]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
38d146bfc70e0517d219a80ef9e4be46d5f84fb4
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 import collections
2 import contextlib
3 import datetime
4 import errno
5 import fileinput
6 import functools
7 import io
8 import itertools
9 import json
10 import locale
11 import operator
12 import os
13 import random
14 import re
15 import shutil
16 import subprocess
17 import sys
18 import tempfile
19 import time
20 import tokenize
21 import traceback
22 import unicodedata
23 import urllib.request
24 from string import ascii_letters
25
26 from .cache import Cache
27 from .compat import HAS_LEGACY as compat_has_legacy
28 from .compat import compat_os_name, compat_shlex_quote
29 from .cookies import load_cookies
30 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
31 from .downloader.rtmp import rtmpdump_version
32 from .extractor import gen_extractor_classes, get_info_extractor
33 from .extractor.openload import PhantomJSwrapper
34 from .minicurses import format_text
35 from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
36 from .postprocessor import (
37 EmbedThumbnailPP,
38 FFmpegFixupDuplicateMoovPP,
39 FFmpegFixupDurationPP,
40 FFmpegFixupM3u8PP,
41 FFmpegFixupM4aPP,
42 FFmpegFixupStretchedPP,
43 FFmpegFixupTimestampPP,
44 FFmpegMergerPP,
45 FFmpegPostProcessor,
46 MoveFilesAfterDownloadPP,
47 get_postprocessor,
48 )
49 from .update import detect_variant
50 from .utils import (
51 DEFAULT_OUTTMPL,
52 IDENTITY,
53 LINK_TEMPLATES,
54 NO_DEFAULT,
55 NUMBER_RE,
56 OUTTMPL_TYPES,
57 POSTPROCESS_WHEN,
58 STR_FORMAT_RE_TMPL,
59 STR_FORMAT_TYPES,
60 ContentTooShortError,
61 DateRange,
62 DownloadCancelled,
63 DownloadError,
64 EntryNotInPlaylist,
65 ExistingVideoReached,
66 ExtractorError,
67 GeoRestrictedError,
68 HEADRequest,
69 ISO3166Utils,
70 LazyList,
71 MaxDownloadsReached,
72 Namespace,
73 PagedList,
74 PerRequestProxyHandler,
75 PlaylistEntries,
76 Popen,
77 PostProcessingError,
78 ReExtractInfo,
79 RejectedVideoReached,
80 SameFileError,
81 UnavailableVideoError,
82 YoutubeDLCookieProcessor,
83 YoutubeDLHandler,
84 YoutubeDLRedirectHandler,
85 age_restricted,
86 args_to_str,
87 date_from_str,
88 determine_ext,
89 determine_protocol,
90 encode_compat_str,
91 encodeFilename,
92 error_to_compat_str,
93 expand_path,
94 filter_dict,
95 float_or_none,
96 format_bytes,
97 format_decimal_suffix,
98 format_field,
99 formatSeconds,
100 get_domain,
101 int_or_none,
102 iri_to_uri,
103 join_nonempty,
104 locked_file,
105 make_dir,
106 make_HTTPS_handler,
107 merge_headers,
108 network_exceptions,
109 number_of_digits,
110 orderedSet,
111 parse_filesize,
112 preferredencoding,
113 prepend_extension,
114 register_socks_protocols,
115 remove_terminal_sequences,
116 render_table,
117 replace_extension,
118 sanitize_filename,
119 sanitize_path,
120 sanitize_url,
121 sanitized_Request,
122 std_headers,
123 str_or_none,
124 strftime_or_none,
125 subtitles_filename,
126 supports_terminal_sequences,
127 system_identifier,
128 timetuple_from_msec,
129 to_high_limit_path,
130 traverse_obj,
131 try_get,
132 url_basename,
133 variadic,
134 version_tuple,
135 windows_enable_vt_mode,
136 write_json_file,
137 write_string,
138 )
139 from .version import RELEASE_GIT_HEAD, __version__
140
141 if compat_os_name == 'nt':
142 import ctypes
143
144
145 class YoutubeDL:
146 """YoutubeDL class.
147
148 YoutubeDL objects are the ones responsible of downloading the
149 actual video file and writing it to disk if the user has requested
150 it, among some other tasks. In most cases there should be one per
151 program. As, given a video URL, the downloader doesn't know how to
152 extract all the needed information, task that InfoExtractors do, it
153 has to pass the URL to one of them.
154
155 For this, YoutubeDL objects have a method that allows
156 InfoExtractors to be registered in a given order. When it is passed
157 a URL, the YoutubeDL object handles it to the first InfoExtractor it
158 finds that reports being able to handle it. The InfoExtractor extracts
159 all the information about the video or videos the URL refers to, and
160 YoutubeDL process the extracted information, possibly using a File
161 Downloader to download the video.
162
163 YoutubeDL objects accept a lot of parameters. In order not to saturate
164 the object constructor with arguments, it receives a dictionary of
165 options instead. These options are available through the params
166 attribute for the InfoExtractors to use. The YoutubeDL also
167 registers itself as the downloader in charge for the InfoExtractors
168 that are added to it, so this is a "mutual registration".
169
170 Available options:
171
172 username: Username for authentication purposes.
173 password: Password for authentication purposes.
174 videopassword: Password for accessing a video.
175 ap_mso: Adobe Pass multiple-system operator identifier.
176 ap_username: Multiple-system operator account username.
177 ap_password: Multiple-system operator account password.
178 usenetrc: Use netrc for authentication instead.
179 verbose: Print additional info to stdout.
180 quiet: Do not print messages to stdout.
181 no_warnings: Do not print out anything for warnings.
182 forceprint: A dict with keys WHEN mapped to a list of templates to
183 print to stdout. The allowed keys are video or any of the
184 items in utils.POSTPROCESS_WHEN.
185 For compatibility, a single list is also accepted
186 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
187 a list of tuples with (template, filename)
188 forcejson: Force printing info_dict as JSON.
189 dump_single_json: Force printing the info_dict of the whole playlist
190 (or video) as a single JSON line.
191 force_write_download_archive: Force writing download archive regardless
192 of 'skip_download' or 'simulate'.
193 simulate: Do not download the video files. If unset (or None),
194 simulate only if listsubtitles, listformats or list_thumbnails is used
195 format: Video format code. see "FORMAT SELECTION" for more details.
196 You can also pass a function. The function takes 'ctx' as
197 argument and returns the formats to download.
198 See "build_format_selector" for an implementation
199 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
200 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
201 extracting metadata even if the video is not actually
202 available for download (experimental)
203 format_sort: A list of fields by which to sort the video formats.
204 See "Sorting Formats" for more details.
205 format_sort_force: Force the given format_sort. see "Sorting Formats"
206 for more details.
207 prefer_free_formats: Whether to prefer video formats with free containers
208 over non-free ones of same quality.
209 allow_multiple_video_streams: Allow multiple video streams to be merged
210 into a single file
211 allow_multiple_audio_streams: Allow multiple audio streams to be merged
212 into a single file
213 check_formats Whether to test if the formats are downloadable.
214 Can be True (check all), False (check none),
215 'selected' (check selected formats),
216 or None (check only if requested by extractor)
217 paths: Dictionary of output paths. The allowed keys are 'home'
218 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
219 outtmpl: Dictionary of templates for output names. Allowed keys
220 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
221 For compatibility with youtube-dl, a single string can also be used
222 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
223 restrictfilenames: Do not allow "&" and spaces in file names
224 trim_file_name: Limit length of filename (extension excluded)
225 windowsfilenames: Force the filenames to be windows compatible
226 ignoreerrors: Do not stop on download/postprocessing errors.
227 Can be 'only_download' to ignore only download errors.
228 Default is 'only_download' for CLI, but False for API
229 skip_playlist_after_errors: Number of allowed failures until the rest of
230 the playlist is skipped
231 force_generic_extractor: Force downloader to use the generic extractor
232 overwrites: Overwrite all video and metadata files if True,
233 overwrite only non-video files if None
234 and don't overwrite any file if False
235 For compatibility with youtube-dl,
236 "nooverwrites" may also be used instead
237 playlist_items: Specific indices of playlist to download.
238 playlistrandom: Download playlist items in random order.
239 lazy_playlist: Process playlist entries as they are received.
240 matchtitle: Download only matching titles.
241 rejecttitle: Reject downloads for matching titles.
242 logger: Log messages to a logging.Logger instance.
243 logtostderr: Log messages to stderr instead of stdout.
244 consoletitle: Display progress in console window's titlebar.
245 writedescription: Write the video description to a .description file
246 writeinfojson: Write the video description to a .info.json file
247 clean_infojson: Remove private fields from the infojson
248 getcomments: Extract video comments. This will not be written to disk
249 unless writeinfojson is also given
250 writeannotations: Write the video annotations to a .annotations.xml file
251 writethumbnail: Write the thumbnail image to a file
252 allow_playlist_files: Whether to write playlists' description, infojson etc
253 also to disk when using the 'write*' options
254 write_all_thumbnails: Write all thumbnail formats to files
255 writelink: Write an internet shortcut file, depending on the
256 current platform (.url/.webloc/.desktop)
257 writeurllink: Write a Windows internet shortcut file (.url)
258 writewebloclink: Write a macOS internet shortcut file (.webloc)
259 writedesktoplink: Write a Linux internet shortcut file (.desktop)
260 writesubtitles: Write the video subtitles to a file
261 writeautomaticsub: Write the automatically generated subtitles to a file
262 listsubtitles: Lists all available subtitles for the video
263 subtitlesformat: The format code for subtitles
264 subtitleslangs: List of languages of the subtitles to download (can be regex).
265 The list may contain "all" to refer to all the available
266 subtitles. The language can be prefixed with a "-" to
267 exclude it from the requested languages. Eg: ['all', '-live_chat']
268 keepvideo: Keep the video file after post-processing
269 daterange: A DateRange object, download only if the upload_date is in the range.
270 skip_download: Skip the actual download of the video file
271 cachedir: Location of the cache files in the filesystem.
272 False to disable filesystem cache.
273 noplaylist: Download single video instead of a playlist if in doubt.
274 age_limit: An integer representing the user's age in years.
275 Unsuitable videos for the given age are skipped.
276 min_views: An integer representing the minimum view count the video
277 must have in order to not be skipped.
278 Videos without view count information are always
279 downloaded. None for no limit.
280 max_views: An integer representing the maximum view count.
281 Videos that are more popular than that are not
282 downloaded.
283 Videos without view count information are always
284 downloaded. None for no limit.
285 download_archive: File name of a file where all downloads are recorded.
286 Videos already present in the file are not downloaded
287 again.
288 break_on_existing: Stop the download process after attempting to download a
289 file that is in the archive.
290 break_on_reject: Stop the download process when encountering a video that
291 has been filtered out.
292 break_per_url: Whether break_on_reject and break_on_existing
293 should act on each input URL as opposed to for the entire queue
294 cookiefile: File name or text stream from where cookies should be read and dumped to
295 cookiesfrombrowser: A tuple containing the name of the browser, the profile
296 name/pathfrom where cookies are loaded, and the name of the
297 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
298 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
299 support RFC 5746 secure renegotiation
300 nocheckcertificate: Do not verify SSL certificates
301 client_certificate: Path to client certificate file in PEM format. May include the private key
302 client_certificate_key: Path to private key file for client certificate
303 client_certificate_password: Password for client certificate private key, if encrypted.
304 If not provided and the key is encrypted, yt-dlp will ask interactively
305 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
306 At the moment, this is only supported by YouTube.
307 http_headers: A dictionary of custom headers to be used for all requests
308 proxy: URL of the proxy server to use
309 geo_verification_proxy: URL of the proxy to use for IP address verification
310 on geo-restricted sites.
311 socket_timeout: Time to wait for unresponsive hosts, in seconds
312 bidi_workaround: Work around buggy terminals without bidirectional text
313 support, using fridibi
314 debug_printtraffic:Print out sent and received HTTP traffic
315 default_search: Prepend this string if an input url is not valid.
316 'auto' for elaborate guessing
317 encoding: Use this encoding instead of the system-specified.
318 extract_flat: Do not resolve URLs, return the immediate result.
319 Pass in 'in_playlist' to only show this behavior for
320 playlist items.
321 wait_for_video: If given, wait for scheduled streams to become available.
322 The value should be a tuple containing the range
323 (min_secs, max_secs) to wait between retries
324 postprocessors: A list of dictionaries, each with an entry
325 * key: The name of the postprocessor. See
326 yt_dlp/postprocessor/__init__.py for a list.
327 * when: When to run the postprocessor. Allowed values are
328 the entries of utils.POSTPROCESS_WHEN
329 Assumed to be 'post_process' if not given
330 progress_hooks: A list of functions that get called on download
331 progress, with a dictionary with the entries
332 * status: One of "downloading", "error", or "finished".
333 Check this first and ignore unknown values.
334 * info_dict: The extracted info_dict
335
336 If status is one of "downloading", or "finished", the
337 following properties may also be present:
338 * filename: The final filename (always present)
339 * tmpfilename: The filename we're currently writing to
340 * downloaded_bytes: Bytes on disk
341 * total_bytes: Size of the whole file, None if unknown
342 * total_bytes_estimate: Guess of the eventual file size,
343 None if unavailable.
344 * elapsed: The number of seconds since download started.
345 * eta: The estimated time in seconds, None if unknown
346 * speed: The download speed in bytes/second, None if
347 unknown
348 * fragment_index: The counter of the currently
349 downloaded video fragment.
350 * fragment_count: The number of fragments (= individual
351 files that will be merged)
352
353 Progress hooks are guaranteed to be called at least once
354 (with status "finished") if the download is successful.
355 postprocessor_hooks: A list of functions that get called on postprocessing
356 progress, with a dictionary with the entries
357 * status: One of "started", "processing", or "finished".
358 Check this first and ignore unknown values.
359 * postprocessor: Name of the postprocessor
360 * info_dict: The extracted info_dict
361
362 Progress hooks are guaranteed to be called at least twice
363 (with status "started" and "finished") if the processing is successful.
364 merge_output_format: Extension to use when merging formats.
365 final_ext: Expected final extension; used to detect when the file was
366 already downloaded and converted
367 fixup: Automatically correct known faults of the file.
368 One of:
369 - "never": do nothing
370 - "warn": only emit a warning
371 - "detect_or_warn": check whether we can do anything
372 about it, warn otherwise (default)
373 source_address: Client-side IP address to bind to.
374 sleep_interval_requests: Number of seconds to sleep between requests
375 during extraction
376 sleep_interval: Number of seconds to sleep before each download when
377 used alone or a lower bound of a range for randomized
378 sleep before each download (minimum possible number
379 of seconds to sleep) when used along with
380 max_sleep_interval.
381 max_sleep_interval:Upper bound of a range for randomized sleep before each
382 download (maximum possible number of seconds to sleep).
383 Must only be used along with sleep_interval.
384 Actual sleep time will be a random float from range
385 [sleep_interval; max_sleep_interval].
386 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
387 listformats: Print an overview of available video formats and exit.
388 list_thumbnails: Print a table of all thumbnails and exit.
389 match_filter: A function that gets called for every video with the signature
390 (info_dict, *, incomplete: bool) -> Optional[str]
391 For backward compatibility with youtube-dl, the signature
392 (info_dict) -> Optional[str] is also allowed.
393 - If it returns a message, the video is ignored.
394 - If it returns None, the video is downloaded.
395 - If it returns utils.NO_DEFAULT, the user is interactively
396 asked whether to download the video.
397 match_filter_func in utils.py is one example for this.
398 no_color: Do not emit color codes in output.
399 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
400 HTTP header
401 geo_bypass_country:
402 Two-letter ISO 3166-2 country code that will be used for
403 explicit geographic restriction bypassing via faking
404 X-Forwarded-For HTTP header
405 geo_bypass_ip_block:
406 IP range in CIDR notation that will be used similarly to
407 geo_bypass_country
408 external_downloader: A dictionary of protocol keys and the executable of the
409 external downloader to use for it. The allowed protocols
410 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
411 Set the value to 'native' to use the native downloader
412 compat_opts: Compatibility options. See "Differences in default behavior".
