infofn:
3768 self.write_debug(f'Skipping writing {label} infojson')
3769 return False
3770 elif not self._ensure_dir_exists(infofn):
3771 return None
3772 elif not overwrite and os.path.exists(infofn):
3773 self.to_screen(f'[info] {label.title()} metadata is already present')
3774 return 'exists'
3775
3776 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3777 try:
3778 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3779 return True
3780 except OSError:
3781 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3782 return None
3783
3784 def _write_description(self, label, ie_result, descfn):
3785 ''' Write description and returns True = written, False = skip, None = error '''
3786 if not self.params.get('writedescription'):
3787 return False
3788 elif not descfn:
3789 self.write_debug(f'Skipping writing {label} description')
3790 return False
3791 elif not self._ensure_dir_exists(descfn):
3792 return None
3793 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3794 self.to_screen(f'[info] {label.title()} description is already present')
3795 elif ie_result.get('description') is None:
3796 self.report_warning(f'There\'s no {label} description to write')
3797 return False
3798 else:
3799 try:
3800 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3801 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3802 descfile.write(ie_result['description'])
3803 except OSError:
3804 self.report_error(f'Cannot write {label} description file {descfn}')
3805 return None
3806 return True
3807
3808 def _write_subtitles(self, info_dict, filename):
3809 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3810 ret = []
3811 subtitles = info_dict.get('requested_subtitles')
3812 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3813 # subtitles download errors are already managed as troubles in relevant IE
3814 # that way it will silently go on when used with unsupporting IE
3815 return ret
3816
3817 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3818 if not sub_filename_base:
3819 self.to_screen('[info] Skipping writing video subtitles')
3820 return ret
3821 for sub_lang, sub_info in subtitles.items():
3822 sub_format = sub_info['ext']
3823 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3824 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3825 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3826 if existing_sub:
3827 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3828 sub_info['filepath'] = existing_sub
3829 ret.append((existing_sub, sub_filename_final))
3830 continue
3831
3832 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3833 if sub_info.get('data') is not None:
3834 try:
3835 # Use newline='' to prevent conversion of newline characters
3836 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3837 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3838 subfile.write(sub_info['data'])
3839 sub_info['filepath'] = sub_filename
3840 ret.append((sub_filename, sub_filename_final))
3841 continue
3842 except OSError:
3843 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3844 return None
3845
3846 try:
3847 sub_copy = sub_info.copy()
3848 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3849 self.dl(sub_filename, sub_copy, subtitle=True)
3850 sub_info['filepath'] = sub_filename
3851 ret.append((sub_filename, sub_filename_final))
3852 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3853 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3854 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3855 if not self.params.get('ignoreerrors'):
3856 self.report_error(msg)
3857 raise DownloadError(msg)
3858 self.report_warning(msg)
3859 return ret
3860
3861 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3862 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3863 write_all = self.params.get('write_all_thumbnails', False)
3864 thumbnails, ret = [], []
3865 if write_all or self.params.get('writethumbnail', False):
3866 thumbnails = info_dict.get('thumbnails') or []
3867 multiple = write_all and len(thumbnails) > 1
3868
3869 if thumb_filename_base is None:
3870 thumb_filename_base = filename
3871 if thumbnails and not thumb_filename_base:
3872 self.write_debug(f'Skipping writing {label} thumbnail')
3873 return ret
3874
3875 for idx, t in list(enumerate(thumbnails))[::-1]:
3876 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3877 thumb_display_id = f'{label} thumbnail {t["id"]}'
3878 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3879 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3880
3881 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3882 if existing_thumb:
3883 self.to_screen('[info] %s is already present' % (
3884 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3885 t['filepath'] = existing_thumb
3886 ret.append((existing_thumb, thumb_filename_final))
3887 else:
3888 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3889 try:
3890 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3891 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3892 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3893 shutil.copyfileobj(uf, thumbf)
3894 ret.append((thumb_filename, thumb_filename_final))
3895 t['filepath'] = thumb_filename
3896 except network_exceptions as err:
3897 thumbnails.pop(idx)
3898 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3899 if ret and not write_all:
3900 break
3901 return ret