]> jfr.im git - yt-dlp.git/blame_incremental - yt_dlp/YoutubeDL.py
Add slicing notation to `--playlist-items`
[yt-dlp.git] / yt_dlp / YoutubeDL.py
... / ...
CommitLineData
1#!/usr/bin/env python3
2import collections
3import contextlib
4import datetime
5import errno
6import fileinput
7import functools
8import io
9import itertools
10import json
11import locale
12import operator
13import os
14import platform
15import random
16import re
17import shutil
18import subprocess
19import sys
20import tempfile
21import time
22import tokenize
23import traceback
24import unicodedata
25import urllib.request
26from string import ascii_letters
27
28from .cache import Cache
29from .compat import (
30 HAS_LEGACY as compat_has_legacy,
31 compat_get_terminal_size,
32 compat_os_name,
33 compat_shlex_quote,
34 compat_str,
35 compat_urllib_error,
36 compat_urllib_request,
37)
38from .cookies import load_cookies
39from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
40from .downloader.rtmp import rtmpdump_version
41from .extractor import gen_extractor_classes, get_info_extractor
42from .extractor.openload import PhantomJSwrapper
43from .minicurses import format_text
44from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
45from .postprocessor import (
46 EmbedThumbnailPP,
47 FFmpegFixupDuplicateMoovPP,
48 FFmpegFixupDurationPP,
49 FFmpegFixupM3u8PP,
50 FFmpegFixupM4aPP,
51 FFmpegFixupStretchedPP,
52 FFmpegFixupTimestampPP,
53 FFmpegMergerPP,
54 FFmpegPostProcessor,
55 MoveFilesAfterDownloadPP,
56 get_postprocessor,
57)
58from .update import detect_variant
59from .utils import (
60 DEFAULT_OUTTMPL,
61 LINK_TEMPLATES,
62 NO_DEFAULT,
63 NUMBER_RE,
64 OUTTMPL_TYPES,
65 POSTPROCESS_WHEN,
66 STR_FORMAT_RE_TMPL,
67 STR_FORMAT_TYPES,
68 ContentTooShortError,
69 DateRange,
70 DownloadCancelled,
71 DownloadError,
72 EntryNotInPlaylist,
73 ExistingVideoReached,
74 ExtractorError,
75 GeoRestrictedError,
76 HEADRequest,
77 ISO3166Utils,
78 LazyList,
79 MaxDownloadsReached,
80 Namespace,
81 PagedList,
82 PerRequestProxyHandler,
83 PlaylistEntries,
84 Popen,
85 PostProcessingError,
86 ReExtractInfo,
87 RejectedVideoReached,
88 SameFileError,
89 UnavailableVideoError,
90 YoutubeDLCookieProcessor,
91 YoutubeDLHandler,
92 YoutubeDLRedirectHandler,
93 age_restricted,
94 args_to_str,
95 date_from_str,
96 determine_ext,
97 determine_protocol,
98 encode_compat_str,
99 encodeFilename,
100 error_to_compat_str,
101 expand_path,
102 filter_dict,
103 float_or_none,
104 format_bytes,
105 format_decimal_suffix,
106 format_field,
107 formatSeconds,
108 get_domain,
109 int_or_none,
110 iri_to_uri,
111 join_nonempty,
112 locked_file,
113 make_dir,
114 make_HTTPS_handler,
115 merge_headers,
116 network_exceptions,
117 number_of_digits,
118 orderedSet,
119 parse_filesize,
120 platform_name,
121 preferredencoding,
122 prepend_extension,
123 register_socks_protocols,
124 remove_terminal_sequences,
125 render_table,
126 replace_extension,
127 sanitize_filename,
128 sanitize_path,
129 sanitize_url,
130 sanitized_Request,
131 std_headers,
132 str_or_none,
133 strftime_or_none,
134 subtitles_filename,
135 supports_terminal_sequences,
136 timetuple_from_msec,
137 to_high_limit_path,
138 traverse_obj,
139 try_get,
140 url_basename,
141 variadic,
142 version_tuple,
143 windows_enable_vt_mode,
144 write_json_file,
145 write_string,
146)
147from .version import RELEASE_GIT_HEAD, __version__
148
149if compat_os_name == 'nt':
150 import ctypes
151
152
153class YoutubeDL:
154 """YoutubeDL class.
155
156 YoutubeDL objects are the ones responsible of downloading the
157 actual video file and writing it to disk if the user has requested
158 it, among some other tasks. In most cases there should be one per
159 program. As, given a video URL, the downloader doesn't know how to
160 extract all the needed information, task that InfoExtractors do, it
161 has to pass the URL to one of them.
162
163 For this, YoutubeDL objects have a method that allows
164 InfoExtractors to be registered in a given order. When it is passed
165 a URL, the YoutubeDL object handles it to the first InfoExtractor it
166 finds that reports being able to handle it. The InfoExtractor extracts
167 all the information about the video or videos the URL refers to, and
168 YoutubeDL process the extracted information, possibly using a File
169 Downloader to download the video.
170
171 YoutubeDL objects accept a lot of parameters. In order not to saturate
172 the object constructor with arguments, it receives a dictionary of
173 options instead. These options are available through the params
174 attribute for the InfoExtractors to use. The YoutubeDL also
175 registers itself as the downloader in charge for the InfoExtractors
176 that are added to it, so this is a "mutual registration".
177
178 Available options:
179
180 username: Username for authentication purposes.
181 password: Password for authentication purposes.
182 videopassword: Password for accessing a video.
183 ap_mso: Adobe Pass multiple-system operator identifier.
184 ap_username: Multiple-system operator account username.
185 ap_password: Multiple-system operator account password.
186 usenetrc: Use netrc for authentication instead.
187 verbose: Print additional info to stdout.
188 quiet: Do not print messages to stdout.
189 no_warnings: Do not print out anything for warnings.
190 forceprint: A dict with keys WHEN mapped to a list of templates to
191 print to stdout. The allowed keys are video or any of the
192 items in utils.POSTPROCESS_WHEN.
193 For compatibility, a single list is also accepted
194 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
195 a list of tuples with (template, filename)
196 forcejson: Force printing info_dict as JSON.
197 dump_single_json: Force printing the info_dict of the whole playlist
198 (or video) as a single JSON line.
199 force_write_download_archive: Force writing download archive regardless
200 of 'skip_download' or 'simulate'.
201 simulate: Do not download the video files. If unset (or None),
202 simulate only if listsubtitles, listformats or list_thumbnails is used
203 format: Video format code. see "FORMAT SELECTION" for more details.
204 You can also pass a function. The function takes 'ctx' as
205 argument and returns the formats to download.
206 See "build_format_selector" for an implementation
207 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
208 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
209 extracting metadata even if the video is not actually
210 available for download (experimental)
211 format_sort: A list of fields by which to sort the video formats.
212 See "Sorting Formats" for more details.
213 format_sort_force: Force the given format_sort. see "Sorting Formats"
214 for more details.
215 prefer_free_formats: Whether to prefer video formats with free containers
216 over non-free ones of same quality.
217 allow_multiple_video_streams: Allow multiple video streams to be merged
218 into a single file
219 allow_multiple_audio_streams: Allow multiple audio streams to be merged
220 into a single file
221 check_formats Whether to test if the formats are downloadable.
222 Can be True (check all), False (check none),
223 'selected' (check selected formats),
224 or None (check only if requested by extractor)
225 paths: Dictionary of output paths. The allowed keys are 'home'
226 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
227 outtmpl: Dictionary of templates for output names. Allowed keys
228 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
229 For compatibility with youtube-dl, a single string can also be used
230 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
231 restrictfilenames: Do not allow "&" and spaces in file names
232 trim_file_name: Limit length of filename (extension excluded)
233 windowsfilenames: Force the filenames to be windows compatible
234 ignoreerrors: Do not stop on download/postprocessing errors.