413 The following options do not work when used through the API:
414 filename, abort-on-error, multistreams, no-live-chat, format-sort
415 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
416 Refer __init__.py for their implementation
417 progress_template: Dictionary of templates for progress outputs.
418 Allowed keys are 'download', 'postprocess',
419 'download-title' (console title) and 'postprocess-title'.
420 The template is mapped on a dictionary with keys 'progress' and 'info'
421 retry_sleep_functions: Dictionary of functions that takes the number of attempts
422 as argument and returns the time to sleep in seconds.
423 Allowed keys are 'http', 'fragment', 'file_access'
424 download_ranges: A function that gets called for every video with the signature
425 (info_dict, *, ydl) -> Iterable[Section].
426 Only the returned sections will be downloaded. Each Section contains:
427 * start_time: Start time of the section in seconds
428 * end_time: End time of the section in seconds
429 * title: Section title (Optional)
430 * index: Section number (Optional)
431
432 The following parameters are not used by YoutubeDL itself, they are used by
433 the downloader (see yt_dlp/downloader/common.py):
434 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
435 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
436 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
437 external_downloader_args, concurrent_fragment_downloads.
438
439 The following options are used by the post processors:
440 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
441 to the binary or its containing directory.
442 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
443 and a list of additional command-line arguments for the
444 postprocessor/executable. The dict can also have "PP+EXE" keys
445 which are used when the given exe is used by the given PP.
446 Use 'default' as the name for arguments to passed to all PP
447 For compatibility with youtube-dl, a single list of args
448 can also be used
449
450 The following options are used by the extractors:
451 extractor_retries: Number of times to retry for known errors
452 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
453 hls_split_discontinuity: Split HLS playlists to different formats at
454 discontinuities such as ad breaks (default: False)
455 extractor_args: A dictionary of arguments to be passed to the extractors.
456 See "EXTRACTOR ARGUMENTS" for details.
457 Eg: {'youtube': {'skip': ['dash', 'hls']}}
458 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
459
460 The following options are deprecated and may be removed in the future:
461
462 playliststart: - Use playlist_items
463 Playlist item to start at.
464 playlistend: - Use playlist_items
465 Playlist item to end at.
466 playlistreverse: - Use playlist_items
467 Download playlist items in reverse order.
468 forceurl: - Use forceprint
469 Force printing final URL.
470 forcetitle: - Use forceprint
471 Force printing title.
472 forceid: - Use forceprint
473 Force printing ID.
474 forcethumbnail: - Use forceprint
475 Force printing thumbnail URL.
476 forcedescription: - Use forceprint
477 Force printing description.
478 forcefilename: - Use forceprint
479 Force printing final filename.
480 forceduration: - Use forceprint
481 Force printing duration.
482 allsubtitles: - Use subtitleslangs = ['all']
483 Downloads all the subtitles of the video
484 (requires writesubtitles or writeautomaticsub)
485 include_ads: - Doesn't work
486 Download ads as well
487 call_home: - Not implemented
488 Boolean, true iff we are allowed to contact the
489 yt-dlp servers for debugging.
490 post_hooks: - Register a custom postprocessor
491 A list of functions that get called as the final step
492 for each video file, after all postprocessors have been
493 called. The filename will be passed as the only argument.
494 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
495 Use the native HLS downloader instead of ffmpeg/avconv
496 if True, otherwise use ffmpeg/avconv if False, otherwise
497 use downloader suggested by extractor if None.
498 prefer_ffmpeg: - avconv support is deprecated
499 If False, use avconv instead of ffmpeg if both are available,
500 otherwise prefer ffmpeg.
501 youtube_include_dash_manifest: - Use extractor_args
502 If True (default), DASH manifests and related
503 data will be downloaded and processed by extractor.
504 You can reduce network I/O by disabling it if you don't
505 care about DASH. (only for youtube)
506 youtube_include_hls_manifest: - Use extractor_args
507 If True (default), HLS manifests and related
508 data will be downloaded and processed by extractor.
509 You can reduce network I/O by disabling it if you don't
510 care about HLS. (only for youtube)
511 """
512
513 _NUMERIC_FIELDS = {
514 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
515 'timestamp', 'release_timestamp',
516 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
517 'average_rating', 'comment_count', 'age_limit',
518 'start_time', 'end_time',
519 'chapter_number', 'season_number', 'episode_number',
520 'track_number', 'disc_number', 'release_year',
521 }
522
523 _format_fields = {
524 # NB: Keep in sync with the docstring of extractor/common.py
525 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
526 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
527 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
528 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
529 'preference', 'language', 'language_preference', 'quality', 'source_preference',
530 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
531 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
532 }
533 _format_selection_exts = {
534 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
535 'video': {'mp4', 'flv', 'webm', '3gp'},
536 'storyboards': {'mhtml'},
537 }
538
539 def __init__(self, params=None, auto_init=True):
540 """Create a FileDownloader object with the given options.
541 @param auto_init Whether to load the default extractors and print header (if verbose).
542 Set to 'no_verbose_header' to not print the header
543 """
544 if params is None:
545 params = {}
546 self.params = params
547 self._ies = {}
548 self._ies_instances = {}
549 self._pps = {k: [] for k in POSTPROCESS_WHEN}
550 self._printed_messages = set()
551 self._first_webpage_request = True
552 self._post_hooks = []
553 self._progress_hooks = []
554 self._postprocessor_hooks = []
555 self._download_retcode = 0
556 self._num_downloads = 0
557 self._num_videos = 0
558 self._playlist_level = 0
559 self._playlist_urls = set()
560 self.cache = Cache(self)
561
562 windows_enable_vt_mode()
563 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
564 self._out_files = Namespace(
565 out=stdout,
566 error=sys.stderr,
567 screen=sys.stderr if self.params.get('quiet') else stdout,
568 console=None if compat_os_name == 'nt' else next(
569 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
570 )
571 self._allow_colors = Namespace(**{
572 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
573 for type_, stream in self._out_files.items_ if type_ != 'console'
574 })
575
576 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 6), (3, 7)
577 current_version = sys.version_info[:2]
578 if current_version < MIN_RECOMMENDED:
579 msg = ('Support for Python version %d.%d has been deprecated. '
580 'See https://github.com/yt-dlp/yt-dlp/issues/3764 for more details. '
581 'You will recieve only one more update on this version')
582 if current_version < MIN_SUPPORTED:
583 msg = 'Python version %d.%d is no longer supported'
584 self.deprecation_warning(
585 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
586
587 if self.params.get('allow_unplayable_formats'):
588 self.report_warning(
589 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
590 'This is a developer option intended for debugging. \n'
591 ' If you experience any issues while using this option, '
592 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
593
594 def check_deprecated(param, option, suggestion):
595 if self.params.get(param) is not None:
596 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
597 return True
598 return False
599
600 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
601 if self.params.get('geo_verification_proxy') is None:
602 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
603
604 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
605 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
606 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
607
608 for msg in self.params.get('_warnings', []):
609 self.report_warning(msg)
610 for msg in self.params.get('_deprecation_warnings', []):
611 self.deprecation_warning(msg)
612
613 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
614 if not compat_has_legacy:
615 self.params['compat_opts'].add('no-compat-legacy')
616 if 'list-formats' in self.params['compat_opts']:
617 self.params['listformats_table'] = False
618
619 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
620 # nooverwrites was unnecessarily changed to overwrites
621 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
622 # This ensures compatibility with both keys
623 self.params['overwrites'] = not self.params['nooverwrites']
624 elif self.params.get('overwrites') is None:
625 self.params.pop('overwrites', None)
626 else:
627 self.params['nooverwrites'] = not self.params['overwrites']
628
629 self.params.setdefault('forceprint', {})
630 self.params.setdefault('print_to_file', {})
631
632 # Compatibility with older syntax
633 if not isinstance(params['forceprint'], dict):
634 self.params['forceprint'] = {'video': params['forceprint']}
635
636 if self.params.get('bidi_workaround', False):
637 try:
638 import pty
639 master, slave = pty.openpty()
640 width = shutil.get_terminal_size().columns
641 width_args = [] if width is None else ['-w', str(width)]
642 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
643 try:
644 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
645 except OSError:
646 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
647 self._output_channel = os.fdopen(master, 'rb')
648 except OSError as ose:
649 if ose.errno == errno.ENOENT:
650 self.report_warning(
651 'Could not find fribidi executable, ignoring --bidi-workaround. '
652 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
653 else:
654 raise
655
656 if auto_init:
657 if auto_init != 'no_verbose_header':
658 self.print_debug_header()
659 self.add_default_info_extractors()
660
661 if (sys.platform != 'win32'
662 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
663 and not self.params.get('restrictfilenames', False)):
664 # Unicode filesystem API will throw errors (#1474, #13027)
665 self.report_warning(
666 'Assuming --restrict-filenames since file system encoding '
667 'cannot encode all characters. '
668 'Set the LC_ALL environment variable to fix this.')
669 self.params['restrictfilenames'] = True
670
671 self._parse_outtmpl()
672
673 # Creating format selector here allows us to catch syntax errors before the extraction
674 self.format_selector = (
675 self.params.get('format') if self.params.get('format') in (None, '-')
676 else self.params['format'] if callable(self.params['format'])
677 else self.build_format_selector(self.params['format']))
678
679 # Set http_headers defaults according to std_headers
680 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
681
682 hooks = {
683 'post_hooks': self.add_post_hook,
684 'progress_hooks': self.add_progress_hook,
685 'postprocessor_hooks': self.add_postprocessor_hook,
686 }
687 for opt, fn in hooks.items():
688 for ph in self.params.get(opt, []):
689 fn(ph)
690
691 for pp_def_raw in self.params.get('postprocessors', []):
692 pp_def = dict(pp_def_raw)
693 when = pp_def.pop('when', 'post_process')
694 self.add_post_processor(
695 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
696 when=when)
697
698 self._setup_opener()
699 register_socks_protocols()
700
701 def preload_download_archive(fn):
702 """Preload the archive, if any is specified"""
703 if fn is None:
704 return False
705 self.write_debug(f'Loading archive file {fn!r}')
706 try:
707 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
708 for line in archive_file:
709 self.archive.add(line.strip())
710 except OSError as ioe:
711 if ioe.errno != errno.ENOENT:
712 raise
713 return False
714 return True
715
716 self.archive = set()
717 preload_download_archive(self.params.get('download_archive'))
718
719 def warn_if_short_id(self, argv):
720 # short YouTube ID starting with dash?
721 idxs = [
722 i for i, a in enumerate(argv)
723 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
724 if idxs:
725 correct_argv = (
726 ['yt-dlp']
727 + [a for i, a in enumerate(argv) if i not in idxs]
728 + ['--'] + [argv[i] for i in idxs]
729 )
730 self.report_warning(
731 'Long argument string detected. '
732 'Use -- to separate parameters and URLs, like this:\n%s' %
733 args_to_str(correct_argv))
734
735 def add_info_extractor(self, ie):
736 """Add an InfoExtractor object to the end of the list."""
737 ie_key = ie.ie_key()
738 self._ies[ie_key] = ie
739 if not isinstance(ie, type):
740 self._ies_instances[ie_key] = ie
741 ie.set_downloader(self)
742
743 def _get_info_extractor_class(self, ie_key):
744 ie = self._ies.get(ie_key)
745 if ie is None:
746 ie = get_info_extractor(ie_key)
747 self.add_info_extractor(ie)
748 return ie
749
750 def get_info_extractor(self, ie_key):
751 """
752 Get an instance of an IE with name ie_key, it will try to get one from
753 the _ies list, if there's no instance it will create a new one and add
754 it to the extractor list.
755 """
756 ie = self._ies_instances.get(ie_key)
757 if ie is None:
758 ie = get_info_extractor(ie_key)()
759 self.add_info_extractor(ie)
760 return ie
761
762 def add_default_info_extractors(self):
763 """
764 Add the InfoExtractors returned by gen_extractors to the end of the list
765 """
766 for ie in gen_extractor_classes():
767 self.add_info_extractor(ie)
768
769 def add_post_processor(self, pp, when='post_process'):
770 """Add a PostProcessor object to the end of the chain."""
771 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
772 self._pps[when].append(pp)
773 pp.set_downloader(self)
774
775 def add_post_hook(self, ph):
776 """Add the post hook"""
777 self._post_hooks.append(ph)
778
779 def add_progress_hook(self, ph):
780 """Add the download progress hook"""
781 self._progress_hooks.append(ph)
782
783 def add_postprocessor_hook(self, ph):
784 """Add the postprocessing progress hook"""
785 self._postprocessor_hooks.append(ph)
786 for pps in self._pps.values():
787 for pp in pps:
788 pp.add_progress_hook(ph)
789
790 def _bidi_workaround(self, message):
791 if not hasattr(self, '_output_channel'):
792 return message
793
794 assert hasattr(self, '_output_process')
795 assert isinstance(message, str)
796 line_count = message.count('\n') + 1
797 self._output_process.stdin.write((message + '\n').encode())
798 self._output_process.stdin.flush()
799 res = ''.join(self._output_channel.readline().decode()
800 for _ in range(line_count))
801 return res[:-len('\n')]
802
803 def _write_string(self, message, out=None, only_once=False):
804 if only_once:
805 if message in self._printed_messages:
806 return
807 self._printed_messages.add(message)
808 write_string(message, out=out, encoding=self.params.get('encoding'))
809
810 def to_stdout(self, message, skip_eol=False, quiet=None):
811 """Print message to stdout"""
812 if quiet is not None:
813 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
814 if skip_eol is not False:
815 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. Use "YoutubeDL.to_screen" instead')
816 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
817
818 def to_screen(self, message, skip_eol=False, quiet=None):
819 """Print message to screen if not in quiet mode"""
820 if self.params.get('logger'):
821 self.params['logger'].debug(message)
822 return
823 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
824 return
825 self._write_string(
826 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
827 self._out_files.screen)
828
829 def to_stderr(self, message, only_once=False):
830 """Print message to stderr"""
831 assert isinstance(message, str)
832 if self.params.get('logger'):
833 self.params['logger'].error(message)
834 else:
835 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
836
837 def _send_console_code(self, code):
838 if compat_os_name == 'nt' or not self._out_files.console:
839 return
840 self._write_string(code, self._out_files.console)
841
842 def to_console_title(self, message):
843 if not self.params.get('consoletitle', False):
844 return
845 message = remove_terminal_sequences(message)
846 if compat_os_name == 'nt':
847 if ctypes.windll.kernel32.GetConsoleWindow():
848 # c_wchar_p() might not be necessary if `message` is
849 # already of type unicode()
850 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
851 else:
852 self._send_console_code(f'\033]0;{message}\007')
853
854 def save_console_title(self):
855 if not self.params.get('consoletitle') or self.params.get('simulate'):
856 return
857 self._send_console_code('\033[22;0t') # Save the title on stack
858
859 def restore_console_title(self):
860 if not self.params.get('consoletitle') or self.params.get('simulate'):
861 return
862 self._send_console_code('\033[23;0t') # Restore the title from stack
863
864 def __enter__(self):
865 self.save_console_title()
866 return self
867
868 def __exit__(self, *args):
869 self.restore_console_title()
870
871 if self.params.get('cookiefile') is not None:
872 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
873
874 def trouble(self, message=None, tb=None, is_error=True):
875 """Determine action to take when a download problem appears.
876
877 Depending on if the downloader has been configured to ignore
878 download errors or not, this method may throw an exception or
879 not when errors are found, after printing the message.