235 Can be 'only_download' to ignore only download errors.
236 Default is 'only_download' for CLI, but False for API
237 skip_playlist_after_errors: Number of allowed failures until the rest of
238 the playlist is skipped
239 force_generic_extractor: Force downloader to use the generic extractor
240 overwrites: Overwrite all video and metadata files if True,
241 overwrite only non-video files if None
242 and don't overwrite any file if False
243 For compatibility with youtube-dl,
244 "nooverwrites" may also be used instead
245 playliststart: Playlist item to start at.
246 playlistend: Playlist item to end at.
247 playlist_items: Specific indices of playlist to download.
248 playlistreverse: Download playlist items in reverse order.
249 playlistrandom: Download playlist items in random order.
250 matchtitle: Download only matching titles.
251 rejecttitle: Reject downloads for matching titles.
252 logger: Log messages to a logging.Logger instance.
253 logtostderr: Log messages to stderr instead of stdout.
254 consoletitle: Display progress in console window's titlebar.
255 writedescription: Write the video description to a .description file
256 writeinfojson: Write the video description to a .info.json file
257 clean_infojson: Remove private fields from the infojson
258 getcomments: Extract video comments. This will not be written to disk
259 unless writeinfojson is also given
260 writeannotations: Write the video annotations to a .annotations.xml file
261 writethumbnail: Write the thumbnail image to a file
262 allow_playlist_files: Whether to write playlists' description, infojson etc
263 also to disk when using the 'write*' options
264 write_all_thumbnails: Write all thumbnail formats to files
265 writelink: Write an internet shortcut file, depending on the
266 current platform (.url/.webloc/.desktop)
267 writeurllink: Write a Windows internet shortcut file (.url)
268 writewebloclink: Write a macOS internet shortcut file (.webloc)
269 writedesktoplink: Write a Linux internet shortcut file (.desktop)
270 writesubtitles: Write the video subtitles to a file
271 writeautomaticsub: Write the automatically generated subtitles to a file
272 listsubtitles: Lists all available subtitles for the video
273 subtitlesformat: The format code for subtitles
274 subtitleslangs: List of languages of the subtitles to download (can be regex).
275 The list may contain "all" to refer to all the available
276 subtitles. The language can be prefixed with a "-" to
277 exclude it from the requested languages. Eg: ['all', '-live_chat']
278 keepvideo: Keep the video file after post-processing
279 daterange: A DateRange object, download only if the upload_date is in the range.
280 skip_download: Skip the actual download of the video file
281 cachedir: Location of the cache files in the filesystem.
282 False to disable filesystem cache.
283 noplaylist: Download single video instead of a playlist if in doubt.
284 age_limit: An integer representing the user's age in years.
285 Unsuitable videos for the given age are skipped.
286 min_views: An integer representing the minimum view count the video
287 must have in order to not be skipped.
288 Videos without view count information are always
289 downloaded. None for no limit.
290 max_views: An integer representing the maximum view count.
291 Videos that are more popular than that are not
292 downloaded.
293 Videos without view count information are always
294 downloaded. None for no limit.
295 download_archive: File name of a file where all downloads are recorded.
296 Videos already present in the file are not downloaded
297 again.
298 break_on_existing: Stop the download process after attempting to download a
299 file that is in the archive.
300 break_on_reject: Stop the download process when encountering a video that
301 has been filtered out.
302 break_per_url: Whether break_on_reject and break_on_existing
303 should act on each input URL as opposed to for the entire queue
304 cookiefile: File name or text stream from where cookies should be read and dumped to
305 cookiesfrombrowser: A tuple containing the name of the browser, the profile
306 name/pathfrom where cookies are loaded, and the name of the
307 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
308 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
309 support RFC 5746 secure renegotiation
310 nocheckcertificate: Do not verify SSL certificates
311 client_certificate: Path to client certificate file in PEM format. May include the private key
312 client_certificate_key: Path to private key file for client certificate
313 client_certificate_password: Password for client certificate private key, if encrypted.
314 If not provided and the key is encrypted, yt-dlp will ask interactively
315 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
316 At the moment, this is only supported by YouTube.
317 http_headers: A dictionary of custom headers to be used for all requests
318 proxy: URL of the proxy server to use
319 geo_verification_proxy: URL of the proxy to use for IP address verification
320 on geo-restricted sites.
321 socket_timeout: Time to wait for unresponsive hosts, in seconds
322 bidi_workaround: Work around buggy terminals without bidirectional text
323 support, using fridibi
324 debug_printtraffic:Print out sent and received HTTP traffic
325 default_search: Prepend this string if an input url is not valid.
326 'auto' for elaborate guessing
327 encoding: Use this encoding instead of the system-specified.
328 extract_flat: Do not resolve URLs, return the immediate result.
329 Pass in 'in_playlist' to only show this behavior for
330 playlist items.
331 wait_for_video: If given, wait for scheduled streams to become available.
332 The value should be a tuple containing the range
333 (min_secs, max_secs) to wait between retries
334 postprocessors: A list of dictionaries, each with an entry
335 * key: The name of the postprocessor. See
336 yt_dlp/postprocessor/__init__.py for a list.
337 * when: When to run the postprocessor. Allowed values are
338 the entries of utils.POSTPROCESS_WHEN
339 Assumed to be 'post_process' if not given
340 progress_hooks: A list of functions that get called on download
341 progress, with a dictionary with the entries
342 * status: One of "downloading", "error", or "finished".
343 Check this first and ignore unknown values.
344 * info_dict: The extracted info_dict
345
346 If status is one of "downloading", or "finished", the
347 following properties may also be present:
348 * filename: The final filename (always present)
349 * tmpfilename: The filename we're currently writing to
350 * downloaded_bytes: Bytes on disk
351 * total_bytes: Size of the whole file, None if unknown
352 * total_bytes_estimate: Guess of the eventual file size,
353 None if unavailable.
354 * elapsed: The number of seconds since download started.
355 * eta: The estimated time in seconds, None if unknown
356 * speed: The download speed in bytes/second, None if
357 unknown
358 * fragment_index: The counter of the currently
359 downloaded video fragment.
360 * fragment_count: The number of fragments (= individual
361 files that will be merged)
362
363 Progress hooks are guaranteed to be called at least once
364 (with status "finished") if the download is successful.
365 postprocessor_hooks: A list of functions that get called on postprocessing
366 progress, with a dictionary with the entries
367 * status: One of "started", "processing", or "finished".
368 Check this first and ignore unknown values.
369 * postprocessor: Name of the postprocessor
370 * info_dict: The extracted info_dict
371
372 Progress hooks are guaranteed to be called at least twice
373 (with status "started" and "finished") if the processing is successful.
374 merge_output_format: Extension to use when merging formats.
375 final_ext: Expected final extension; used to detect when the file was
376 already downloaded and converted
377 fixup: Automatically correct known faults of the file.
378 One of:
379 - "never": do nothing
380 - "warn": only emit a warning
381 - "detect_or_warn": check whether we can do anything
382 about it, warn otherwise (default)
383 source_address: Client-side IP address to bind to.
384 sleep_interval_requests: Number of seconds to sleep between requests
385 during extraction
386 sleep_interval: Number of seconds to sleep before each download when
387 used alone or a lower bound of a range for randomized
388 sleep before each download (minimum possible number
389 of seconds to sleep) when used along with
390 max_sleep_interval.
391 max_sleep_interval:Upper bound of a range for randomized sleep before each
392 download (maximum possible number of seconds to sleep).
393 Must only be used along with sleep_interval.
394 Actual sleep time will be a random float from range
395 [sleep_interval; max_sleep_interval].
396 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
397 listformats: Print an overview of available video formats and exit.
398 list_thumbnails: Print a table of all thumbnails and exit.
399 match_filter: A function that gets called for every video with the signature
400 (info_dict, *, incomplete: bool) -> Optional[str]
401 For backward compatibility with youtube-dl, the signature
402 (info_dict) -> Optional[str] is also allowed.
403 - If it returns a message, the video is ignored.
404 - If it returns None, the video is downloaded.
405 - If it returns utils.NO_DEFAULT, the user is interactively
406 asked whether to download the video.
407 match_filter_func in utils.py is one example for this.
408 no_color: Do not emit color codes in output.
409 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
410 HTTP header
411 geo_bypass_country:
412 Two-letter ISO 3166-2 country code that will be used for
413 explicit geographic restriction bypassing via faking
414 X-Forwarded-For HTTP header
415 geo_bypass_ip_block:
416 IP range in CIDR notation that will be used similarly to
417 geo_bypass_country
418 external_downloader: A dictionary of protocol keys and the executable of the
419 external downloader to use for it. The allowed protocols
420 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
421 Set the value to 'native' to use the native downloader
422 compat_opts: Compatibility options. See "Differences in default behavior".
423 The following options do not work when used through the API:
424 filename, abort-on-error, multistreams, no-live-chat, format-sort
425 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
426 Refer __init__.py for their implementation
427 progress_template: Dictionary of templates for progress outputs.
428 Allowed keys are 'download', 'postprocess',
429 'download-title' (console title) and 'postprocess-title'.
430 The template is mapped on a dictionary with keys 'progress' and 'info'
431 retry_sleep_functions: Dictionary of functions that takes the number of attempts
432 as argument and returns the time to sleep in seconds.
433 Allowed keys are 'http', 'fragment', 'file_access'
434 download_ranges: A function that gets called for every video with the signature
435 (info_dict, *, ydl) -> Iterable[Section].
436 Only the returned sections will be downloaded. Each Section contains:
437 * start_time: Start time of the section in seconds
438 * end_time: End time of the section in seconds
439 * title: Section title (Optional)
440 * index: Section number (Optional)
441
442 The following parameters are not used by YoutubeDL itself, they are used by
443 the downloader (see yt_dlp/downloader/common.py):
444 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
445 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
446 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
447 external_downloader_args, concurrent_fragment_downloads.
448
449 The following options are used by the post processors:
450 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
451 to the binary or its containing directory.
452 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
453 and a list of additional command-line arguments for the
454 postprocessor/executable. The dict can also have "PP+EXE" keys
455 which are used when the given exe is used by the given PP.
456 Use 'default' as the name for arguments to passed to all PP
457 For compatibility with youtube-dl, a single list of args
458 can also be used
459
460 The following options are used by the extractors:
461 extractor_retries: Number of times to retry for known errors
462 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
463 hls_split_discontinuity: Split HLS playlists to different formats at
464 discontinuities such as ad breaks (default: False)
465 extractor_args: A dictionary of arguments to be passed to the extractors.
466 See "EXTRACTOR ARGUMENTS" for details.
467 Eg: {'youtube': {'skip': ['dash', 'hls']}}
468 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
469
470 The following options are deprecated and may be removed in the future:
471
472 forceurl: - Use forceprint
473 Force printing final URL.
474 forcetitle: - Use forceprint
475 Force printing title.
476 forceid: - Use forceprint
477 Force printing ID.
478 forcethumbnail: - Use forceprint
479 Force printing thumbnail URL.
480 forcedescription: - Use forceprint
481 Force printing description.
482 forcefilename: - Use forceprint
483 Force printing final filename.
484 forceduration: - Use forceprint
485 Force printing duration.
486 allsubtitles: - Use subtitleslangs = ['all']
487 Downloads all the subtitles of the video
488 (requires writesubtitles or writeautomaticsub)
489 include_ads: - Doesn't work
490 Download ads as well
491 call_home: - Not implemented
492 Boolean, true iff we are allowed to contact the
493 yt-dlp servers for debugging.
494 post_hooks: - Register a custom postprocessor
495 A list of functions that get called as the final step
496 for each video file, after all postprocessors have been
497 called. The filename will be passed as the only argument.
498 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
499 Use the native HLS downloader instead of ffmpeg/avconv
500 if True, otherwise use ffmpeg/avconv if False, otherwise
501 use downloader suggested by extractor if None.
502 prefer_ffmpeg: - avconv support is deprecated
503 If False, use avconv instead of ffmpeg if both are available,
504 otherwise prefer ffmpeg.
505 youtube_include_dash_manifest: - Use extractor_args
506 If True (default), DASH manifests and related
507 data will be downloaded and processed by extractor.
508 You can reduce network I/O by disabling it if you don't
509 care about DASH. (only for youtube)
510 youtube_include_hls_manifest: - Use extractor_args
511 If True (default), HLS manifests and related
512 data will be downloaded and processed by extractor.
513 You can reduce network I/O by disabling it if you don't
514 care about HLS. (only for youtube)
515 """
516
517 _NUMERIC_FIELDS = {
518 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
519 'timestamp', 'release_timestamp',
520 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
521 'average_rating', 'comment_count', 'age_limit',
522 'start_time', 'end_time',
523 'chapter_number', 'season_number', 'episode_number',
524 'track_number', 'disc_number', 'release_year',
525 }
526
527 _format_fields = {
528 # NB: Keep in sync with the docstring of extractor/common.py
529 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
530 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
531 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
532 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
533 'preference', 'language', 'language_preference', 'quality', 'source_preference',
534 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
535 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
536 }
537 _format_selection_exts = {
538 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
539 'video': {'mp4', 'flv', 'webm', '3gp'},
540 'storyboards': {'mhtml'},
541 }
542
543 def __init__(self, params=None, auto_init=True):
544 """Create a FileDownloader object with the given options.
545 @param auto_init Whether to load the default extractors and print header (if verbose).