880
881 @param tb If given, is additional traceback information
882 @param is_error Whether to raise error according to ignorerrors
883 """
884 if message is not None:
885 self.to_stderr(message)
886 if self.params.get('verbose'):
887 if tb is None:
888 if sys.exc_info()[0]: # if .trouble has been called from an except block
889 tb = ''
890 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
891 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
892 tb += encode_compat_str(traceback.format_exc())
893 else:
894 tb_data = traceback.format_list(traceback.extract_stack())
895 tb = ''.join(tb_data)
896 if tb:
897 self.to_stderr(tb)
898 if not is_error:
899 return
900 if not self.params.get('ignoreerrors'):
901 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
902 exc_info = sys.exc_info()[1].exc_info
903 else:
904 exc_info = sys.exc_info()
905 raise DownloadError(message, exc_info)
906 self._download_retcode = 1
907
908 Styles = Namespace(
909 HEADERS='yellow',
910 EMPHASIS='light blue',
911 FILENAME='green',
912 ID='green',
913 DELIM='blue',
914 ERROR='red',
915 WARNING='yellow',
916 SUPPRESS='light black',
917 )
918
919 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
920 text = str(text)
921 if test_encoding:
922 original_text = text
923 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
924 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
925 text = text.encode(encoding, 'ignore').decode(encoding)
926 if fallback is not None and text != original_text:
927 text = fallback
928 return format_text(text, f) if allow_colors else text if fallback is None else fallback
929
930 def _format_out(self, *args, **kwargs):
931 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
932
933 def _format_screen(self, *args, **kwargs):
934 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
935
936 def _format_err(self, *args, **kwargs):
937 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
938
939 def report_warning(self, message, only_once=False):
940 '''
941 Print the message to stderr, it will be prefixed with 'WARNING:'
942 If stderr is a tty file the 'WARNING:' will be colored
943 '''
944 if self.params.get('logger') is not None:
945 self.params['logger'].warning(message)
946 else:
947 if self.params.get('no_warnings'):
948 return
949 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
950
951 def deprecation_warning(self, message):
952 if self.params.get('logger') is not None:
953 self.params['logger'].warning(f'DeprecationWarning: {message}')
954 else:
955 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
956
957 def report_error(self, message, *args, **kwargs):
958 '''
959 Do the same as trouble, but prefixes the message with 'ERROR:', colored
960 in red if stderr is a tty file.
961 '''
962 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
963
964 def write_debug(self, message, only_once=False):
965 '''Log debug message or Print message to stderr'''
966 if not self.params.get('verbose', False):
967 return
968 message = f'[debug] {message}'
969 if self.params.get('logger'):
970 self.params['logger'].debug(message)
971 else:
972 self.to_stderr(message, only_once)
973
974 def report_file_already_downloaded(self, file_name):
975 """Report file has already been fully downloaded."""
976 try:
977 self.to_screen('[download] %s has already been downloaded' % file_name)
978 except UnicodeEncodeError:
979 self.to_screen('[download] The file has already been downloaded')
980
981 def report_file_delete(self, file_name):
982 """Report that existing file will be deleted."""
983 try:
984 self.to_screen('Deleting existing file %s' % file_name)
985 except UnicodeEncodeError:
986 self.to_screen('Deleting existing file')
987
988 def raise_no_formats(self, info, forced=False, *, msg=None):
989 has_drm = info.get('_has_drm')
990 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
991 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
992 if forced or not ignored:
993 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
994 expected=has_drm or ignored or expected)
995 else:
996 self.report_warning(msg)
997
998 def parse_outtmpl(self):
999 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1000 self._parse_outtmpl()
1001 return self.params['outtmpl']
1002
1003 def _parse_outtmpl(self):
1004 sanitize = IDENTITY
1005 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1006 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1007
1008 outtmpl = self.params.setdefault('outtmpl', {})
1009 if not isinstance(outtmpl, dict):
1010 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1011 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1012
1013 def get_output_path(self, dir_type='', filename=None):
1014 paths = self.params.get('paths', {})
1015 assert isinstance(paths, dict)
1016 path = os.path.join(
1017 expand_path(paths.get('home', '').strip()),
1018 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1019 filename or '')
1020 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1021
1022 @staticmethod
1023 def _outtmpl_expandpath(outtmpl):
1024 # expand_path translates '%%' into '%' and '$$' into '$'
1025 # correspondingly that is not what we want since we need to keep
1026 # '%%' intact for template dict substitution step. Working around
1027 # with boundary-alike separator hack.
1028 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1029 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1030
1031 # outtmpl should be expand_path'ed before template dict substitution
1032 # because meta fields may contain env variables we don't want to
1033 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1034 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1035 return expand_path(outtmpl).replace(sep, '')
1036
1037 @staticmethod
1038 def escape_outtmpl(outtmpl):
1039 ''' Escape any remaining strings like %s, %abc% etc. '''
1040 return re.sub(
1041 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1042 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1043 outtmpl)
1044
1045 @classmethod
1046 def validate_outtmpl(cls, outtmpl):
1047 ''' @return None or Exception object '''
1048 outtmpl = re.sub(
1049 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1050 lambda mobj: f'{mobj.group(0)[:-1]}s',
1051 cls._outtmpl_expandpath(outtmpl))
1052 try:
1053 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1054 return None
1055 except ValueError as err:
1056 return err
1057
1058 @staticmethod
1059 def _copy_infodict(info_dict):
1060 info_dict = dict(info_dict)
1061 info_dict.pop('__postprocessors', None)
1062 info_dict.pop('__pending_error', None)
1063 return info_dict
1064
1065 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1066 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1067 @param sanitize Whether to sanitize the output as a filename.
1068 For backward compatibility, a function can also be passed
1069 """
1070
1071 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1072
1073 info_dict = self._copy_infodict(info_dict)
1074 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1075 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1076 if info_dict.get('duration', None) is not None
1077 else None)
1078 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1079 info_dict['video_autonumber'] = self._num_videos
1080 if info_dict.get('resolution') is None:
1081 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1082
1083 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1084 # of %(field)s to %(field)0Nd for backward compatibility
1085 field_size_compat_map = {
1086 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1087 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1088 'autonumber': self.params.get('autonumber_size') or 5,
1089 }
1090
1091 TMPL_DICT = {}
1092 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1093 MATH_FUNCTIONS = {
1094 '+': float.__add__,
1095 '-': float.__sub__,
1096 }
1097 # Field is of the form key1.key2...
1098 # where keys (except first) can be string, int or slice
1099 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1100 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1101 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1102 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1103 (?P<negate>-)?
1104 (?P<fields>{FIELD_RE})
1105 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1106 (?:>(?P<strf_format>.+?))?
1107 (?P<remaining>
1108 (?P<alternate>(?<!\\),[^|&)]+)?
1109 (?:&(?P<replacement>.*?))?
1110 (?:\|(?P<default>.*?))?
1111 )$''')
1112
1113 def _traverse_infodict(k):
1114 k = k.split('.')
1115 if k[0] == '':
1116 k.pop(0)
1117 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1118
1119 def get_value(mdict):
1120 # Object traversal
1121 value = _traverse_infodict(mdict['fields'])
1122 # Negative
1123 if mdict['negate']:
1124 value = float_or_none(value)
1125 if value is not None:
1126 value *= -1
1127 # Do maths
1128 offset_key = mdict['maths']
1129 if offset_key:
1130 value = float_or_none(value)
1131 operator = None
1132 while offset_key:
1133 item = re.match(
1134 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1135 offset_key).group(0)
1136 offset_key = offset_key[len(item):]
1137 if operator is None:
1138 operator = MATH_FUNCTIONS[item]
1139 continue
1140 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1141 offset = float_or_none(item)
1142 if offset is None:
1143 offset = float_or_none(_traverse_infodict(item))
1144 try:
1145 value = operator(value, multiplier * offset)
1146 except (TypeError, ZeroDivisionError):
1147 return None
1148 operator = None
1149 # Datetime formatting
1150 if mdict['strf_format']:
1151 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1152
1153 return value
1154
1155 na = self.params.get('outtmpl_na_placeholder', 'NA')
1156
1157 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1158 return sanitize_filename(str(value), restricted=restricted, is_id=(
1159 bool(re.search(r'(^|[_.])id(\.|$)', key))
1160 if 'filename-sanitization' in self.params['compat_opts']
1161 else NO_DEFAULT))
1162
1163 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1164 sanitize = bool(sanitize)
1165
1166 def _dumpjson_default(obj):
1167 if isinstance(obj, (set, LazyList)):
1168 return list(obj)
1169 return repr(obj)
1170
1171 def create_key(outer_mobj):
1172 if not outer_mobj.group('has_key'):
1173 return outer_mobj.group(0)
1174 key = outer_mobj.group('key')
1175 mobj = re.match(INTERNAL_FORMAT_RE, key)
1176 initial_field = mobj.group('fields') if mobj else ''
1177 value, replacement, default = None, None, na
1178 while mobj:
1179 mobj = mobj.groupdict()
1180 default = mobj['default'] if mobj['default'] is not None else default
1181 value = get_value(mobj)
1182 replacement = mobj['replacement']
1183 if value is None and mobj['alternate']:
1184 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1185 else:
1186 break
1187
1188 fmt = outer_mobj.group('format')
1189 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1190 fmt = f'0{field_size_compat_map[key]:d}d'
1191
1192 value = default if value is None else value if replacement is None else replacement
1193
1194 flags = outer_mobj.group('conversion') or ''
1195 str_fmt = f'{fmt[:-1]}s'
1196 if fmt[-1] == 'l': # list
1197 delim = '\n' if '#' in flags else ', '
1198 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1199 elif fmt[-1] == 'j': # json
1200 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1201 elif fmt[-1] == 'q': # quoted
1202 value = map(str, variadic(value) if '#' in flags else [value])
1203 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1204 elif fmt[-1] == 'B': # bytes
1205 value = f'%{str_fmt}'.encode() % str(value).encode()
1206 value, fmt = value.decode('utf-8', 'ignore'), 's'
1207 elif fmt[-1] == 'U': # unicode normalized
1208 value, fmt = unicodedata.normalize(
1209 # "+" = compatibility equivalence, "#" = NFD
1210 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1211 value), str_fmt
1212 elif fmt[-1] == 'D': # decimal suffix
1213 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1214 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1215 factor=1024 if '#' in flags else 1000)
1216 elif fmt[-1] == 'S': # filename sanitization
1217 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1218 elif fmt[-1] == 'c':
1219 if value:
1220 value = str(value)[0]
1221 else:
1222 fmt = str_fmt
1223 elif fmt[-1] not in 'rs': # numeric
1224 value = float_or_none(value)
1225 if value is None:
1226 value, fmt = default, 's'
1227
1228 if sanitize:
1229 if fmt[-1] == 'r':
1230 # If value is an object, sanitize might convert it to a string
1231 # So we convert it to repr first
1232 value, fmt = repr(value), str_fmt
1233 if fmt[-1] in 'csr':
1234 value = sanitizer(initial_field, value)
1235
1236 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1237 TMPL_DICT[key] = value
1238 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1239
1240 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1241
1242 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1243 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1244 return self.escape_outtmpl(outtmpl) % info_dict
1245
1246 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1247 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1248 if outtmpl is None:
1249 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1250 try:
1251 outtmpl = self._outtmpl_expandpath(outtmpl)
1252 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1253 if not filename:
1254 return None
1255
1256 if tmpl_type in ('', 'temp'):
1257 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1258 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1259 filename = replace_extension(filename, ext, final_ext)
1260 elif tmpl_type:
1261 force_ext = OUTTMPL_TYPES[tmpl_type]
1262 if force_ext:
1263 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1264
1265 # https://github.com/blackjack4494/youtube-dlc/issues/85
1266 trim_file_name = self.params.get('trim_file_name', False)
1267 if trim_file_name:
1268 no_ext, *ext = filename.rsplit('.', 2)
1269 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1270
1271 return filename
1272 except ValueError as err:
1273 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1274 return None
1275
1276 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1277 """Generate the output filename"""
1278 if outtmpl:
1279 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1280 dir_type = None
1281 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1282 if not filename and dir_type not in ('', 'temp'):
1283 return ''
1284
1285 if warn:
1286 if not self.params.get('paths'):
1287 pass
1288 elif filename == '-':
1289 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1290 elif os.path.isabs(filename):
1291 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1292 if filename == '-' or not filename:
1293 return filename
1294
1295 return self.get_output_path(dir_type, filename)
1296
1297 def _match_entry(self, info_dict, incomplete=False, silent=False):
1298 """ Returns None if the file should be downloaded """
1299
1300 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1301
1302 def check_filter():
1303 if 'title' in info_dict:
1304 # This can happen when we're just evaluating the playlist
1305 title = info_dict['title']
1306 matchtitle = self.params.get('matchtitle', False)
1307 if matchtitle:
1308 if not re.search(matchtitle, title, re.IGNORECASE):
1309 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1310 rejecttitle = self.params.get('rejecttitle', False)
1311 if rejecttitle:
1312 if re.search(rejecttitle, title, re.IGNORECASE):
1313 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1314 date = info_dict.get('upload_date')
1315 if date is not None:
1316 dateRange = self.params.get('daterange', DateRange())
1317 if date not in dateRange:
1318 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1319 view_count = info_dict.get('view_count')
1320 if view_count is not None:
1321 min_views = self.params.get('min_views')
1322 if min_views is not None and view_count < min_views:
1323 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1324 max_views = self.params.get('max_views')
1325 if max_views is not None and view_count > max_views:
1326 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1327 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1328 return 'Skipping "%s" because it is age restricted' % video_title
1329
1330 match_filter = self.params.get('match_filter')
1331 if match_filter is not None:
1332 try:
1333 ret = match_filter(info_dict, incomplete=incomplete)
1334 except TypeError:
1335 # For backward compatibility
1336 ret = None if incomplete else match_filter(info_dict)
1337 if ret is NO_DEFAULT:
1338 while True:
1339 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1340 reply = input(self._format_screen(
1341 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1342 if reply in {'y', ''}:
1343 return None
1344 elif reply == 'n':
1345 return f'Skipping {video_title}'
1346 elif ret is not None:
1347 return ret
1348 return None
1349
1350 if self.in_download_archive(info_dict):
1351 reason = '%s has already been recorded in the archive' % video_title
1352 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1353 else:
1354 reason = check_filter()
1355 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1356 if reason is not None:
1357 if not silent:
1358 self.to_screen('[download] ' + reason)
1359 if self.params.get(break_opt, False):
1360 raise break_err()
1361 return reason
1362
1363 @staticmethod
1364 def add_extra_info(info_dict, extra_info):
1365 '''Set the keys from extra_info in info dict if they are missing'''
1366 for key, value in extra_info.items():
1367 info_dict.setdefault(key, value)
1368
1369 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1370 process=True, force_generic_extractor=False):
1371 """
1372 Return a list with a dictionary for each video extracted.
1373
1374 Arguments:
1375 url -- URL to extract
1376
1377 Keyword arguments:
1378 download -- whether to download videos during extraction
1379 ie_key -- extractor key hint
1380 extra_info -- dictionary containing the extra values to add to each result
1381 process -- whether to resolve all unresolved references (URLs, playlist items),
1382 must be True for download to work.
1383 force_generic_extractor -- force using the generic extractor
1384 """
1385
1386 if extra_info is None:
1387 extra_info = {}
1388
1389 if not ie_key and force_generic_extractor:
1390 ie_key = 'Generic'
1391
1392 if ie_key:
1393 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1394 else:
1395 ies = self._ies
1396
1397 for ie_key, ie in ies.items():
1398 if not ie.suitable(url):
1399 continue
1400
1401 if not ie.working():
1402 self.report_warning('The program functionality for this site has been marked as broken, '
1403 'and will probably not work.')