546 Set to 'no_verbose_header' to not print the header
547 """
548 if params is None:
549 params = {}
550 self.params = params
551 self._ies = {}
552 self._ies_instances = {}
553 self._pps = {k: [] for k in POSTPROCESS_WHEN}
554 self._printed_messages = set()
555 self._first_webpage_request = True
556 self._post_hooks = []
557 self._progress_hooks = []
558 self._postprocessor_hooks = []
559 self._download_retcode = 0
560 self._num_downloads = 0
561 self._num_videos = 0
562 self._playlist_level = 0
563 self._playlist_urls = set()
564 self.cache = Cache(self)
565
566 windows_enable_vt_mode()
567 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
568 self._out_files = Namespace(
569 out=stdout,
570 error=sys.stderr,
571 screen=sys.stderr if self.params.get('quiet') else stdout,
572 console=None if compat_os_name == 'nt' else next(
573 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
574 )
575 self._allow_colors = Namespace(**{
576 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
577 for type_, stream in self._out_files.items_ if type_ != 'console'
578 })
579
580 if sys.version_info < (3, 6):
581 self.report_warning(
582 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
583
584 if self.params.get('allow_unplayable_formats'):
585 self.report_warning(
586 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
587 'This is a developer option intended for debugging. \n'
588 ' If you experience any issues while using this option, '
589 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
590
591 def check_deprecated(param, option, suggestion):
592 if self.params.get(param) is not None:
593 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
594 return True
595 return False
596
597 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
598 if self.params.get('geo_verification_proxy') is None:
599 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
600
601 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
602 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
603 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
604
605 for msg in self.params.get('_warnings', []):
606 self.report_warning(msg)
607 for msg in self.params.get('_deprecation_warnings', []):
608 self.deprecation_warning(msg)
609
610 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
611 if not compat_has_legacy:
612 self.params['compat_opts'].add('no-compat-legacy')
613 if 'list-formats' in self.params['compat_opts']:
614 self.params['listformats_table'] = False
615
616 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
617 # nooverwrites was unnecessarily changed to overwrites
618 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
619 # This ensures compatibility with both keys
620 self.params['overwrites'] = not self.params['nooverwrites']
621 elif self.params.get('overwrites') is None:
622 self.params.pop('overwrites', None)
623 else:
624 self.params['nooverwrites'] = not self.params['overwrites']
625
626 self.params.setdefault('forceprint', {})
627 self.params.setdefault('print_to_file', {})
628
629 # Compatibility with older syntax
630 if not isinstance(params['forceprint'], dict):
631 self.params['forceprint'] = {'video': params['forceprint']}
632
633 if self.params.get('bidi_workaround', False):
634 try:
635 import pty
636 master, slave = pty.openpty()
637 width = compat_get_terminal_size().columns
638 width_args = [] if width is None else ['-w', str(width)]
639 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
640 try:
641 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
642 except OSError:
643 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
644 self._output_channel = os.fdopen(master, 'rb')
645 except OSError as ose:
646 if ose.errno == errno.ENOENT:
647 self.report_warning(
648 'Could not find fribidi executable, ignoring --bidi-workaround. '
649 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
650 else:
651 raise
652
653 if auto_init:
654 if auto_init != 'no_verbose_header':
655 self.print_debug_header()
656 self.add_default_info_extractors()
657
658 if (sys.platform != 'win32'
659 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
660 and not self.params.get('restrictfilenames', False)):
661 # Unicode filesystem API will throw errors (#1474, #13027)
662 self.report_warning(
663 'Assuming --restrict-filenames since file system encoding '
664 'cannot encode all characters. '
665 'Set the LC_ALL environment variable to fix this.')
666 self.params['restrictfilenames'] = True
667
668 self.outtmpl_dict = self.parse_outtmpl()
669
670 # Creating format selector here allows us to catch syntax errors before the extraction
671 self.format_selector = (
672 self.params.get('format') if self.params.get('format') in (None, '-')
673 else self.params['format'] if callable(self.params['format'])
674 else self.build_format_selector(self.params['format']))
675
676 # Set http_headers defaults according to std_headers
677 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
678
679 hooks = {
680 'post_hooks': self.add_post_hook,
681 'progress_hooks': self.add_progress_hook,
682 'postprocessor_hooks': self.add_postprocessor_hook,
683 }
684 for opt, fn in hooks.items():
685 for ph in self.params.get(opt, []):
686 fn(ph)
687
688 for pp_def_raw in self.params.get('postprocessors', []):
689 pp_def = dict(pp_def_raw)
690 when = pp_def.pop('when', 'post_process')
691 self.add_post_processor(
692 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
693 when=when)
694
695 self._setup_opener()
696 register_socks_protocols()
697
698 def preload_download_archive(fn):
699 """Preload the archive, if any is specified"""
700 if fn is None:
701 return False
702 self.write_debug(f'Loading archive file {fn!r}')
703 try:
704 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
705 for line in archive_file:
706 self.archive.add(line.strip())
707 except OSError as ioe:
708 if ioe.errno != errno.ENOENT:
709 raise
710 return False
711 return True
712
713 self.archive = set()
714 preload_download_archive(self.params.get('download_archive'))
715
716 def warn_if_short_id(self, argv):
717 # short YouTube ID starting with dash?
718 idxs = [
719 i for i, a in enumerate(argv)
720 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
721 if idxs:
722 correct_argv = (
723 ['yt-dlp']
724 + [a for i, a in enumerate(argv) if i not in idxs]
725 + ['--'] + [argv[i] for i in idxs]
726 )
727 self.report_warning(
728 'Long argument string detected. '
729 'Use -- to separate parameters and URLs, like this:\n%s' %
730 args_to_str(correct_argv))
731
732 def add_info_extractor(self, ie):
733 """Add an InfoExtractor object to the end of the list."""
734 ie_key = ie.ie_key()
735 self._ies[ie_key] = ie
736 if not isinstance(ie, type):
737 self._ies_instances[ie_key] = ie
738 ie.set_downloader(self)
739
740 def _get_info_extractor_class(self, ie_key):
741 ie = self._ies.get(ie_key)
742 if ie is None:
743 ie = get_info_extractor(ie_key)
744 self.add_info_extractor(ie)
745 return ie
746
747 def get_info_extractor(self, ie_key):
748 """
749 Get an instance of an IE with name ie_key, it will try to get one from
750 the _ies list, if there's no instance it will create a new one and add
751 it to the extractor list.
752 """
753 ie = self._ies_instances.get(ie_key)
754 if ie is None:
755 ie = get_info_extractor(ie_key)()
756 self.add_info_extractor(ie)
757 return ie
758
759 def add_default_info_extractors(self):
760 """
761 Add the InfoExtractors returned by gen_extractors to the end of the list
762 """
763 for ie in gen_extractor_classes():
764 self.add_info_extractor(ie)
765
766 def add_post_processor(self, pp, when='post_process'):
767 """Add a PostProcessor object to the end of the chain."""
768 self._pps[when].append(pp)
769 pp.set_downloader(self)
770
771 def add_post_hook(self, ph):
772 """Add the post hook"""
773 self._post_hooks.append(ph)
774
775 def add_progress_hook(self, ph):
776 """Add the download progress hook"""
777 self._progress_hooks.append(ph)
778
779 def add_postprocessor_hook(self, ph):
780 """Add the postprocessing progress hook"""
781 self._postprocessor_hooks.append(ph)
782 for pps in self._pps.values():
783 for pp in pps:
784 pp.add_progress_hook(ph)
785
786 def _bidi_workaround(self, message):
787 if not hasattr(self, '_output_channel'):
788 return message
789
790 assert hasattr(self, '_output_process')
791 assert isinstance(message, compat_str)
792 line_count = message.count('\n') + 1
793 self._output_process.stdin.write((message + '\n').encode())
794 self._output_process.stdin.flush()
795 res = ''.join(self._output_channel.readline().decode()
796 for _ in range(line_count))
797 return res[:-len('\n')]
798
799 def _write_string(self, message, out=None, only_once=False):
800 if only_once:
801 if message in self._printed_messages:
802 return
803 self._printed_messages.add(message)
804 write_string(message, out=out, encoding=self.params.get('encoding'))
805
806 def to_stdout(self, message, skip_eol=False, quiet=None):
807 """Print message to stdout"""
808 if quiet is not None:
809 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
810 if skip_eol is not False:
811 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. Use "YoutubeDL.to_screen" instead')
812 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
813
814 def to_screen(self, message, skip_eol=False, quiet=None):
815 """Print message to screen if not in quiet mode"""
816 if self.params.get('logger'):
817 self.params['logger'].debug(message)
818 return
819 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
820 return
821 self._write_string(
822 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
823 self._out_files.screen)
824
825 def to_stderr(self, message, only_once=False):
826 """Print message to stderr"""
827 assert isinstance(message, compat_str)
828 if self.params.get('logger'):
829 self.params['logger'].error(message)
830 else:
831 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
832
833 def _send_console_code(self, code):
834 if compat_os_name == 'nt' or not self._out_files.console:
835 return
836 self._write_string(code, self._out_files.console)
837
838 def to_console_title(self, message):
839 if not self.params.get('consoletitle', False):
840 return
841 message = remove_terminal_sequences(message)
842 if compat_os_name == 'nt':
843 if ctypes.windll.kernel32.GetConsoleWindow():
844 # c_wchar_p() might not be necessary if `message` is
845 # already of type unicode()
846 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
847 else:
848 self._send_console_code(f'\033]0;{message}\007')
849
850 def save_console_title(self):
851 if not self.params.get('consoletitle') or self.params.get('simulate'):
852 return
853 self._send_console_code('\033[22;0t') # Save the title on stack
854
855 def restore_console_title(self):
856 if not self.params.get('consoletitle') or self.params.get('simulate'):
857 return
858 self._send_console_code('\033[23;0t') # Restore the title from stack
859
860 def __enter__(self):
861 self.save_console_title()
862 return self
863
864 def __exit__(self, *args):
865 self.restore_console_title()
866
867 if self.params.get('cookiefile') is not None:
868 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
869
870 def trouble(self, message=None, tb=None, is_error=True):
871 """Determine action to take when a download problem appears.