1404
1405 temp_id = ie.get_temp_id(url)
1406 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1407 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1408 if self.params.get('break_on_existing', False):
1409 raise ExistingVideoReached()
1410 break
1411 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1412 else:
1413 self.report_error('no suitable InfoExtractor for URL %s' % url)
1414
1415 def _handle_extraction_exceptions(func):
1416 @functools.wraps(func)
1417 def wrapper(self, *args, **kwargs):
1418 while True:
1419 try:
1420 return func(self, *args, **kwargs)
1421 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1422 raise
1423 except ReExtractInfo as e:
1424 if e.expected:
1425 self.to_screen(f'{e}; Re-extracting data')
1426 else:
1427 self.to_stderr('\r')
1428 self.report_warning(f'{e}; Re-extracting data')
1429 continue
1430 except GeoRestrictedError as e:
1431 msg = e.msg
1432 if e.countries:
1433 msg += '\nThis video is available in %s.' % ', '.join(
1434 map(ISO3166Utils.short2full, e.countries))
1435 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1436 self.report_error(msg)
1437 except ExtractorError as e: # An error we somewhat expected
1438 self.report_error(str(e), e.format_traceback())
1439 except Exception as e:
1440 if self.params.get('ignoreerrors'):
1441 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1442 else:
1443 raise
1444 break
1445 return wrapper
1446
1447 def _wait_for_video(self, ie_result):
1448 if (not self.params.get('wait_for_video')
1449 or ie_result.get('_type', 'video') != 'video'
1450 or ie_result.get('formats') or ie_result.get('url')):
1451 return
1452
1453 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1454 last_msg = ''
1455
1456 def progress(msg):
1457 nonlocal last_msg
1458 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1459 last_msg = msg
1460
1461 min_wait, max_wait = self.params.get('wait_for_video')
1462 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1463 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1464 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1465 self.report_warning('Release time of video is not known')
1466 elif (diff or 0) <= 0:
1467 self.report_warning('Video should already be available according to extracted info')
1468 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1469 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1470
1471 wait_till = time.time() + diff
1472 try:
1473 while True:
1474 diff = wait_till - time.time()
1475 if diff <= 0:
1476 progress('')
1477 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1478 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1479 time.sleep(1)
1480 except KeyboardInterrupt:
1481 progress('')
1482 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1483 except BaseException as e:
1484 if not isinstance(e, ReExtractInfo):
1485 self.to_screen('')
1486 raise
1487
1488 @_handle_extraction_exceptions
1489 def __extract_info(self, url, ie, download, extra_info, process):
1490 ie_result = ie.extract(url)
1491 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1492 return
1493 if isinstance(ie_result, list):
1494 # Backwards compatibility: old IE result format
1495 ie_result = {
1496 '_type': 'compat_list',
1497 'entries': ie_result,
1498 }
1499 if extra_info.get('original_url'):
1500 ie_result.setdefault('original_url', extra_info['original_url'])
1501 self.add_default_extra_info(ie_result, ie, url)
1502 if process:
1503 self._wait_for_video(ie_result)
1504 return self.process_ie_result(ie_result, download, extra_info)
1505 else:
1506 return ie_result
1507
1508 def add_default_extra_info(self, ie_result, ie, url):
1509 if url is not None:
1510 self.add_extra_info(ie_result, {
1511 'webpage_url': url,
1512 'original_url': url,
1513 })
1514 webpage_url = ie_result.get('webpage_url')
1515 if webpage_url:
1516 self.add_extra_info(ie_result, {
1517 'webpage_url_basename': url_basename(webpage_url),
1518 'webpage_url_domain': get_domain(webpage_url),
1519 })
1520 if ie is not None:
1521 self.add_extra_info(ie_result, {
1522 'extractor': ie.IE_NAME,
1523 'extractor_key': ie.ie_key(),
1524 })
1525
1526 def process_ie_result(self, ie_result, download=True, extra_info=None):
1527 """
1528 Take the result of the ie(may be modified) and resolve all unresolved
1529 references (URLs, playlist items).
1530
1531 It will also download the videos if 'download'.
1532 Returns the resolved ie_result.
1533 """
1534 if extra_info is None:
1535 extra_info = {}
1536 result_type = ie_result.get('_type', 'video')
1537
1538 if result_type in ('url', 'url_transparent'):
1539 ie_result['url'] = sanitize_url(ie_result['url'])
1540 if ie_result.get('original_url'):
1541 extra_info.setdefault('original_url', ie_result['original_url'])
1542
1543 extract_flat = self.params.get('extract_flat', False)
1544 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1545 or extract_flat is True):
1546 info_copy = ie_result.copy()
1547 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1548 if ie and not ie_result.get('id'):
1549 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1550 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1551 self.add_extra_info(info_copy, extra_info)
1552 info_copy, _ = self.pre_process(info_copy)
1553 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1554 self._raise_pending_errors(info_copy)
1555 if self.params.get('force_write_download_archive', False):
1556 self.record_download_archive(info_copy)
1557 return ie_result
1558
1559 if result_type == 'video':
1560 self.add_extra_info(ie_result, extra_info)
1561 ie_result = self.process_video_result(ie_result, download=download)
1562 self._raise_pending_errors(ie_result)
1563 additional_urls = (ie_result or {}).get('additional_urls')
1564 if additional_urls:
1565 # TODO: Improve MetadataParserPP to allow setting a list
1566 if isinstance(additional_urls, str):
1567 additional_urls = [additional_urls]
1568 self.to_screen(
1569 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1570 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1571 ie_result['additional_entries'] = [
1572 self.extract_info(
1573 url, download, extra_info=extra_info,
1574 force_generic_extractor=self.params.get('force_generic_extractor'))
1575 for url in additional_urls
1576 ]
1577 return ie_result
1578 elif result_type == 'url':
1579 # We have to add extra_info to the results because it may be
1580 # contained in a playlist
1581 return self.extract_info(
1582 ie_result['url'], download,
1583 ie_key=ie_result.get('ie_key'),
1584 extra_info=extra_info)
1585 elif result_type == 'url_transparent':
1586 # Use the information from the embedding page
1587 info = self.extract_info(
1588 ie_result['url'], ie_key=ie_result.get('ie_key'),
1589 extra_info=extra_info, download=False, process=False)
1590
1591 # extract_info may return None when ignoreerrors is enabled and
1592 # extraction failed with an error, don't crash and return early
1593 # in this case
1594 if not info:
1595 return info
1596
1597 exempted_fields = {'_type', 'url', 'ie_key'}
1598 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1599 # For video clips, the id etc of the clip extractor should be used
1600 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1601
1602 new_result = info.copy()
1603 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1604
1605 # Extracted info may not be a video result (i.e.
1606 # info.get('_type', 'video') != video) but rather an url or
1607 # url_transparent. In such cases outer metadata (from ie_result)
1608 # should be propagated to inner one (info). For this to happen
1609 # _type of info should be overridden with url_transparent. This
1610 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1611 if new_result.get('_type') == 'url':
1612 new_result['_type'] = 'url_transparent'
1613
1614 return self.process_ie_result(
1615 new_result, download=download, extra_info=extra_info)
1616 elif result_type in ('playlist', 'multi_video'):
1617 # Protect from infinite recursion due to recursively nested playlists
1618 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1619 webpage_url = ie_result['webpage_url']
1620 if webpage_url in self._playlist_urls:
1621 self.to_screen(
1622 '[download] Skipping already downloaded playlist: %s'
1623 % ie_result.get('title') or ie_result.get('id'))
1624 return
1625
1626 self._playlist_level += 1
1627 self._playlist_urls.add(webpage_url)
1628 self._fill_common_fields(ie_result, False)
1629 self._sanitize_thumbnails(ie_result)
1630 try:
1631 return self.__process_playlist(ie_result, download)
1632 finally:
1633 self._playlist_level -= 1
1634 if not self._playlist_level:
1635 self._playlist_urls.clear()
1636 elif result_type == 'compat_list':
1637 self.report_warning(
1638 'Extractor %s returned a compat_list result. '
1639 'It needs to be updated.' % ie_result.get('extractor'))
1640
1641 def _fixup(r):
1642 self.add_extra_info(r, {
1643 'extractor': ie_result['extractor'],
1644 'webpage_url': ie_result['webpage_url'],
1645 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1646 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1647 'extractor_key': ie_result['extractor_key'],
1648 })
1649 return r
1650 ie_result['entries'] = [
1651 self.process_ie_result(_fixup(r), download, extra_info)
1652 for r in ie_result['entries']
1653 ]
1654 return ie_result
1655 else:
1656 raise Exception('Invalid result type: %s' % result_type)
1657
1658 def _ensure_dir_exists(self, path):
1659 return make_dir(path, self.report_error)
1660
1661 @staticmethod
1662 def _playlist_infodict(ie_result, **kwargs):
1663 return {
1664 **ie_result,
1665 'playlist': ie_result.get('title') or ie_result.get('id'),
1666 'playlist_id': ie_result.get('id'),
1667 'playlist_title': ie_result.get('title'),
1668 'playlist_uploader': ie_result.get('uploader'),
1669 'playlist_uploader_id': ie_result.get('uploader_id'),
1670 'playlist_index': 0,
1671 **kwargs,
1672 }
1673
1674 def __process_playlist(self, ie_result, download):
1675 """Process each entry in the playlist"""
1676 title = ie_result.get('title') or ie_result.get('id') or '<Untitled>'
1677 self.to_screen(f'[download] Downloading playlist: {title}')
1678
1679 all_entries = PlaylistEntries(self, ie_result)
1680 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1681
1682 lazy = self.params.get('lazy_playlist')
1683 if lazy:
1684 resolved_entries, n_entries = [], 'N/A'
1685 ie_result['requested_entries'], ie_result['entries'] = None, None
1686 else:
1687 entries = resolved_entries = list(entries)
1688 n_entries = len(resolved_entries)
1689 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1690 if not ie_result.get('playlist_count'):
1691 # Better to do this after potentially exhausting entries
1692 ie_result['playlist_count'] = all_entries.get_full_count()
1693
1694 _infojson_written = False
1695 write_playlist_files = self.params.get('allow_playlist_files', True)
1696 if write_playlist_files and self.params.get('list_thumbnails'):
1697 self.list_thumbnails(ie_result)
1698 if write_playlist_files and not self.params.get('simulate'):
1699 ie_copy = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1700 _infojson_written = self._write_info_json(
1701 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1702 if _infojson_written is None:
1703 return
1704 if self._write_description('playlist', ie_result,
1705 self.prepare_filename(ie_copy, 'pl_description')) is None:
1706 return
1707 # TODO: This should be passed to ThumbnailsConvertor if necessary
1708 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1709
1710 if lazy:
1711 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1712 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1713 elif self.params.get('playlistreverse'):
1714 entries.reverse()
1715 elif self.params.get('playlistrandom'):
1716 random.shuffle(entries)
1717
1718 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} videos'
1719 f'{format_field(ie_result, "playlist_count", " of %s")}')
1720
1721 failures = 0
1722 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1723 for i, (playlist_index, entry) in enumerate(entries):
1724 if lazy:
1725 resolved_entries.append((playlist_index, entry))
1726
1727 # TODO: Add auto-generated fields
1728 if not entry or self._match_entry(entry, incomplete=True) is not None:
1729 continue
1730
1731 self.to_screen('[download] Downloading video %s of %s' % (
1732 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1733
1734 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1735 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1736 playlist_index = ie_result['requested_entries'][i]
1737
1738 entry_result = self.__process_iterable_entry(entry, download, {
1739 'n_entries': int_or_none(n_entries),
1740 '__last_playlist_index': max(ie_result['requested_entries'] or (0, 0)),
1741 'playlist_count': ie_result.get('playlist_count'),
1742 'playlist_index': playlist_index,
1743 'playlist_autonumber': i + 1,
1744 'playlist': title,
1745 'playlist_id': ie_result.get('id'),
1746 'playlist_title': ie_result.get('title'),
1747 'playlist_uploader': ie_result.get('uploader'),
1748 'playlist_uploader_id': ie_result.get('uploader_id'),
1749 'extractor': ie_result['extractor'],
1750 'webpage_url': ie_result['webpage_url'],
1751 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1752 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1753 'extractor_key': ie_result['extractor_key'],
1754 })
1755 if not entry_result:
1756 failures += 1
1757 if failures >= max_failures:
1758 self.report_error(
1759 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1760 break
1761 resolved_entries[i] = (playlist_index, entry_result)
1762
1763 # Update with processed data
1764 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1765
1766 # Write the updated info to json
1767 if _infojson_written is True and self._write_info_json(
1768 'updated playlist', ie_result,
1769 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1770 return
1771
1772 ie_result = self.run_all_pps('playlist', ie_result)
1773 self.to_screen(f'[download] Finished downloading playlist: {title}')
1774 return ie_result
1775
1776 @_handle_extraction_exceptions
1777 def __process_iterable_entry(self, entry, download, extra_info):
1778 return self.process_ie_result(
1779 entry, download=download, extra_info=extra_info)
1780
1781 def _build_format_filter(self, filter_spec):
1782 " Returns a function to filter the formats according to the filter_spec "
1783
1784 OPERATORS = {
1785 '<': operator.lt,
1786 '<=': operator.le,
1787 '>': operator.gt,
1788 '>=': operator.ge,
1789 '=': operator.eq,
1790 '!=': operator.ne,
1791 }
1792 operator_rex = re.compile(r'''(?x)\s*
1793 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1794 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1795 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1796 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1797 m = operator_rex.fullmatch(filter_spec)
1798 if m:
1799 try:
1800 comparison_value = int(m.group('value'))
1801 except ValueError:
1802 comparison_value = parse_filesize(m.group('value'))
1803 if comparison_value is None:
1804 comparison_value = parse_filesize(m.group('value') + 'B')
1805 if comparison_value is None:
1806 raise ValueError(
1807 'Invalid value %r in format specification %r' % (
1808 m.group('value'), filter_spec))
1809 op = OPERATORS[m.group('op')]
1810
1811 if not m:
1812 STR_OPERATORS = {
1813 '=': operator.eq,
1814 '^=': lambda attr, value: attr.startswith(value),
1815 '$=': lambda attr, value: attr.endswith(value),
1816 '*=': lambda attr, value: value in attr,
1817 '~=': lambda attr, value: value.search(attr) is not None
1818 }
1819 str_operator_rex = re.compile(r'''(?x)\s*
1820 (?P<key>[a-zA-Z0-9._-]+)\s*
1821 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1822 (?P<quote>["'])?