872
873 Depending on if the downloader has been configured to ignore
874 download errors or not, this method may throw an exception or
875 not when errors are found, after printing the message.
876
877 @param tb If given, is additional traceback information
878 @param is_error Whether to raise error according to ignorerrors
879 """
880 if message is not None:
881 self.to_stderr(message)
882 if self.params.get('verbose'):
883 if tb is None:
884 if sys.exc_info()[0]: # if .trouble has been called from an except block
885 tb = ''
886 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
887 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
888 tb += encode_compat_str(traceback.format_exc())
889 else:
890 tb_data = traceback.format_list(traceback.extract_stack())
891 tb = ''.join(tb_data)
892 if tb:
893 self.to_stderr(tb)
894 if not is_error:
895 return
896 if not self.params.get('ignoreerrors'):
897 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
898 exc_info = sys.exc_info()[1].exc_info
899 else:
900 exc_info = sys.exc_info()
901 raise DownloadError(message, exc_info)
902 self._download_retcode = 1
903
904 Styles = Namespace(
905 HEADERS='yellow',
906 EMPHASIS='light blue',
907 FILENAME='green',
908 ID='green',
909 DELIM='blue',
910 ERROR='red',
911 WARNING='yellow',
912 SUPPRESS='light black',
913 )
914
915 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
916 text = str(text)
917 if test_encoding:
918 original_text = text
919 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
920 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
921 text = text.encode(encoding, 'ignore').decode(encoding)
922 if fallback is not None and text != original_text:
923 text = fallback
924 return format_text(text, f) if allow_colors else text if fallback is None else fallback
925
926 def _format_out(self, *args, **kwargs):
927 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
928
929 def _format_screen(self, *args, **kwargs):
930 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
931
932 def _format_err(self, *args, **kwargs):
933 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
934
935 def report_warning(self, message, only_once=False):
936 '''
937 Print the message to stderr, it will be prefixed with 'WARNING:'
938 If stderr is a tty file the 'WARNING:' will be colored
939 '''
940 if self.params.get('logger') is not None:
941 self.params['logger'].warning(message)
942 else:
943 if self.params.get('no_warnings'):
944 return
945 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
946
947 def deprecation_warning(self, message):
948 if self.params.get('logger') is not None:
949 self.params['logger'].warning(f'DeprecationWarning: {message}')
950 else:
951 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
952
953 def report_error(self, message, *args, **kwargs):
954 '''
955 Do the same as trouble, but prefixes the message with 'ERROR:', colored
956 in red if stderr is a tty file.
957 '''
958 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
959
960 def write_debug(self, message, only_once=False):
961 '''Log debug message or Print message to stderr'''
962 if not self.params.get('verbose', False):
963 return
964 message = f'[debug] {message}'
965 if self.params.get('logger'):
966 self.params['logger'].debug(message)
967 else:
968 self.to_stderr(message, only_once)
969
970 def report_file_already_downloaded(self, file_name):
971 """Report file has already been fully downloaded."""
972 try:
973 self.to_screen('[download] %s has already been downloaded' % file_name)
974 except UnicodeEncodeError:
975 self.to_screen('[download] The file has already been downloaded')
976
977 def report_file_delete(self, file_name):
978 """Report that existing file will be deleted."""
979 try:
980 self.to_screen('Deleting existing file %s' % file_name)
981 except UnicodeEncodeError:
982 self.to_screen('Deleting existing file')
983
984 def raise_no_formats(self, info, forced=False, *, msg=None):
985 has_drm = info.get('_has_drm')
986 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
987 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
988 if forced or not ignored:
989 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
990 expected=has_drm or ignored or expected)
991 else:
992 self.report_warning(msg)
993
994 def parse_outtmpl(self):
995 outtmpl_dict = self.params.get('outtmpl', {})
996 if not isinstance(outtmpl_dict, dict):
997 outtmpl_dict = {'default': outtmpl_dict}
998 # Remove spaces in the default template
999 if self.params.get('restrictfilenames'):
1000 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1001 else:
1002 sanitize = lambda x: x
1003 outtmpl_dict.update({
1004 k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items()
1005 if outtmpl_dict.get(k) is None})
1006 for _, val in outtmpl_dict.items():
1007 if isinstance(val, bytes):
1008 self.report_warning('Parameter outtmpl is bytes, but should be a unicode string')
1009 return outtmpl_dict
1010
1011 def get_output_path(self, dir_type='', filename=None):
1012 paths = self.params.get('paths', {})
1013 assert isinstance(paths, dict)
1014 path = os.path.join(
1015 expand_path(paths.get('home', '').strip()),
1016 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1017 filename or '')
1018 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1019
1020 @staticmethod
1021 def _outtmpl_expandpath(outtmpl):
1022 # expand_path translates '%%' into '%' and '$$' into '$'
1023 # correspondingly that is not what we want since we need to keep
1024 # '%%' intact for template dict substitution step. Working around
1025 # with boundary-alike separator hack.
1026 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1027 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1028
1029 # outtmpl should be expand_path'ed before template dict substitution
1030 # because meta fields may contain env variables we don't want to
1031 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1032 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1033 return expand_path(outtmpl).replace(sep, '')
1034
1035 @staticmethod
1036 def escape_outtmpl(outtmpl):
1037 ''' Escape any remaining strings like %s, %abc% etc. '''
1038 return re.sub(
1039 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1040 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1041 outtmpl)
1042
1043 @classmethod
1044 def validate_outtmpl(cls, outtmpl):
1045 ''' @return None or Exception object '''
1046 outtmpl = re.sub(
1047 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1048 lambda mobj: f'{mobj.group(0)[:-1]}s',
1049 cls._outtmpl_expandpath(outtmpl))
1050 try:
1051 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1052 return None
1053 except ValueError as err:
1054 return err
1055
1056 @staticmethod
1057 def _copy_infodict(info_dict):
1058 info_dict = dict(info_dict)
1059 info_dict.pop('__postprocessors', None)
1060 info_dict.pop('__pending_error', None)
1061 return info_dict
1062
1063 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1064 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1065 @param sanitize Whether to sanitize the output as a filename.
1066 For backward compatibility, a function can also be passed
1067 """
1068
1069 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1070
1071 info_dict = self._copy_infodict(info_dict)
1072 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1073 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1074 if info_dict.get('duration', None) is not None
1075 else None)
1076 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1077 info_dict['video_autonumber'] = self._num_videos
1078 if info_dict.get('resolution') is None:
1079 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1080
1081 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1082 # of %(field)s to %(field)0Nd for backward compatibility
1083 field_size_compat_map = {
1084 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1085 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1086 'autonumber': self.params.get('autonumber_size') or 5,
1087 }
1088
1089 TMPL_DICT = {}
1090 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1091 MATH_FUNCTIONS = {
1092 '+': float.__add__,
1093 '-': float.__sub__,
1094 }
1095 # Field is of the form key1.key2...
1096 # where keys (except first) can be string, int or slice
1097 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1098 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1099 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1100 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1101 (?P<negate>-)?
1102 (?P<fields>{FIELD_RE})
1103 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1104 (?:>(?P<strf_format>.+?))?
1105 (?P<remaining>
1106 (?P<alternate>(?<!\\),[^|&)]+)?
1107 (?:&(?P<replacement>.*?))?
1108 (?:\|(?P<default>.*?))?
1109 )$''')
1110
1111 def _traverse_infodict(k):
1112 k = k.split('.')
1113 if k[0] == '':
1114 k.pop(0)
1115 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1116
1117 def get_value(mdict):
1118 # Object traversal
1119 value = _traverse_infodict(mdict['fields'])
1120 # Negative
1121 if mdict['negate']:
1122 value = float_or_none(value)
1123 if value is not None:
1124 value *= -1
1125 # Do maths
1126 offset_key = mdict['maths']
1127 if offset_key:
1128 value = float_or_none(value)
1129 operator = None
1130 while offset_key:
1131 item = re.match(
1132 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1133 offset_key).group(0)
1134 offset_key = offset_key[len(item):]
1135 if operator is None:
1136 operator = MATH_FUNCTIONS[item]
1137 continue
1138 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1139 offset = float_or_none(item)
1140 if offset is None:
1141 offset = float_or_none(_traverse_infodict(item))
1142 try:
1143 value = operator(value, multiplier * offset)
1144 except (TypeError, ZeroDivisionError):
1145 return None
1146 operator = None
1147 # Datetime formatting
1148 if mdict['strf_format']:
1149 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1150
1151 return value
1152
1153 na = self.params.get('outtmpl_na_placeholder', 'NA')
1154
1155 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1156 return sanitize_filename(str(value), restricted=restricted, is_id=(
1157 bool(re.search(r'(^|[_.])id(\.|$)', key))
1158 if 'filename-sanitization' in self.params['compat_opts']
1159 else NO_DEFAULT))
1160
1161 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1162 sanitize = bool(sanitize)
1163
1164 def _dumpjson_default(obj):
1165 if isinstance(obj, (set, LazyList)):
1166 return list(obj)
1167 return repr(obj)
1168
1169 def create_key(outer_mobj):
1170 if not outer_mobj.group('has_key'):
1171 return outer_mobj.group(0)
1172 key = outer_mobj.group('key')
1173 mobj = re.match(INTERNAL_FORMAT_RE, key)
1174 initial_field = mobj.group('fields') if mobj else ''
1175 value, replacement, default = None, None, na
1176 while mobj:
1177 mobj = mobj.groupdict()
1178 default = mobj['default'] if mobj['default'] is not None else default
1179 value = get_value(mobj)
1180 replacement = mobj['replacement']
1181 if value is None and mobj['alternate']:
1182 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1183 else:
1184 break
1185
1186 fmt = outer_mobj.group('format')
1187 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1188 fmt = f'0{field_size_compat_map[key]:d}d'
1189
1190 value = default if value is None else value if replacement is None else replacement
1191
1192 flags = outer_mobj.group('conversion') or ''
1193 str_fmt = f'{fmt[:-1]}s'
1194 if fmt[-1] == 'l': # list
1195 delim = '\n' if '#' in flags else ', '
1196 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1197 elif fmt[-1] == 'j': # json
1198 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1199 elif fmt[-1] == 'q': # quoted
1200 value = map(str, variadic(value) if '#' in flags else [value])
1201 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1202 elif fmt[-1] == 'B': # bytes
1203 value = f'%{str_fmt}'.encode() % str(value).encode()
1204 value, fmt = value.decode('utf-8', 'ignore'), 's'
1205 elif fmt[-1] == 'U': # unicode normalized
1206 value, fmt = unicodedata.normalize(
1207 # "+" = compatibility equivalence, "#" = NFD
1208 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1209 value), str_fmt
1210 elif fmt[-1] == 'D': # decimal suffix
1211 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1212 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1213 factor=1024 if '#' in flags else 1000)
1214 elif fmt[-1] == 'S': # filename sanitization
1215 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1216 elif fmt[-1] == 'c':
1217 if value:
1218 value = str(value)[0]
1219 else:
1220 fmt = str_fmt
1221 elif fmt[-1] not in 'rs': # numeric
1222 value = float_or_none(value)
1223 if value is None:
1224 value, fmt = default, 's'
1225
1226 if sanitize:
1227 if fmt[-1] == 'r':
1228 # If value is an object, sanitize might convert it to a string
1229 # So we convert it to repr first
1230 value, fmt = repr(value), str_fmt
1231 if fmt[-1] in 'csr':
1232 value = sanitizer(initial_field, value)
1233
1234 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1235 TMPL_DICT[key] = value
1236 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1237
1238 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1239
1240 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1241 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1242 return self.escape_outtmpl(outtmpl) % info_dict
1243
1244 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1245 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1246 if outtmpl is None:
1247 outtmpl = self.outtmpl_dict.get(tmpl_type or 'default', self.outtmpl_dict['default'])
1248 try:
1249 outtmpl = self._outtmpl_expandpath(outtmpl)
1250 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1251 if not filename:
1252 return None
1253
1254 if tmpl_type in ('', 'temp'):
1255 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1256 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1257 filename = replace_extension(filename, ext, final_ext)
1258 elif tmpl_type:
1259 force_ext = OUTTMPL_TYPES[tmpl_type]
1260 if force_ext:
1261 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1262
1263 # https://github.com/blackjack4494/youtube-dlc/issues/85
1264 trim_file_name = self.params.get('trim_file_name', False)
1265 if trim_file_name:
1266 no_ext, *ext = filename.rsplit('.', 2)
1267 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1268
1269 return filename
1270 except ValueError as err:
1271 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1272 return None
1273
1274 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1275 """Generate the output filename"""
1276 if outtmpl:
1277 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1278 dir_type = None
1279 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1280 if not filename and dir_type not in ('', 'temp'):
1281 return ''
1282
1283 if warn:
1284 if not self.params.get('paths'):
1285 pass
1286 elif filename == '-':
1287 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1288 elif os.path.isabs(filename):
1289 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1290 if filename == '-' or not filename:
1291 return filename
1292
1293 return self.get_output_path(dir_type, filename)
1294
1295 def _match_entry(self, info_dict, incomplete=False, silent=False):
1296 """ Returns None if the file should be downloaded """
1297
1298 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1299
1300 def check_filter():
1301 if 'title' in info_dict:
1302 # This can happen when we're just evaluating the playlist
1303 title = info_dict['title']
1304 matchtitle = self.params.get('matchtitle', False)
1305 if matchtitle:
1306 if not re.search(matchtitle, title, re.IGNORECASE):
1307 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1308 rejecttitle = self.params.get('rejecttitle', False)
1309 if rejecttitle:
1310 if re.search(rejecttitle, title, re.IGNORECASE):
1311 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1312 date = info_dict.get('upload_date')
1313 if date is not None:
1314 dateRange = self.params.get('daterange', DateRange())
1315 if date not in dateRange:
1316 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1317 view_count = info_dict.get('view_count')
1318 if view_count is not None:
1319 min_views = self.params.get('min_views')
1320 if min_views is not None and view_count < min_views:
1321 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1322 max_views = self.params.get('max_views')
1323 if max_views is not None and view_count > max_views:
1324 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1325 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1326 return 'Skipping "%s" because it is age restricted' % video_title
1327
1328 match_filter = self.params.get('match_filter')
1329 if match_filter is not None:
1330 try:
1331 ret = match_filter(info_dict, incomplete=incomplete)
1332 except TypeError:
1333 # For backward compatibility
1334 ret = None if incomplete else match_filter(info_dict)
1335 if ret is NO_DEFAULT:
1336 while True:
1337 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1338 reply = input(self._format_screen(
1339 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1340 if reply in {'y', ''}:
1341 return None
1342 elif reply == 'n':
1343 return f'Skipping {video_title}'
1344 elif ret is not None:
1345 return ret
1346 return None
1347
1348 if self.in_download_archive(info_dict):
1349 reason = '%s has already been recorded in the archive' % video_title
1350 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1351 else:
1352 reason = check_filter()
1353 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1354 if reason is not None:
1355 if not silent:
1356 self.to_screen('[download] ' + reason)
1357 if self.params.get(break_opt, False):
1358 raise break_err()
1359 return reason
1360
1361 @staticmethod
1362 def add_extra_info(info_dict, extra_info):
1363 '''Set the keys from extra_info in info dict if they are missing'''
1364 for key, value in extra_info.items():
1365 info_dict.setdefault(key, value)
1366
1367 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1368 process=True, force_generic_extractor=False):
1369 """