1823 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1824 (?(quote)(?P=quote))\s*
1825 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1826 m = str_operator_rex.fullmatch(filter_spec)
1827 if m:
1828 if m.group('op') == '~=':
1829 comparison_value = re.compile(m.group('value'))
1830 else:
1831 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1832 str_op = STR_OPERATORS[m.group('op')]
1833 if m.group('negation'):
1834 op = lambda attr, value: not str_op(attr, value)
1835 else:
1836 op = str_op
1837
1838 if not m:
1839 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1840
1841 def _filter(f):
1842 actual_value = f.get(m.group('key'))
1843 if actual_value is None:
1844 return m.group('none_inclusive')
1845 return op(actual_value, comparison_value)
1846 return _filter
1847
1848 def _check_formats(self, formats):
1849 for f in formats:
1850 self.to_screen('[info] Testing format %s' % f['format_id'])
1851 path = self.get_output_path('temp')
1852 if not self._ensure_dir_exists(f'{path}/'):
1853 continue
1854 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1855 temp_file.close()
1856 try:
1857 success, _ = self.dl(temp_file.name, f, test=True)
1858 except (DownloadError, OSError, ValueError) + network_exceptions:
1859 success = False
1860 finally:
1861 if os.path.exists(temp_file.name):
1862 try:
1863 os.remove(temp_file.name)
1864 except OSError:
1865 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1866 if success:
1867 yield f
1868 else:
1869 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1870
1871 def _default_format_spec(self, info_dict, download=True):
1872
1873 def can_merge():
1874 merger = FFmpegMergerPP(self)
1875 return merger.available and merger.can_merge()
1876
1877 prefer_best = (
1878 not self.params.get('simulate')
1879 and download
1880 and (
1881 not can_merge()
1882 or info_dict.get('is_live') and not self.params.get('live_from_start')
1883 or self.params['outtmpl']['default'] == '-'))
1884 compat = (
1885 prefer_best
1886 or self.params.get('allow_multiple_audio_streams', False)
1887 or 'format-spec' in self.params['compat_opts'])
1888
1889 return (
1890 'best/bestvideo+bestaudio' if prefer_best
1891 else 'bestvideo*+bestaudio/best' if not compat
1892 else 'bestvideo+bestaudio/best')
1893
1894 def build_format_selector(self, format_spec):
1895 def syntax_error(note, start):
1896 message = (
1897 'Invalid format specification: '
1898 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1899 return SyntaxError(message)
1900
1901 PICKFIRST = 'PICKFIRST'
1902 MERGE = 'MERGE'
1903 SINGLE = 'SINGLE'
1904 GROUP = 'GROUP'
1905 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1906
1907 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1908 'video': self.params.get('allow_multiple_video_streams', False)}
1909
1910 check_formats = self.params.get('check_formats') == 'selected'
1911
1912 def _parse_filter(tokens):
1913 filter_parts = []
1914 for type, string, start, _, _ in tokens:
1915 if type == tokenize.OP and string == ']':
1916 return ''.join(filter_parts)
1917 else:
1918 filter_parts.append(string)
1919
1920 def _remove_unused_ops(tokens):
1921 # Remove operators that we don't use and join them with the surrounding strings
1922 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1923 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1924 last_string, last_start, last_end, last_line = None, None, None, None
1925 for type, string, start, end, line in tokens:
1926 if type == tokenize.OP and string == '[':
1927 if last_string:
1928 yield tokenize.NAME, last_string, last_start, last_end, last_line
1929 last_string = None
1930 yield type, string, start, end, line
1931 # everything inside brackets will be handled by _parse_filter
1932 for type, string, start, end, line in tokens:
1933 yield type, string, start, end, line
1934 if type == tokenize.OP and string == ']':
1935 break
1936 elif type == tokenize.OP and string in ALLOWED_OPS:
1937 if last_string:
1938 yield tokenize.NAME, last_string, last_start, last_end, last_line
1939 last_string = None
1940 yield type, string, start, end, line
1941 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1942 if not last_string:
1943 last_string = string
1944 last_start = start
1945 last_end = end
1946 else:
1947 last_string += string
1948 if last_string:
1949 yield tokenize.NAME, last_string, last_start, last_end, last_line
1950
1951 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1952 selectors = []
1953 current_selector = None
1954 for type, string, start, _, _ in tokens:
1955 # ENCODING is only defined in python 3.x
1956 if type == getattr(tokenize, 'ENCODING', None):
1957 continue
1958 elif type in [tokenize.NAME, tokenize.NUMBER]:
1959 current_selector = FormatSelector(SINGLE, string, [])
1960 elif type == tokenize.OP:
1961 if string == ')':
1962 if not inside_group:
1963 # ')' will be handled by the parentheses group
1964 tokens.restore_last_token()
1965 break
1966 elif inside_merge and string in ['/', ',']:
1967 tokens.restore_last_token()
1968 break
1969 elif inside_choice and string == ',':
1970 tokens.restore_last_token()
1971 break
1972 elif string == ',':
1973 if not current_selector:
1974 raise syntax_error('"," must follow a format selector', start)
1975 selectors.append(current_selector)
1976 current_selector = None
1977 elif string == '/':
1978 if not current_selector:
1979 raise syntax_error('"/" must follow a format selector', start)
1980 first_choice = current_selector
1981 second_choice = _parse_format_selection(tokens, inside_choice=True)
1982 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1983 elif string == '[':
1984 if not current_selector:
1985 current_selector = FormatSelector(SINGLE, 'best', [])
1986 format_filter = _parse_filter(tokens)
1987 current_selector.filters.append(format_filter)
1988 elif string == '(':
1989 if current_selector:
1990 raise syntax_error('Unexpected "("', start)
1991 group = _parse_format_selection(tokens, inside_group=True)
1992 current_selector = FormatSelector(GROUP, group, [])
1993 elif string == '+':
1994 if not current_selector:
1995 raise syntax_error('Unexpected "+"', start)
1996 selector_1 = current_selector
1997 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1998 if not selector_2:
1999 raise syntax_error('Expected a selector', start)
2000 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2001 else:
2002 raise syntax_error(f'Operator not recognized: "{string}"', start)
2003 elif type == tokenize.ENDMARKER:
2004 break
2005 if current_selector:
2006 selectors.append(current_selector)
2007 return selectors
2008
2009 def _merge(formats_pair):
2010 format_1, format_2 = formats_pair
2011
2012 formats_info = []
2013 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2014 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2015
2016 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2017 get_no_more = {'video': False, 'audio': False}
2018 for (i, fmt_info) in enumerate(formats_info):
2019 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2020 formats_info.pop(i)
2021 continue
2022 for aud_vid in ['audio', 'video']:
2023 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2024 if get_no_more[aud_vid]:
2025 formats_info.pop(i)
2026 break
2027 get_no_more[aud_vid] = True
2028
2029 if len(formats_info) == 1:
2030 return formats_info[0]
2031
2032 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2033 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2034
2035 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2036 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2037
2038 output_ext = self.params.get('merge_output_format')
2039 if not output_ext:
2040 if the_only_video:
2041 output_ext = the_only_video['ext']
2042 elif the_only_audio and not video_fmts:
2043 output_ext = the_only_audio['ext']
2044 else:
2045 output_ext = 'mkv'
2046
2047 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2048
2049 new_dict = {
2050 'requested_formats': formats_info,
2051 'format': '+'.join(filtered('format')),
2052 'format_id': '+'.join(filtered('format_id')),
2053 'ext': output_ext,
2054 'protocol': '+'.join(map(determine_protocol, formats_info)),
2055 'language': '+'.join(orderedSet(filtered('language'))) or None,
2056 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2057 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2058 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2059 }
2060
2061 if the_only_video:
2062 new_dict.update({
2063 'width': the_only_video.get('width'),
2064 'height': the_only_video.get('height'),
2065 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2066 'fps': the_only_video.get('fps'),
2067 'dynamic_range': the_only_video.get('dynamic_range'),
2068 'vcodec': the_only_video.get('vcodec'),
2069 'vbr': the_only_video.get('vbr'),
2070 'stretched_ratio': the_only_video.get('stretched_ratio'),
2071 })
2072
2073 if the_only_audio:
2074 new_dict.update({
2075 'acodec': the_only_audio.get('acodec'),
2076 'abr': the_only_audio.get('abr'),
2077 'asr': the_only_audio.get('asr'),
2078 })
2079
2080 return new_dict
2081
2082 def _check_formats(formats):
2083 if not check_formats:
2084 yield from formats
2085 return
2086 yield from self._check_formats(formats)
2087
2088 def _build_selector_function(selector):
2089 if isinstance(selector, list): # ,
2090 fs = [_build_selector_function(s) for s in selector]
2091
2092 def selector_function(ctx):
2093 for f in fs:
2094 yield from f(ctx)
2095 return selector_function
2096
2097 elif selector.type == GROUP: # ()
2098 selector_function = _build_selector_function(selector.selector)
2099
2100 elif selector.type == PICKFIRST: # /
2101 fs = [_build_selector_function(s) for s in selector.selector]
2102
2103 def selector_function(ctx):
2104 for f in fs:
2105 picked_formats = list(f(ctx))
2106 if picked_formats:
2107 return picked_formats
2108 return []
2109
2110 elif selector.type == MERGE: # +
2111 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2112
2113 def selector_function(ctx):
2114 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2115 yield _merge(pair)
2116
2117 elif selector.type == SINGLE: # atom
2118 format_spec = selector.selector or 'best'
2119
2120 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2121 if format_spec == 'all':
2122 def selector_function(ctx):
2123 yield from _check_formats(ctx['formats'][::-1])
2124 elif format_spec == 'mergeall':
2125 def selector_function(ctx):
2126 formats = list(_check_formats(
2127 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2128 if not formats:
2129 return
2130 merged_format = formats[-1]
2131 for f in formats[-2::-1]:
2132 merged_format = _merge((merged_format, f))
2133 yield merged_format
2134
2135 else:
2136 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2137 mobj = re.match(
2138 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2139 format_spec)
2140 if mobj is not None:
2141 format_idx = int_or_none(mobj.group('n'), default=1)
2142 format_reverse = mobj.group('bw')[0] == 'b'
2143 format_type = (mobj.group('type') or [None])[0]
2144 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2145 format_modified = mobj.group('mod') is not None
2146
2147 format_fallback = not format_type and not format_modified # for b, w
2148 _filter_f = (
2149 (lambda f: f.get('%scodec' % format_type) != 'none')
2150 if format_type and format_modified # bv*, ba*, wv*, wa*
2151 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2152 if format_type # bv, ba, wv, wa
2153 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2154 if not format_modified # b, w
2155 else lambda f: True) # b*, w*
2156 filter_f = lambda f: _filter_f(f) and (
2157 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2158 else:
2159 if format_spec in self._format_selection_exts['audio']:
2160 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2161 elif format_spec in self._format_selection_exts['video']:
2162 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2163 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2164 elif format_spec in self._format_selection_exts['storyboards']:
2165 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2166 else:
2167 filter_f = lambda f: f.get('format_id') == format_spec # id
2168
2169 def selector_function(ctx):
2170 formats = list(ctx['formats'])
2171 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2172 if not matches:
2173 if format_fallback and ctx['incomplete_formats']:
2174 # for extractors with incomplete formats (audio only (soundcloud)
2175 # or video only (imgur)) best/worst will fallback to
2176 # best/worst {video,audio}-only format
2177 matches = formats
2178 elif seperate_fallback and not ctx['has_merged_format']:
2179 # for compatibility with youtube-dl when there is no pre-merged format
2180 matches = list(filter(seperate_fallback, formats))
2181 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2182 try:
2183 yield matches[format_idx - 1]
2184 except LazyList.IndexError:
2185 return
2186
2187 filters = [self._build_format_filter(f) for f in selector.filters]
2188
2189 def final_selector(ctx):
2190 ctx_copy = dict(ctx)
2191 for _filter in filters:
2192 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2193 return selector_function(ctx_copy)
2194 return final_selector
2195
2196 stream = io.BytesIO(format_spec.encode())
2197 try:
2198 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2199 except tokenize.TokenError:
2200 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2201
2202 class TokenIterator:
2203 def __init__(self, tokens):
2204 self.tokens = tokens
2205 self.counter = 0
2206
2207 def __iter__(self):
2208 return self
2209
2210 def __next__(self):
2211 if self.counter >= len(self.tokens):
2212 raise StopIteration()
2213 value = self.tokens[self.counter]
2214 self.counter += 1
2215 return value
2216
2217 next = __next__
2218
2219 def restore_last_token(self):
2220 self.counter -= 1
2221
2222 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2223 return _build_selector_function(parsed_selector)
2224
2225 def _calc_headers(self, info_dict):
2226 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2227
2228 cookies = self._calc_cookies(info_dict['url'])
2229 if cookies:
2230 res['Cookie'] = cookies
2231
2232 if 'X-Forwarded-For' not in res:
2233 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2234 if x_forwarded_for_ip:
2235 res['X-Forwarded-For'] = x_forwarded_for_ip
2236
2237 return res
2238
2239 def _calc_cookies(self, url):
2240 pr = sanitized_Request(url)
2241 self.cookiejar.add_cookie_header(pr)
2242 return pr.get_header('Cookie')
2243
2244 def _sort_thumbnails(self, thumbnails):
2245 thumbnails.sort(key=lambda t: (
2246 t.get('preference') if t.get('preference') is not None else -1,
2247 t.get('width') if t.get('width') is not None else -1,
2248 t.get('height') if t.get('height') is not None else -1,
2249 t.get('id') if t.get('id') is not None else '',
2250 t.get('url')))
2251
2252 def _sanitize_thumbnails(self, info_dict):
2253 thumbnails = info_dict.get('thumbnails')
2254 if thumbnails is None:
2255 thumbnail = info_dict.get('thumbnail')
2256 if thumbnail:
2257 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2258 if not thumbnails:
2259 return
2260
2261 def check_thumbnails(thumbnails):
2262 for t in thumbnails:
2263 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2264 try:
2265 self.urlopen(HEADRequest(t['url']))
2266 except network_exceptions as err:
2267 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2268 continue
2269 yield t
2270
2271 self._sort_thumbnails(thumbnails)
2272 for i, t in enumerate(thumbnails):
2273 if t.get('id') is None:
2274 t['id'] = '%d' % i
2275 if t.get('width') and t.get('height'):
2276 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2277 t['url'] = sanitize_url(t['url'])
2278
2279 if self.params.get('check_formats') is True:
2280 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2281 else:
2282 info_dict['thumbnails'] = thumbnails
2283
2284 def _fill_common_fields(self, info_dict, is_video=True):
2285 # TODO: move sanitization here
2286 if is_video:
2287 # playlists are allowed to lack "title"
2288 title = info_dict.get('title', NO_DEFAULT)
2289 if title is NO_DEFAULT:
2290 raise ExtractorError('Missing "title" field in extractor result',
2291 video_id=info_dict['id'], ie=info_dict['extractor'])
2292 info_dict['fulltitle'] = title
2293 if not title:
2294 if title == '':
2295 self.write_debug('Extractor gave empty title. Creating a generic title')
2296 else:
2297 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2298 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2299
2300 if info_dict.get('duration') is not None:
2301 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2302
2303 for ts_key, date_key in (
2304 ('timestamp', 'upload_date'),
2305 ('release_timestamp', 'release_date'),
2306 ('modified_timestamp', 'modified_date'),
2307 ):
2308 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2309 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2310 # see http://bugs.python.org/issue1646728)
2311 with contextlib.suppress(ValueError, OverflowError, OSError):
2312 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2313 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2314
2315 live_keys = ('is_live', 'was_live')
2316 live_status = info_dict.get('live_status')
2317 if live_status is None:
2318 for key in live_keys:
2319 if info_dict.get(key) is False:
2320 continue
2321 if info_dict.get(key):
2322 live_status = key
2323 break
2324 if all(info_dict.get(key) is False for key in live_keys):
2325 live_status = 'not_live'
2326 if live_status:
2327 info_dict['live_status'] = live_status
2328 for key in live_keys:
2329 if info_dict.get(key) is None:
2330 info_dict[key] = (live_status == key)
2331
2332 # Auto generate title fields corresponding to the *_number fields when missing
2333 # in order to always have clean titles. This is very common for TV series.
2334 for field in ('chapter', 'season', 'episode'):
2335 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2336 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2337
2338 def _raise_pending_errors(self, info):
2339 err = info.pop('__pending_error', None)
2340 if err:
2341 self.report_error(err, tb=False)
2342
2343 def process_video_result(self, info_dict, download=True):
2344 assert info_dict.get('_type', 'video') == 'video'
2345 self._num_videos += 1
2346
2347 if 'id' not in info_dict:
2348 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2349 elif not info_dict.get('id'):
2350 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2351
2352 def report_force_conversion(field, field_not, conversion):
2353 self.report_warning(
2354 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2355 % (field, field_not, conversion))
2356
2357 def sanitize_string_field(info, string_field):
2358 field = info.get(string_field)
2359 if field is None or isinstance(field, str):
2360 return
2361 report_force_conversion(string_field, 'a string', 'string')
2362 info[string_field] = str(field)
2363
2364 def sanitize_numeric_fields(info):
2365 for numeric_field in self._NUMERIC_FIELDS:
2366 field = info.get(numeric_field)
2367 if field is None or isinstance(field, (int, float)):
2368 continue
2369 report_force_conversion(numeric_field, 'numeric', 'int')
2370 info[numeric_field] = int_or_none(field)
2371
2372 sanitize_string_field(info_dict, 'id')
2373 sanitize_numeric_fields(info_dict)
2374 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2375 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2376 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2377 self.report_warning('"duration" field is negative, there is an error in extractor')
2378
2379 chapters = info_dict.get('chapters') or []
2380 if chapters and chapters[0].get('start_time'):
2381 chapters.insert(0, {'start_time': 0})
2382
2383 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2384 for idx, (prev, current, next_) in enumerate(zip(
2385 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2386 if current.get('start_time') is None:
2387 current['start_time'] = prev.get('end_time')
2388 if not current.get('end_time'):
2389 current['end_time'] = next_.get('start_time')
2390 if not current.get('title'):
2391 current['title'] = f'<Untitled Chapter {idx}>'
2392
2393 if 'playlist' not in info_dict:
2394 # It isn't part of a playlist
2395 info_dict['playlist'] = None
2396 info_dict['playlist_index'] = None
2397
2398 self._sanitize_thumbnails(info_dict)
2399
2400 thumbnail = info_dict.get('thumbnail')
2401 thumbnails = info_dict.get('thumbnails')
2402 if thumbnail:
2403 info_dict['thumbnail'] = sanitize_url(thumbnail)
2404 elif thumbnails:
2405 info_dict['thumbnail'] = thumbnails[-1]['url']
2406
2407 if info_dict.get('display_id') is None and 'id' in info_dict:
2408 info_dict['display_id'] = info_dict['id']
2409
2410 self._fill_common_fields(info_dict)
2411
2412 for cc_kind in ('subtitles', 'automatic_captions'):
2413 cc = info_dict.get(cc_kind)
2414 if cc:
2415 for _, subtitle in cc.items():
2416 for subtitle_format in subtitle:
2417 if subtitle_format.get('url'):
2418 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2419 if subtitle_format.get('ext') is None:
2420 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2421
2422 automatic_captions = info_dict.get('automatic_captions')
2423 subtitles = info_dict.get('subtitles')
2424
2425 info_dict['requested_subtitles'] = self.process_subtitles(
2426 info_dict['id'], subtitles, automatic_captions)
2427
2428 if info_dict.get('formats') is None:
2429 # There's only one format available
2430 formats = [info_dict]
2431 else:
2432 formats = info_dict['formats']
2433
2434 # or None ensures --clean-infojson removes it
2435 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2436 if not self.params.get('allow_unplayable_formats'):
2437 formats = [f for f in formats if not f.get('has_drm')]
2438 if info_dict['_has_drm'] and all(
2439 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2440 self.report_warning(
2441 'This video is DRM protected and only images are available for download. '
2442 'Use --list-formats to see them')
2443
2444 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2445 if not get_from_start:
2446 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2447 if info_dict.get('is_live') and formats:
2448 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2449 if get_from_start and not formats:
2450 self.raise_no_formats(info_dict, msg=(
2451 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2452 'If you want to download from the current time, use --no-live-from-start'))
2453
2454 if not formats:
2455 self.raise_no_formats(info_dict)
2456
2457 def is_wellformed(f):
2458 url = f.get('url')
2459 if not url:
2460 self.report_warning(
2461 '"url" field is missing or empty - skipping format, '
2462 'there is an error in extractor')
2463 return False
2464 if isinstance(url, bytes):
2465 sanitize_string_field(f, 'url')
2466 return True
2467
2468 # Filter out malformed formats for better extraction robustness
2469 formats = list(filter(is_wellformed, formats))
2470
2471 formats_dict = {}
2472
2473 # We check that all the formats have the format and format_id fields
2474 for i, format in enumerate(formats):
2475 sanitize_string_field(format, 'format_id')
2476 sanitize_numeric_fields(format)
2477 format['url'] = sanitize_url(format['url'])
2478 if not format.get('format_id'):
2479 format['format_id'] = str(i)
2480 else:
2481 # Sanitize format_id from characters used in format selector expression
2482 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2483 format_id = format['format_id']
2484 if format_id not in formats_dict:
2485 formats_dict[format_id] = []
2486 formats_dict[format_id].append(format)
2487
2488 # Make sure all formats have unique format_id
2489 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2490 for format_id, ambiguous_formats in formats_dict.items():
2491 ambigious_id = len(ambiguous_formats) > 1
2492 for i, format in enumerate(ambiguous_formats):
2493 if ambigious_id:
2494 format['format_id'] = '%s-%d' % (format_id, i)
2495 if format.get('ext') is None:
2496 format['ext'] = determine_ext(format['url']).lower()
2497 # Ensure there is no conflict between id and ext in format selection
2498 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2499 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2500 format['format_id'] = 'f%s' % format['format_id']
2501
2502 for i, format in enumerate(formats):
2503 if format.get('format') is None:
2504 format['format'] = '{id} - {res}{note}'.format(
2505 id=format['format_id'],
2506 res=self.format_resolution(format),
2507 note=format_field(format, 'format_note', ' (%s)'),
2508 )
2509 if format.get('protocol') is None:
2510 format['protocol'] = determine_protocol(format)
2511 if format.get('resolution') is None:
2512 format['resolution'] = self.format_resolution(format, default=None)
2513 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2514 format['dynamic_range'] = 'SDR'
2515 if (info_dict.get('duration') and format.get('tbr')
2516 and not format.get('filesize') and not format.get('filesize_approx')):
2517 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2518
2519 # Add HTTP headers, so that external programs can use them from the
2520 # json output
2521 full_format_info = info_dict.copy()
2522 full_format_info.update(format)
2523 format['http_headers'] = self._calc_headers(full_format_info)
2524 # Remove private housekeeping stuff
2525 if '__x_forwarded_for_ip' in info_dict:
2526 del info_dict['__x_forwarded_for_ip']
2527
2528 if self.params.get('check_formats') is True:
2529 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2530
2531 if not formats or formats[0] is not info_dict:
2532 # only set the 'formats' fields if the original info_dict list them
2533 # otherwise we end up with a circular reference, the first (and unique)
2534 # element in the 'formats' field in info_dict is info_dict itself,
2535 # which can't be exported to json
2536 info_dict['formats'] = formats
2537
2538 info_dict, _ = self.pre_process(info_dict)
2539
2540 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2541 return info_dict
2542
2543 self.post_extract(info_dict)
2544 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2545
2546 # The pre-processors may have modified the formats
2547 formats = info_dict.get('formats', [info_dict])
2548
2549 list_only = self.params.get('simulate') is None and (
2550 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2551 interactive_format_selection = not list_only and self.format_selector == '-'
2552 if self.params.get('list_thumbnails'):
2553 self.list_thumbnails(info_dict)
2554 if self.params.get('listsubtitles'):
2555 if 'automatic_captions' in info_dict:
2556 self.list_subtitles(
2557 info_dict['id'], automatic_captions, 'automatic captions')
2558 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2559 if self.params.get('listformats') or interactive_format_selection:
2560 self.list_formats(info_dict)
2561 if list_only:
2562 # Without this printing, -F --print-json will not work
2563 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2564 return info_dict
2565
2566 format_selector = self.format_selector
2567 if format_selector is None:
2568 req_format = self._default_format_spec(info_dict, download=download)
2569 self.write_debug('Default format spec: %s' % req_format)
2570 format_selector = self.build_format_selector(req_format)
2571
2572 while True:
2573 if interactive_format_selection:
2574 req_format = input(
2575 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2576 try:
2577 format_selector = self.build_format_selector(req_format)
2578 except SyntaxError as err:
2579 self.report_error(err, tb=False, is_error=False)
2580 continue
2581
2582 formats_to_download = list(format_selector({
2583 'formats': formats,
2584 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2585 'incomplete_formats': (
2586 # All formats are video-only or
2587 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2588 # all formats are audio-only
2589 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2590 }))
2591 if interactive_format_selection and not formats_to_download:
2592 self.report_error('Requested format is not available', tb=False, is_error=False)
2593 continue
2594 break
2595
2596 if not formats_to_download:
2597 if not self.params.get('ignore_no_formats_error'):
2598 raise ExtractorError(
2599 'Requested format is not available. Use --list-formats for a list of available formats',
2600 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2601 self.report_warning('Requested format is not available')
2602 # Process what we can, even without any available formats.
2603 formats_to_download = [{}]
2604
2605 requested_ranges = self.params.get('download_ranges')
2606 if requested_ranges:
2607 requested_ranges = tuple(requested_ranges(info_dict, self))
2608
2609 best_format, downloaded_formats = formats_to_download[-1], []
2610 if download:
2611 if best_format:
2612 def to_screen(*msg):
2613 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2614
2615 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2616 (f['format_id'] for f in formats_to_download))
2617 if requested_ranges:
2618 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2619 (f'{int(c["start_time"])}-{int(c["end_time"])}' for c in requested_ranges))
2620 max_downloads_reached = False
2621
2622 for fmt, chapter in itertools.product(formats_to_download, requested_ranges or [{}]):
2623 new_info = self._copy_infodict(info_dict)
2624 new_info.update(fmt)
2625 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2626 if chapter or offset:
2627 new_info.update({
2628 'section_start': offset + chapter.get('start_time', 0),
2629 'section_end': offset + min(chapter.get('end_time', duration), duration),
2630 'section_title': chapter.get('title'),
2631 'section_number': chapter.get('index'),
2632 })
2633 downloaded_formats.append(new_info)
2634 try:
2635 self.process_info(new_info)
2636 except MaxDownloadsReached:
2637 max_downloads_reached = True
2638 self._raise_pending_errors(new_info)
2639 # Remove copied info
2640 for key, val in tuple(new_info.items()):
2641 if info_dict.get(key) == val:
2642 new_info.pop(key)
2643 if max_downloads_reached:
2644 break
2645
2646 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2647 assert write_archive.issubset({True, False, 'ignore'})
2648 if True in write_archive and False not in write_archive:
2649 self.record_download_archive(info_dict)
2650
2651 info_dict['requested_downloads'] = downloaded_formats
2652 info_dict = self.run_all_pps('after_video', info_dict)
2653 if max_downloads_reached:
2654 raise MaxDownloadsReached()
2655
2656 # We update the info dict with the selected best quality format (backwards compatibility)
2657 info_dict.update(best_format)
2658 return info_dict
2659
2660 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2661 """Select the requested subtitles and their format"""
2662 available_subs, normal_sub_langs = {}, []
2663 if normal_subtitles and self.params.get('writesubtitles'):
2664 available_subs.update(normal_subtitles)
2665 normal_sub_langs = tuple(normal_subtitles.keys())
2666 if automatic_captions and self.params.get('writeautomaticsub'):
2667 for lang, cap_info in automatic_captions.items():
2668 if lang not in available_subs:
2669 available_subs[lang] = cap_info
2670
2671 if (not self.params.get('writesubtitles') and not
2672 self.params.get('writeautomaticsub') or not
2673 available_subs):
2674 return None
2675
2676 all_sub_langs = tuple(available_subs.keys())
2677 if self.params.get('allsubtitles', False):
2678 requested_langs = all_sub_langs
2679 elif self.params.get('subtitleslangs', False):
2680 # A list is used so that the order of languages will be the same as
2681 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2682 requested_langs = []
2683 for lang_re in self.params.get('subtitleslangs'):
2684 discard = lang_re[0] == '-'
2685 if discard:
2686 lang_re = lang_re[1:]
2687 if lang_re == 'all':
2688 if discard:
2689 requested_langs = []
2690 else:
2691 requested_langs.extend(all_sub_langs)
2692 continue
2693 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2694 if discard:
2695 for lang in current_langs:
2696 while lang in requested_langs:
2697 requested_langs.remove(lang)
2698 else:
2699 requested_langs.extend(current_langs)
2700 requested_langs = orderedSet(requested_langs)
2701 elif normal_sub_langs:
2702 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2703 else:
2704 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2705 if requested_langs:
2706 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2707
2708 formats_query = self.params.get('subtitlesformat', 'best')
2709 formats_preference = formats_query.split('/') if formats_query else []
2710 subs = {}
2711 for lang in requested_langs:
2712 formats = available_subs.get(lang)
2713 if formats is None:
2714 self.report_warning(f'{lang} subtitles not available for {video_id}')
2715 continue
2716 for ext in formats_preference:
2717 if ext == 'best':
2718 f = formats[-1]
2719 break
2720 matches = list(filter(lambda f: f['ext'] == ext, formats))
2721 if matches:
2722 f = matches[-1]
2723 break
2724 else:
2725 f = formats[-1]
2726 self.report_warning(
2727 'No subtitle format found matching "%s" for language %s, '
2728 'using %s' % (formats_query, lang, f['ext']))
2729 subs[lang] = f
2730 return subs
2731
2732 def _forceprint(self, key, info_dict):
2733 if info_dict is None:
2734 return
2735 info_copy = info_dict.copy()
2736 info_copy['formats_table'] = self.render_formats_table(info_dict)
2737 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2738 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2739 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2740
2741 def format_tmpl(tmpl):
2742 mobj = re.match(r'\w+(=?)$', tmpl)
2743 if mobj and mobj.group(1):
2744 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2745 elif mobj:
2746 return f'%({tmpl})s'
2747 return tmpl
2748
2749 for tmpl in self.params['forceprint'].get(key, []):
2750 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2751
2752 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2753 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2754 tmpl = format_tmpl(tmpl)
2755 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2756 if self._ensure_dir_exists(filename):
2757 with open(filename, 'a', encoding='utf-8') as f:
2758 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2759
2760 def __forced_printings(self, info_dict, filename, incomplete):
2761 def print_mandatory(field, actual_field=None):
2762 if actual_field is None:
2763 actual_field = field
2764 if (self.params.get('force%s' % field, False)
2765 and (not incomplete or info_dict.get(actual_field) is not None)):
2766 self.to_stdout(info_dict[actual_field])
2767
2768 def print_optional(field):
2769 if (self.params.get('force%s' % field, False)
2770 and info_dict.get(field) is not None):
2771 self.to_stdout(info_dict[field])
2772
2773 info_dict = info_dict.copy()
2774 if filename is not None:
2775 info_dict['filename'] = filename
2776 if info_dict.get('requested_formats') is not None:
2777 # For RTMP URLs, also include the playpath
2778 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2779 elif info_dict.get('url'):
2780 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2781
2782 if (self.params.get('forcejson')
2783 or self.params['forceprint'].get('video')
2784 or self.params['print_to_file'].get('video')):
2785 self.post_extract(info_dict)
2786 self._forceprint('video', info_dict)
2787
2788 print_mandatory('title')
2789 print_mandatory('id')
2790 print_mandatory('url', 'urls')
2791 print_optional('thumbnail')
2792 print_optional('description')
2793 print_optional('filename')
2794 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2795 self.to_stdout(formatSeconds(info_dict['duration']))
2796 print_mandatory('format')
2797
2798 if self.params.get('forcejson'):
2799 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2800
2801 def dl(self, name, info, subtitle=False, test=False):
2802 if not info.get('url'):
2803 self.raise_no_formats(info, True)
2804
2805 if test:
2806 verbose = self.params.get('verbose')
2807 params = {
2808 'test': True,
2809 'quiet': self.params.get('quiet') or not verbose,
2810 'verbose': verbose,
2811 'noprogress': not verbose,
2812 'nopart': True,
2813 'skip_unavailable_fragments': False,
2814 'keep_fragments': False,
2815 'overwrites': True,
2816 '_no_ytdl_file': True,
2817 }
2818 else:
2819 params = self.params
2820 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2821 if not test:
2822 for ph in self._progress_hooks:
2823 fd.add_progress_hook(ph)
2824 urls = '", "'.join(
2825 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2826 for f in info.get('requested_formats', []) or [info])
2827 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2828
2829 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2830 # But it may contain objects that are not deep-copyable
2831 new_info = self._copy_infodict(info)
2832 if new_info.get('http_headers') is None:
2833 new_info['http_headers'] = self._calc_headers(new_info)
2834 return fd.download(name, new_info, subtitle)
2835
2836 def existing_file(self, filepaths, *, default_overwrite=True):
2837 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2838 if existing_files and not self.params.get('overwrites', default_overwrite):
2839 return existing_files[0]
2840
2841 for file in existing_files:
2842 self.report_file_delete(file)
2843 os.remove(file)
2844 return None
2845
2846 def process_info(self, info_dict):
2847 """Process a single resolved IE result. (Modifies it in-place)"""
2848
2849 assert info_dict.get('_type', 'video') == 'video'
2850 original_infodict = info_dict
2851
2852 if 'format' not in info_dict and 'ext' in info_dict:
2853 info_dict['format'] = info_dict['ext']
2854
2855 # This is mostly just for backward compatibility of process_info
2856 # As a side-effect, this allows for format-specific filters
2857 if self._match_entry(info_dict) is not None:
2858 info_dict['__write_download_archive'] = 'ignore'
2859 return
2860
2861 # Does nothing under normal operation - for backward compatibility of process_info
2862 self.post_extract(info_dict)
2863 self._num_downloads += 1
2864
2865 # info_dict['_filename'] needs to be set for backward compatibility
2866 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2867 temp_filename = self.prepare_filename(info_dict, 'temp')
2868 files_to_move = {}
2869
2870 # Forced printings
2871 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2872
2873 def check_max_downloads():
2874 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
2875 raise MaxDownloadsReached()
2876
2877 if self.params.get('simulate'):
2878 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2879 check_max_downloads()
2880 return
2881
2882 if full_filename is None:
2883 return
2884 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2885 return
2886 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2887 return
2888
2889 if self._write_description('video', info_dict,
2890 self.prepare_filename(info_dict, 'description')) is None:
2891 return
2892
2893 sub_files = self._write_subtitles(info_dict, temp_filename)
2894 if sub_files is None:
2895 return
2896 files_to_move.update(dict(sub_files))
2897
2898 thumb_files = self._write_thumbnails(
2899 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2900 if thumb_files is None:
2901 return
2902 files_to_move.update(dict(thumb_files))
2903
2904 infofn = self.prepare_filename(info_dict, 'infojson')
2905 _infojson_written = self._write_info_json('video', info_dict, infofn)
2906 if _infojson_written:
2907 info_dict['infojson_filename'] = infofn
2908 # For backward compatibility, even though it was a private field
2909 info_dict['__infojson_filename'] = infofn
2910 elif _infojson_written is None:
2911 return
2912
2913 # Note: Annotations are deprecated
2914 annofn = None
2915 if self.params.get('writeannotations', False):
2916 annofn = self.prepare_filename(info_dict, 'annotation')
2917 if annofn:
2918 if not self._ensure_dir_exists(encodeFilename(annofn)):
2919 return
2920 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2921 self.to_screen('[info] Video annotations are already present')
2922 elif not info_dict.get('annotations'):
2923 self.report_warning('There are no annotations to write.')