1370 Return a list with a dictionary for each video extracted.
1371
1372 Arguments:
1373 url -- URL to extract
1374
1375 Keyword arguments:
1376 download -- whether to download videos during extraction
1377 ie_key -- extractor key hint
1378 extra_info -- dictionary containing the extra values to add to each result
1379 process -- whether to resolve all unresolved references (URLs, playlist items),
1380 must be True for download to work.
1381 force_generic_extractor -- force using the generic extractor
1382 """
1383
1384 if extra_info is None:
1385 extra_info = {}
1386
1387 if not ie_key and force_generic_extractor:
1388 ie_key = 'Generic'
1389
1390 if ie_key:
1391 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1392 else:
1393 ies = self._ies
1394
1395 for ie_key, ie in ies.items():
1396 if not ie.suitable(url):
1397 continue
1398
1399 if not ie.working():
1400 self.report_warning('The program functionality for this site has been marked as broken, '
1401 'and will probably not work.')
1402
1403 temp_id = ie.get_temp_id(url)
1404 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1405 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1406 if self.params.get('break_on_existing', False):
1407 raise ExistingVideoReached()
1408 break
1409 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1410 else:
1411 self.report_error('no suitable InfoExtractor for URL %s' % url)
1412
1413 def _handle_extraction_exceptions(func):
1414 @functools.wraps(func)
1415 def wrapper(self, *args, **kwargs):
1416 while True:
1417 try:
1418 return func(self, *args, **kwargs)
1419 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1420 raise
1421 except ReExtractInfo as e:
1422 if e.expected:
1423 self.to_screen(f'{e}; Re-extracting data')
1424 else:
1425 self.to_stderr('\r')
1426 self.report_warning(f'{e}; Re-extracting data')
1427 continue
1428 except GeoRestrictedError as e:
1429 msg = e.msg
1430 if e.countries:
1431 msg += '\nThis video is available in %s.' % ', '.join(
1432 map(ISO3166Utils.short2full, e.countries))
1433 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1434 self.report_error(msg)
1435 except ExtractorError as e: # An error we somewhat expected
1436 self.report_error(str(e), e.format_traceback())
1437 except Exception as e:
1438 if self.params.get('ignoreerrors'):
1439 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1440 else:
1441 raise
1442 break
1443 return wrapper
1444
1445 def _wait_for_video(self, ie_result):
1446 if (not self.params.get('wait_for_video')
1447 or ie_result.get('_type', 'video') != 'video'
1448 or ie_result.get('formats') or ie_result.get('url')):
1449 return
1450
1451 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1452 last_msg = ''
1453
1454 def progress(msg):
1455 nonlocal last_msg
1456 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1457 last_msg = msg
1458
1459 min_wait, max_wait = self.params.get('wait_for_video')
1460 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1461 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1462 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1463 self.report_warning('Release time of video is not known')
1464 elif (diff or 0) <= 0:
1465 self.report_warning('Video should already be available according to extracted info')
1466 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1467 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1468
1469 wait_till = time.time() + diff
1470 try:
1471 while True:
1472 diff = wait_till - time.time()
1473 if diff <= 0:
1474 progress('')
1475 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1476 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1477 time.sleep(1)
1478 except KeyboardInterrupt:
1479 progress('')
1480 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1481 except BaseException as e:
1482 if not isinstance(e, ReExtractInfo):
1483 self.to_screen('')
1484 raise
1485
1486 @_handle_extraction_exceptions
1487 def __extract_info(self, url, ie, download, extra_info, process):
1488 ie_result = ie.extract(url)
1489 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1490 return
1491 if isinstance(ie_result, list):
1492 # Backwards compatibility: old IE result format
1493 ie_result = {
1494 '_type': 'compat_list',
1495 'entries': ie_result,
1496 }
1497 if extra_info.get('original_url'):
1498 ie_result.setdefault('original_url', extra_info['original_url'])
1499 self.add_default_extra_info(ie_result, ie, url)
1500 if process:
1501 self._wait_for_video(ie_result)
1502 return self.process_ie_result(ie_result, download, extra_info)
1503 else:
1504 return ie_result
1505
1506 def add_default_extra_info(self, ie_result, ie, url):
1507 if url is not None:
1508 self.add_extra_info(ie_result, {
1509 'webpage_url': url,
1510 'original_url': url,
1511 })
1512 webpage_url = ie_result.get('webpage_url')
1513 if webpage_url:
1514 self.add_extra_info(ie_result, {
1515 'webpage_url_basename': url_basename(webpage_url),
1516 'webpage_url_domain': get_domain(webpage_url),
1517 })
1518 if ie is not None:
1519 self.add_extra_info(ie_result, {
1520 'extractor': ie.IE_NAME,
1521 'extractor_key': ie.ie_key(),
1522 })
1523
1524 def process_ie_result(self, ie_result, download=True, extra_info=None):
1525 """
1526 Take the result of the ie(may be modified) and resolve all unresolved
1527 references (URLs, playlist items).
1528
1529 It will also download the videos if 'download'.
1530 Returns the resolved ie_result.
1531 """
1532 if extra_info is None:
1533 extra_info = {}
1534 result_type = ie_result.get('_type', 'video')
1535
1536 if result_type in ('url', 'url_transparent'):
1537 ie_result['url'] = sanitize_url(ie_result['url'])
1538 if ie_result.get('original_url'):
1539 extra_info.setdefault('original_url', ie_result['original_url'])
1540
1541 extract_flat = self.params.get('extract_flat', False)
1542 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1543 or extract_flat is True):
1544 info_copy = ie_result.copy()
1545 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1546 if ie and not ie_result.get('id'):
1547 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1548 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1549 self.add_extra_info(info_copy, extra_info)
1550 info_copy, _ = self.pre_process(info_copy)
1551 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1552 self._raise_pending_errors(info_copy)
1553 if self.params.get('force_write_download_archive', False):
1554 self.record_download_archive(info_copy)
1555 return ie_result
1556
1557 if result_type == 'video':
1558 self.add_extra_info(ie_result, extra_info)
1559 ie_result = self.process_video_result(ie_result, download=download)
1560 self._raise_pending_errors(ie_result)
1561 additional_urls = (ie_result or {}).get('additional_urls')
1562 if additional_urls:
1563 # TODO: Improve MetadataParserPP to allow setting a list
1564 if isinstance(additional_urls, compat_str):
1565 additional_urls = [additional_urls]
1566 self.to_screen(
1567 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1568 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1569 ie_result['additional_entries'] = [
1570 self.extract_info(
1571 url, download, extra_info=extra_info,
1572 force_generic_extractor=self.params.get('force_generic_extractor'))
1573 for url in additional_urls
1574 ]
1575 return ie_result
1576 elif result_type == 'url':
1577 # We have to add extra_info to the results because it may be
1578 # contained in a playlist
1579 return self.extract_info(
1580 ie_result['url'], download,
1581 ie_key=ie_result.get('ie_key'),
1582 extra_info=extra_info)
1583 elif result_type == 'url_transparent':
1584 # Use the information from the embedding page
1585 info = self.extract_info(
1586 ie_result['url'], ie_key=ie_result.get('ie_key'),
1587 extra_info=extra_info, download=False, process=False)
1588
1589 # extract_info may return None when ignoreerrors is enabled and
1590 # extraction failed with an error, don't crash and return early
1591 # in this case
1592 if not info:
1593 return info
1594
1595 new_result = info.copy()
1596 new_result.update(filter_dict(ie_result, lambda k, v: (
1597 v is not None and k not in {'_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'})))
1598
1599 # Extracted info may not be a video result (i.e.