2924 else:
2925 try:
2926 self.to_screen('[info] Writing video annotations to: ' + annofn)
2927 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2928 annofile.write(info_dict['annotations'])
2929 except (KeyError, TypeError):
2930 self.report_warning('There are no annotations to write.')
2931 except OSError:
2932 self.report_error('Cannot write annotations file: ' + annofn)
2933 return
2934
2935 # Write internet shortcut files
2936 def _write_link_file(link_type):
2937 url = try_get(info_dict['webpage_url'], iri_to_uri)
2938 if not url:
2939 self.report_warning(
2940 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2941 return True
2942 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2943 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2944 return False
2945 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2946 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2947 return True
2948 try:
2949 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2950 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2951 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2952 template_vars = {'url': url}
2953 if link_type == 'desktop':
2954 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2955 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2956 except OSError:
2957 self.report_error(f'Cannot write internet shortcut {linkfn}')
2958 return False
2959 return True
2960
2961 write_links = {
2962 'url': self.params.get('writeurllink'),
2963 'webloc': self.params.get('writewebloclink'),
2964 'desktop': self.params.get('writedesktoplink'),
2965 }
2966 if self.params.get('writelink'):
2967 link_type = ('webloc' if sys.platform == 'darwin'
2968 else 'desktop' if sys.platform.startswith('linux')
2969 else 'url')
2970 write_links[link_type] = True
2971
2972 if any(should_write and not _write_link_file(link_type)
2973 for link_type, should_write in write_links.items()):
2974 return
2975
2976 def replace_info_dict(new_info):
2977 nonlocal info_dict
2978 if new_info == info_dict:
2979 return
2980 info_dict.clear()
2981 info_dict.update(new_info)
2982
2983 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2984 replace_info_dict(new_info)
2985
2986 if self.params.get('skip_download'):
2987 info_dict['filepath'] = temp_filename
2988 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2989 info_dict['__files_to_move'] = files_to_move
2990 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2991 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2992 else:
2993 # Download
2994 info_dict.setdefault('__postprocessors', [])
2995 try:
2996
2997 def existing_video_file(*filepaths):
2998 ext = info_dict.get('ext')
2999 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3000 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3001 default_overwrite=False)
3002 if file:
3003 info_dict['ext'] = os.path.splitext(file)[1][1:]
3004 return file
3005
3006 fd, success = None, True
3007 if info_dict.get('protocol') or info_dict.get('url'):
3008 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3009 if fd is not FFmpegFD and (
3010 info_dict.get('section_start') or info_dict.get('section_end')):
3011 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3012 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3013 self.report_error(f'{msg}. Aborting')
3014 return
3015
3016 if info_dict.get('requested_formats') is not None:
3017
3018 def compatible_formats(formats):
3019 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3020 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3021 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3022 if len(video_formats) > 2 or len(audio_formats) > 2:
3023 return False
3024
3025 # Check extension
3026 exts = {format.get('ext') for format in formats}
3027 COMPATIBLE_EXTS = (
3028 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3029 {'webm'},
3030 )
3031 for ext_sets in COMPATIBLE_EXTS:
3032 if ext_sets.issuperset(exts):
3033 return True
3034 # TODO: Check acodec/vcodec
3035 return False
3036
3037 requested_formats = info_dict['requested_formats']
3038 old_ext = info_dict['ext']
3039 if self.params.get('merge_output_format') is None:
3040 if not compatible_formats(requested_formats):
3041 info_dict['ext'] = 'mkv'
3042 self.report_warning(
3043 'Requested formats are incompatible for merge and will be merged into mkv')
3044 if (info_dict['ext'] == 'webm'
3045 and info_dict.get('thumbnails')
3046 # check with type instead of pp_key, __name__, or isinstance
3047 # since we dont want any custom PPs to trigger this
3048 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3049 info_dict['ext'] = 'mkv'
3050 self.report_warning(
3051 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3052 new_ext = info_dict['ext']
3053
3054 def correct_ext(filename, ext=new_ext):
3055 if filename == '-':
3056 return filename
3057 filename_real_ext = os.path.splitext(filename)[1][1:]
3058 filename_wo_ext = (
3059 os.path.splitext(filename)[0]
3060 if filename_real_ext in (old_ext, new_ext)
3061 else filename)
3062 return f'{filename_wo_ext}.{ext}'
3063
3064 # Ensure filename always has a correct extension for successful merge
3065 full_filename = correct_ext(full_filename)
3066 temp_filename = correct_ext(temp_filename)
3067 dl_filename = existing_video_file(full_filename, temp_filename)
3068 info_dict['__real_download'] = False
3069
3070 merger = FFmpegMergerPP(self)
3071 downloaded = []
3072 if dl_filename is not None:
3073 self.report_file_already_downloaded(dl_filename)
3074 elif fd:
3075 for f in requested_formats if fd != FFmpegFD else []:
3076 f['filepath'] = fname = prepend_extension(
3077 correct_ext(temp_filename, info_dict['ext']),
3078 'f%s' % f['format_id'], info_dict['ext'])
3079 downloaded.append(fname)
3080 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3081 success, real_download = self.dl(temp_filename, info_dict)
3082 info_dict['__real_download'] = real_download
3083 else:
3084 if self.params.get('allow_unplayable_formats'):
3085 self.report_warning(
3086 'You have requested merging of multiple formats '
3087 'while also allowing unplayable formats to be downloaded. '
3088 'The formats won\'t be merged to prevent data corruption.')
3089 elif not merger.available:
3090 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3091 if not self.params.get('ignoreerrors'):
3092 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3093 return
3094 self.report_warning(f'{msg}. The formats won\'t be merged')
3095
3096 if temp_filename == '-':
3097 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3098 else 'but the formats are incompatible for simultaneous download' if merger.available
3099 else 'but ffmpeg is not installed')
3100 self.report_warning(
3101 f'You have requested downloading multiple formats to stdout {reason}. '
3102 'The formats will be streamed one after the other')
3103 fname = temp_filename
3104 for f in requested_formats:
3105 new_info = dict(info_dict)
3106 del new_info['requested_formats']
3107 new_info.update(f)
3108 if temp_filename != '-':
3109 fname = prepend_extension(
3110 correct_ext(temp_filename, new_info['ext']),
3111 'f%s' % f['format_id'], new_info['ext'])
3112 if not self._ensure_dir_exists(fname):
3113 return
3114 f['filepath'] = fname
3115 downloaded.append(fname)
3116 partial_success, real_download = self.dl(fname, new_info)
3117 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3118 success = success and partial_success
3119
3120 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3121 info_dict['__postprocessors'].append(merger)
3122 info_dict['__files_to_merge'] = downloaded
3123 # Even if there were no downloads, it is being merged only now
3124 info_dict['__real_download'] = True
3125 else:
3126 for file in downloaded:
3127 files_to_move[file] = None
3128 else:
3129 # Just a single file
3130 dl_filename = existing_video_file(full_filename, temp_filename)
3131 if dl_filename is None or dl_filename == temp_filename:
3132 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3133 # So we should try to resume the download
3134 success, real_download = self.dl(temp_filename, info_dict)
3135 info_dict['__real_download'] = real_download
3136 else:
3137 self.report_file_already_downloaded(dl_filename)
3138
3139 dl_filename = dl_filename or temp_filename
3140 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3141
3142 except network_exceptions as err:
3143 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3144 return
3145 except OSError as err:
3146 raise UnavailableVideoError(err)
3147 except (ContentTooShortError, ) as err:
3148 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3149 return
3150
3151 self._raise_pending_errors(info_dict)
3152 if success and full_filename != '-':
3153
3154 def fixup():
3155 do_fixup = True
3156 fixup_policy = self.params.get('fixup')
3157 vid = info_dict['id']
3158
3159 if fixup_policy in ('ignore', 'never'):
3160 return
3161 elif fixup_policy == 'warn':
3162 do_fixup = 'warn'
3163 elif fixup_policy != 'force':
3164 assert fixup_policy in ('detect_or_warn', None)
3165 if not info_dict.get('__real_download'):
3166 do_fixup = False
3167
3168 def ffmpeg_fixup(cndn, msg, cls):
3169 if not (do_fixup and cndn):
3170 return
3171 elif do_fixup == 'warn':
3172 self.report_warning(f'{vid}: {msg}')
3173 return
3174 pp = cls(self)
3175 if pp.available:
3176 info_dict['__postprocessors'].append(pp)
3177 else:
3178 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3179
3180 stretched_ratio = info_dict.get('stretched_ratio')
3181 ffmpeg_fixup(
3182 stretched_ratio not in (1, None),
3183 f'Non-uniform pixel ratio {stretched_ratio}',
3184 FFmpegFixupStretchedPP)
3185
3186 ffmpeg_fixup(
3187 (info_dict.get('requested_formats') is None
3188 and info_dict.get('container') == 'm4a_dash'
3189 and info_dict.get('ext') == 'm4a'),
3190 'writing DASH m4a. Only some players support this container',
3191 FFmpegFixupM4aPP)
3192
3193 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3194 downloader = downloader.FD_NAME if downloader else None
3195
3196 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3197 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3198 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3199 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3200 FFmpegFixupM3u8PP)
3201 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3202 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3203
3204 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3205 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3206
3207 fixup()
3208 try:
3209 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3210 except PostProcessingError as err:
3211 self.report_error('Postprocessing: %s' % str(err))
3212 return
3213 try:
3214 for ph in self._post_hooks:
3215 ph(info_dict['filepath'])
3216 except Exception as err:
3217 self.report_error('post hooks: %s' % str(err))
3218 return
3219 info_dict['__write_download_archive'] = True
3220
3221 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3222 if self.params.get('force_write_download_archive'):
3223 info_dict['__write_download_archive'] = True
3224 check_max_downloads()
3225
3226 def __download_wrapper(self, func):
3227 @functools.wraps(func)
3228 def wrapper(*args, **kwargs):
3229 try:
3230 res = func(*args, **kwargs)
3231 except UnavailableVideoError as e:
3232 self.report_error(e)
3233 except DownloadCancelled as e:
3234 self.to_screen(f'[info] {e}')
3235 if not self.params.get('break_per_url'):
3236 raise
3237 else:
3238 if self.params.get('dump_single_json', False):
3239 self.post_extract(res)
3240 self.to_stdout(json.dumps(self.sanitize_info(res)))
3241 return wrapper
3242
3243 def download(self, url_list):
3244 """Download a given list of URLs."""
3245 url_list = variadic(url_list) # Passing a single URL is a common mistake
3246 outtmpl = self.params['outtmpl']['default']
3247 if (len(url_list) > 1
3248 and outtmpl != '-'
3249 and '%' not in outtmpl
3250 and self.params.get('max_downloads') != 1):
3251 raise SameFileError(outtmpl)
3252
3253 for url in url_list:
3254 self.__download_wrapper(self.extract_info)(
3255 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3256
3257 return self._download_retcode
3258
3259 def download_with_info_file(self, info_filename):
3260 with contextlib.closing(fileinput.FileInput(
3261 [info_filename], mode='r',
3262 openhook=fileinput.hook_encoded('utf-8'))) as f:
3263 # FileInput doesn't have a read method, we can't call json.load
3264 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3265 try:
3266 self.__download_wrapper(self.process_ie_result)(info, download=True)
3267 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3268 if not isinstance(e, EntryNotInPlaylist):
3269 self.to_stderr('\r')
3270 webpage_url = info.get('webpage_url')
3271 if webpage_url is not None:
3272 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3273 return self.download([webpage_url])
3274 else:
3275 raise
3276 return self._download_retcode
3277
3278 @staticmethod
3279 def sanitize_info(info_dict, remove_private_keys=False):
3280 ''' Sanitize the infodict for converting to json '''
3281 if info_dict is None:
3282 return info_dict
3283 info_dict.setdefault('epoch', int(time.time()))
3284 info_dict.setdefault('_type', 'video')
3285
3286 if remove_private_keys:
3287 reject = lambda k, v: v is None or k.startswith('__') or k in {
3288 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3289 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3290 }
3291 else:
3292 reject = lambda k, v: False
3293
3294 def filter_fn(obj):
3295 if isinstance(obj, dict):
3296 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3297 elif isinstance(obj, (list, tuple, set, LazyList)):
3298 return list(map(filter_fn, obj))
3299 elif obj is None or isinstance(obj, (str, int, float, bool)):
3300 return obj
3301 else:
3302 return repr(obj)
3303
3304 return filter_fn(info_dict)
3305
3306 @staticmethod
3307 def filter_requested_info(info_dict, actually_filter=True):
3308 ''' Alias of sanitize_info for backward compatibility '''
3309 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3310
3311 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3312 for filename in set(filter(None, files_to_delete)):
3313 if msg:
3314 self.to_screen(msg % filename)
3315 try:
3316 os.remove(filename)
3317 except OSError:
3318 self.report_warning(f'Unable to delete file {filename}')
3319 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3320 del info['__files_to_move'][filename]
3321
3322 @staticmethod
3323 def post_extract(info_dict):
3324 def actual_post_extract(info_dict):
3325 if info_dict.get('_type') in ('playlist', 'multi_video'):
3326 for video_dict in info_dict.get('entries', {}):
3327 actual_post_extract(video_dict or {})
3328 return
3329
3330 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3331 info_dict.update(post_extractor())
3332
3333 actual_post_extract(info_dict or {})
3334
3335 def run_pp(self, pp, infodict):
3336 files_to_delete = []
3337 if '__files_to_move' not in infodict:
3338 infodict['__files_to_move'] = {}
3339 try:
3340 files_to_delete, infodict = pp.run(infodict)
3341 except PostProcessingError as e:
3342 # Must be True and not 'only_download'
3343 if self.params.get('ignoreerrors') is True:
3344 self.report_error(e)
3345 return infodict
3346 raise
3347
3348 if not files_to_delete:
3349 return infodict
3350 if self.params.get('keepvideo', False):
3351 for f in files_to_delete:
3352 infodict['__files_to_move'].setdefault(f, '')
3353 else:
3354 self._delete_downloaded_files(
3355 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3356 return infodict
3357
3358 def run_all_pps(self, key, info, *, additional_pps=None):
3359 self._forceprint(key, info)
3360 for pp in (additional_pps or []) + self._pps[key]:
3361 info = self.run_pp(pp, info)
3362 return info
3363
3364 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3365 info = dict(ie_info)
3366 info['__files_to_move'] = files_to_move or {}
3367 try:
3368 info = self.run_all_pps(key, info)
3369 except PostProcessingError as err:
3370 msg = f'Preprocessing: {err}'
3371 info.setdefault('__pending_error', msg)
3372 self.report_error(msg, is_error=False)
3373 return info, info.pop('__files_to_move', None)
3374
3375 def post_process(self, filename, info, files_to_move=None):
3376 """Run all the postprocessors on the given file."""
3377 info['filepath'] = filename
3378 info['__files_to_move'] = files_to_move or {}
3379 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3380 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3381 del info['__files_to_move']
3382 return self.run_all_pps('after_move', info)
3383
3384 def _make_archive_id(self, info_dict):
3385 video_id = info_dict.get('id')
3386 if not video_id:
3387 return
3388 # Future-proof against any change in case
3389 # and backwards compatibility with prior versions
3390 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3391 if extractor is None:
3392 url = str_or_none(info_dict.get('url'))
3393 if not url:
3394 return
3395 # Try to find matching extractor for the URL and take its ie_key
3396 for ie_key, ie in self._ies.items():
3397 if ie.suitable(url):
3398 extractor = ie_key
3399 break
3400 else:
3401 return
3402 return f'{extractor.lower()} {video_id}'
3403
3404 def in_download_archive(self, info_dict):
3405 fn = self.params.get('download_archive')
3406 if fn is None:
3407 return False
3408
3409 vid_id = self._make_archive_id(info_dict)
3410 if not vid_id:
3411 return False # Incomplete video information
3412
3413 return vid_id in self.archive
3414
3415 def record_download_archive(self, info_dict):
3416 fn = self.params.get('download_archive')
3417 if fn is None:
3418 return
3419 vid_id = self._make_archive_id(info_dict)
3420 assert vid_id
3421 self.write_debug(f'Adding to archive: {vid_id}')
3422 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3423 archive_file.write(vid_id + '\n')
3424 self.archive.add(vid_id)
3425
3426 @staticmethod
3427 def format_resolution(format, default='unknown'):
3428 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3429 return 'audio only'
3430 if format.get('resolution') is not None:
3431 return format['resolution']
3432 if format.get('width') and format.get('height'):
3433 return '%dx%d' % (format['width'], format['height'])
3434 elif format.get('height'):
3435 return '%sp' % format['height']
3436 elif format.get('width'):
3437 return '%dx?' % format['width']
3438 return default
3439
3440 def _list_format_headers(self, *headers):
3441 if self.params.get('listformats_table', True) is not False:
3442 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3443 return headers
3444
3445 def _format_note(self, fdict):
3446 res = ''
3447 if fdict.get('ext') in ['f4f', 'f4m']:
3448 res += '(unsupported)'
3449 if fdict.get('language'):
3450 if res:
3451 res += ' '
3452 res += '[%s]' % fdict['language']
3453 if fdict.get('format_note') is not None:
3454 if res:
3455 res += ' '
3456 res += fdict['format_note']
3457 if fdict.get('tbr') is not None:
3458 if res:
3459 res += ', '
3460 res += '%4dk' % fdict['tbr']
3461 if fdict.get('container') is not None:
3462 if res:
3463 res += ', '
3464 res += '%s container' % fdict['container']
3465 if (fdict.get('vcodec') is not None
3466 and fdict.get('vcodec') != 'none'):
3467 if res:
3468 res += ', '
3469 res += fdict['vcodec']
3470 if fdict.get('vbr') is not None:
3471 res += '@'
3472 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3473 res += 'video@'
3474 if fdict.get('vbr') is not None:
3475 res += '%4dk' % fdict['vbr']
3476 if fdict.get('fps') is not None:
3477 if res:
3478 res += ', '
3479 res += '%sfps' % fdict['fps']
3480 if fdict.get('acodec') is not None:
3481 if res:
3482 res += ', '
3483 if fdict['acodec'] == 'none':
3484 res += 'video only'
3485 else:
3486 res += '%-5s' % fdict['acodec']
3487 elif fdict.get('abr') is not None:
3488 if res:
3489 res += ', '
3490 res += 'audio'
3491 if fdict.get('abr') is not None:
3492 res += '@%3dk' % fdict['abr']
3493 if fdict.get('asr') is not None:
3494 res += ' (%5dHz)' % fdict['asr']
3495 if fdict.get('filesize') is not None:
3496 if res:
3497 res += ', '
3498 res += format_bytes(fdict['filesize'])
3499 elif fdict.get('filesize_approx') is not None:
3500 if res:
3501 res += ', '
3502 res += '~' + format_bytes(fdict['filesize_approx'])
3503 return res
3504
3505 def render_formats_table(self, info_dict):
3506 if not info_dict.get('formats') and not info_dict.get('url'):
3507 return None
3508
3509 formats = info_dict.get('formats', [info_dict])
3510 if not self.params.get('listformats_table', True) is not False:
3511 table = [
3512 [
3513 format_field(f, 'format_id'),
3514 format_field(f, 'ext'),
3515 self.format_resolution(f),
3516 self._format_note(f)
3517 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3518 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3519
3520 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3521 table = [
3522 [
3523 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3524 format_field(f, 'ext'),
3525 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3526 format_field(f, 'fps', '\t%d'),
3527 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3528 delim,
3529 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3530 format_field(f, 'tbr', '\t%dk'),
3531 shorten_protocol_name(f.get('protocol', '')),
3532 delim,
3533 format_field(f, 'vcodec', default='unknown').replace(
3534 'none', 'images' if f.get('acodec') == 'none'
3535 else self._format_out('audio only', self.Styles.SUPPRESS)),
3536 format_field(f, 'vbr', '\t%dk'),
3537 format_field(f, 'acodec', default='unknown').replace(
3538 'none', '' if f.get('vcodec') == 'none'
3539 else self._format_out('video only', self.Styles.SUPPRESS)),
3540 format_field(f, 'abr', '\t%dk'),
3541 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3542 join_nonempty(
3543 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3544 format_field(f, 'language', '[%s]'),
3545 join_nonempty(format_field(f, 'format_note'),
3546 format_field(f, 'container', ignore=(None, f.get('ext'))),
3547 delim=', '),
3548 delim=' '),
3549 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3550 header_line = self._list_format_headers(
3551 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3552 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3553
3554 return render_table(
3555 header_line, table, hide_empty=True,
3556 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3557
3558 def render_thumbnails_table(self, info_dict):
3559 thumbnails = list(info_dict.get('thumbnails') or [])
3560 if not thumbnails:
3561 return None
3562 return render_table(
3563 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3564 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3565
3566 def render_subtitles_table(self, video_id, subtitles):
3567 def _row(lang, formats):
3568 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3569 if len(set(names)) == 1:
3570 names = [] if names[0] == 'unknown' else names[:1]
3571 return [lang, ', '.join(names), ', '.join(exts)]
3572
3573 if not subtitles:
3574 return None
3575 return render_table(
3576 self._list_format_headers('Language', 'Name', 'Formats'),
3577 [_row(lang, formats) for lang, formats in subtitles.items()],
3578 hide_empty=True)
3579
3580 def __list_table(self, video_id, name, func, *args):
3581 table = func(*args)
3582 if not table:
3583 self.to_screen(f'{video_id} has no {name}')
3584 return
3585 self.to_screen(f'[info] Available {name} for {video_id}:')
3586 self.to_stdout(table)
3587
3588 def list_formats(self, info_dict):
3589 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3590
3591 def list_thumbnails(self, info_dict):
3592 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3593
3594 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3595 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3596
3597 def urlopen(self, req):
3598 """ Start an HTTP download """
3599 if isinstance(req, str):
3600 req = sanitized_Request(req)
3601 return self._opener.open(req, timeout=self._socket_timeout)
3602
3603 def print_debug_header(self):
3604 if not self.params.get('verbose'):
3605 return
3606
3607 # These imports can be slow. So import them only as needed
3608 from .extractor.extractors import _LAZY_LOADER
3609 from .extractor.extractors import _PLUGIN_CLASSES as plugin_extractors
3610
3611 def get_encoding(stream):
3612 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3613 if not supports_terminal_sequences(stream):
3614 from .utils import WINDOWS_VT_MODE # Must be imported locally
3615 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3616 return ret
3617
3618 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3619 locale.getpreferredencoding(),
3620 sys.getfilesystemencoding(),
3621 self.get_encoding(),
3622 ', '.join(
3623 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3624 if stream is not None and key != 'console')
3625 )
3626
3627 logger = self.params.get('logger')
3628 if logger:
3629 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3630 write_debug(encoding_str)
3631 else:
3632 write_string(f'[debug] {encoding_str}\n', encoding=None)
3633 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3634
3635 source = detect_variant()
3636 write_debug(join_nonempty(
3637 'yt-dlp version', __version__,
3638 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3639 '' if source == 'unknown' else f'({source})',
3640 delim=' '))
3641 if not _LAZY_LOADER:
3642 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3643 write_debug('Lazy loading extractors is forcibly disabled')
3644 else:
3645 write_debug('Lazy loading extractors is disabled')
3646 if plugin_extractors or plugin_postprocessors:
3647 write_debug('Plugins: %s' % [
3648 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3649 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3650 if self.params['compat_opts']:
3651 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3652
3653 if source == 'source':
3654 try:
3655 stdout, _, _ = Popen.run(
3656 ['git', 'rev-parse', '--short', 'HEAD'],
3657 text=True, cwd=os.path.dirname(os.path.abspath(__file__)),
3658 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
3659 if re.fullmatch('[0-9a-f]+', stdout.strip()):
3660 write_debug(f'Git HEAD: {stdout.strip()}')
3661 except Exception:
3662 with contextlib.suppress(Exception):
3663 sys.exc_clear()
3664
3665 write_debug(system_identifier())
3666
3667 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3668 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3669 if ffmpeg_features:
3670 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3671
3672 exe_versions['rtmpdump'] = rtmpdump_version()
3673 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3674 exe_str = ', '.join(
3675 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3676 ) or 'none'
3677 write_debug('exe versions: %s' % exe_str)
3678
3679 from .compat.compat_utils import get_package_info
3680 from .dependencies import available_dependencies
3681
3682 write_debug('Optional libraries: %s' % (', '.join(sorted({
3683 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3684 })) or 'none'))
3685
3686 self._setup_opener()
3687 proxy_map = {}
3688 for handler in self._opener.handlers:
3689 if hasattr(handler, 'proxies'):
3690 proxy_map.update(handler.proxies)
3691 write_debug(f'Proxy map: {proxy_map}')
3692
3693 # Not implemented
3694 if False and self.params.get('call_home'):
3695 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3696 write_debug('Public IP address: %s' % ipaddr)
3697 latest_version = self.urlopen(
3698 'https://yt-dl.org/latest/version').read().decode()
3699 if version_tuple(latest_version) > version_tuple(__version__):
3700 self.report_warning(
3701 'You are using an outdated version (newest version: %s)! '
3702 'See https://yt-dl.org/update if you need help updating.' %
3703 latest_version)
3704
3705 def _setup_opener(self):
3706 if hasattr(self, '_opener'):
3707 return
3708 timeout_val = self.params.get('socket_timeout')
3709 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3710
3711 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3712 opts_cookiefile = self.params.get('cookiefile')
3713 opts_proxy = self.params.get('proxy')
3714
3715 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3716
3717 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3718 if opts_proxy is not None:
3719 if opts_proxy == '':
3720 proxies = {}
3721 else:
3722 proxies = {'http': opts_proxy, 'https': opts_proxy}
3723 else:
3724 proxies = urllib.request.getproxies()
3725 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3726 if 'http' in proxies and 'https' not in proxies:
3727 proxies['https'] = proxies['http']
3728 proxy_handler = PerRequestProxyHandler(proxies)
3729
3730 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3731 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3732 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3733 redirect_handler = YoutubeDLRedirectHandler()
3734 data_handler = urllib.request.DataHandler()
3735
3736 # When passing our own FileHandler instance, build_opener won't add the
3737 # default FileHandler and allows us to disable the file protocol, which
3738 # can be used for malicious purposes (see
3739 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3740 file_handler = urllib.request.FileHandler()
3741
3742 def file_open(*args, **kwargs):
3743 raise urllib.error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3744 file_handler.file_open = file_open
3745
3746 opener = urllib.request.build_opener(
3747 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3748
3749 # Delete the default user-agent header, which would otherwise apply in
3750 # cases where our custom HTTP handler doesn't come into play
3751 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3752 opener.addheaders = []
3753 self._opener = opener
3754
3755 def encode(self, s):
3756 if isinstance(s, bytes):
3757 return s # Already encoded
3758
3759 try:
3760 return s.encode(self.get_encoding())
3761 except UnicodeEncodeError as err:
3762 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3763 raise
3764
3765 def get_encoding(self):
3766 encoding = self.params.get('encoding')
3767 if encoding is None:
3768 encoding = preferredencoding()
3769 return encoding
3770
3771 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3772 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3773 if overwrite is None:
3774 overwrite = self.params.get('overwrites', True)
3775 if not self.params.get('writeinfojson'):
3776 return False
3777 elif not infofn:
3778 self.write_debug(f'Skipping writing {label} infojson')
3779 return False
3780 elif not self._ensure_dir_exists(infofn):
3781 return None
3782 elif not overwrite and os.path.exists(infofn):
3783 self.to_screen(f'[info] {label.title()} metadata is already present')
3784 return 'exists'
3785
3786 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3787 try:
3788 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3789 return True
3790 except OSError:
3791 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3792 return None
3793
3794 def _write_description(self, label, ie_result, descfn):
3795 ''' Write description and returns True = written, False = skip, None = error '''
3796 if not self.params.get('writedescription'):
3797 return False
3798 elif not descfn:
3799 self.write_debug(f'Skipping writing {label} description')
3800 return False
3801 elif not self._ensure_dir_exists(descfn):
3802 return None
3803 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3804 self.to_screen(f'[info] {label.title()} description is already present')
3805 elif ie_result.get('description') is None:
3806 self.report_warning(f'There\'s no {label} description to write')
3807 return False
3808 else:
3809 try:
3810 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3811 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3812 descfile.write(ie_result['description'])
3813 except OSError:
3814 self.report_error(f'Cannot write {label} description file {descfn}')
3815 return None
3816 return True
3817
3818 def _write_subtitles(self, info_dict, filename):
3819 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3820 ret = []
3821 subtitles = info_dict.get('requested_subtitles')
3822 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3823 # subtitles download errors are already managed as troubles in relevant IE
3824 # that way it will silently go on when used with unsupporting IE
3825 return ret
3826
3827 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3828 if not sub_filename_base:
3829 self.to_screen('[info] Skipping writing video subtitles')
3830 return ret
3831 for sub_lang, sub_info in subtitles.items():
3832 sub_format = sub_info['ext']
3833 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3834 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3835 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3836 if existing_sub:
3837 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3838 sub_info['filepath'] = existing_sub
3839 ret.append((existing_sub, sub_filename_final))
3840 continue
3841
3842 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3843 if sub_info.get('data') is not None:
3844 try:
3845 # Use newline='' to prevent conversion of newline characters
3846 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3847 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3848 subfile.write(sub_info['data'])
3849 sub_info['filepath'] = sub_filename
3850 ret.append((sub_filename, sub_filename_final))
3851 continue
3852 except OSError:
3853 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3854 return None
3855
3856 try:
3857 sub_copy = sub_info.copy()
3858 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3859 self.dl(sub_filename, sub_copy, subtitle=True)
3860 sub_info['filepath'] = sub_filename
3861 ret.append((sub_filename, sub_filename_final))
3862 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3863 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3864 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3865 if not self.params.get('ignoreerrors'):
3866 self.report_error(msg)
3867 raise DownloadError(msg)
3868 self.report_warning(msg)
3869 return ret
3870
3871 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3872 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3873 write_all = self.params.get('write_all_thumbnails', False)
3874 thumbnails, ret = [], []
3875 if write_all or self.params.get('writethumbnail', False):
3876 thumbnails = info_dict.get('thumbnails') or []
3877 multiple = write_all and len(thumbnails) > 1
3878
3879 if thumb_filename_base is None:
3880 thumb_filename_base = filename
3881 if thumbnails and not thumb_filename_base:
3882 self.write_debug(f'Skipping writing {label} thumbnail')
3883 return ret
3884
3885 for idx, t in list(enumerate(thumbnails))[::-1]:
3886 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3887 thumb_display_id = f'{label} thumbnail {t["id"]}'
3888 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3889 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3890
3891 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3892 if existing_thumb:
3893 self.to_screen('[info] %s is already present' % (
3894 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3895 t['filepath'] = existing_thumb
3896 ret.append((existing_thumb, thumb_filename_final))
3897 else:
3898 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3899 try:
3900 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3901 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3902 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3903 shutil.copyfileobj(uf, thumbf)
3904 ret.append((thumb_filename, thumb_filename_final))
3905 t['filepath'] = thumb_filename
3906 except network_exceptions as err:
3907 thumbnails.pop(idx)
3908 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3909 if ret and not write_all:
3910 break
3911 return ret