1600 # info.get('_type', 'video') != video) but rather an url or
1601 # url_transparent. In such cases outer metadata (from ie_result)
1602 # should be propagated to inner one (info). For this to happen
1603 # _type of info should be overridden with url_transparent. This
1604 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1605 if new_result.get('_type') == 'url':
1606 new_result['_type'] = 'url_transparent'
1607
1608 return self.process_ie_result(
1609 new_result, download=download, extra_info=extra_info)
1610 elif result_type in ('playlist', 'multi_video'):
1611 # Protect from infinite recursion due to recursively nested playlists
1612 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1613 webpage_url = ie_result['webpage_url']
1614 if webpage_url in self._playlist_urls:
1615 self.to_screen(
1616 '[download] Skipping already downloaded playlist: %s'
1617 % ie_result.get('title') or ie_result.get('id'))
1618 return
1619
1620 self._playlist_level += 1
1621 self._playlist_urls.add(webpage_url)
1622 self._fill_common_fields(ie_result, False)
1623 self._sanitize_thumbnails(ie_result)
1624 try:
1625 return self.__process_playlist(ie_result, download)
1626 finally:
1627 self._playlist_level -= 1
1628 if not self._playlist_level:
1629 self._playlist_urls.clear()
1630 elif result_type == 'compat_list':
1631 self.report_warning(
1632 'Extractor %s returned a compat_list result. '
1633 'It needs to be updated.' % ie_result.get('extractor'))
1634
1635 def _fixup(r):
1636 self.add_extra_info(r, {
1637 'extractor': ie_result['extractor'],
1638 'webpage_url': ie_result['webpage_url'],
1639 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1640 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1641 'extractor_key': ie_result['extractor_key'],
1642 })
1643 return r
1644 ie_result['entries'] = [
1645 self.process_ie_result(_fixup(r), download, extra_info)
1646 for r in ie_result['entries']
1647 ]
1648 return ie_result
1649 else:
1650 raise Exception('Invalid result type: %s' % result_type)
1651
1652 def _ensure_dir_exists(self, path):
1653 return make_dir(path, self.report_error)
1654
1655 @staticmethod
1656 def _playlist_infodict(ie_result, **kwargs):
1657 return {
1658 **ie_result,
1659 'playlist': ie_result.get('title') or ie_result.get('id'),
1660 'playlist_id': ie_result.get('id'),
1661 'playlist_title': ie_result.get('title'),
1662 'playlist_uploader': ie_result.get('uploader'),
1663 'playlist_uploader_id': ie_result.get('uploader_id'),
1664 'playlist_index': 0,
1665 **kwargs,
1666 }
1667
1668 def __process_playlist(self, ie_result, download):
1669 """Process each entry in the playlist"""
1670 title = ie_result.get('title') or ie_result.get('id') or '<Untitled>'
1671 self.to_screen(f'[download] Downloading playlist: {title}')
1672
1673 all_entries = PlaylistEntries(self, ie_result)
1674 entries = orderedSet(all_entries.get_requested_items())
1675 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*entries)) or ([], [])
1676 n_entries, ie_result['playlist_count'] = len(entries), all_entries.full_count
1677
1678 _infojson_written = False
1679 write_playlist_files = self.params.get('allow_playlist_files', True)
1680 if write_playlist_files and self.params.get('list_thumbnails'):
1681 self.list_thumbnails(ie_result)
1682 if write_playlist_files and not self.params.get('simulate'):
1683 ie_copy = self._playlist_infodict(ie_result, n_entries=n_entries)
1684 _infojson_written = self._write_info_json(
1685 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1686 if _infojson_written is None:
1687 return
1688 if self._write_description('playlist', ie_result,
1689 self.prepare_filename(ie_copy, 'pl_description')) is None:
1690 return
1691 # TODO: This should be passed to ThumbnailsConvertor if necessary
1692 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1693
1694 if self.params.get('playlistreverse', False):
1695 entries = entries[::-1]
1696 if self.params.get('playlistrandom', False):
1697 random.shuffle(entries)
1698
1699 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} videos'
1700 f'{format_field(ie_result, "playlist_count", " of %s")}')
1701
1702 failures = 0
1703 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1704 for i, (playlist_index, entry) in enumerate(entries, 1):
1705 # TODO: Add auto-generated fields
1706 if self._match_entry(entry, incomplete=True) is not None:
1707 continue
1708
1709 if 'playlist-index' in self.params.get('compat_opts', []):
1710 playlist_index = ie_result['requested_entries'][i - 1]
1711 self.to_screen('[download] Downloading video %s of %s' % (
1712 self._format_screen(i, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1713
1714 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1715 entry_result = self.__process_iterable_entry(entry, download, {
1716 'n_entries': n_entries,
1717 '__last_playlist_index': max(ie_result['requested_entries']),
1718 'playlist_count': ie_result.get('playlist_count'),
1719 'playlist_index': playlist_index,
1720 'playlist_autonumber': i,
1721 'playlist': title,
1722 'playlist_id': ie_result.get('id'),
1723 'playlist_title': ie_result.get('title'),
1724 'playlist_uploader': ie_result.get('uploader'),
1725 'playlist_uploader_id': ie_result.get('uploader_id'),
1726 'extractor': ie_result['extractor'],
1727 'webpage_url': ie_result['webpage_url'],
1728 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1729 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1730 'extractor_key': ie_result['extractor_key'],
1731 })
1732 if not entry_result:
1733 failures += 1
1734 if failures >= max_failures:
1735 self.report_error(
1736 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1737 break
1738 entries[i - 1] = (playlist_index, entry_result)
1739
1740 # Update with processed data
1741 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*entries)) or ([], [])
1742
1743 # Write the updated info to json
1744 if _infojson_written is True and self._write_info_json(
1745 'updated playlist', ie_result,
1746 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1747 return
1748
1749 ie_result = self.run_all_pps('playlist', ie_result)
1750 self.to_screen(f'[download] Finished downloading playlist: {title}')
1751 return ie_result
1752
1753 @_handle_extraction_exceptions
1754 def __process_iterable_entry(self, entry, download, extra_info):
1755 return self.process_ie_result(
1756 entry, download=download, extra_info=extra_info)
1757
1758 def _build_format_filter(self, filter_spec):
1759 " Returns a function to filter the formats according to the filter_spec "
1760
1761 OPERATORS = {
1762 '<': operator.lt,
1763 '<=': operator.le,
1764 '>': operator.gt,
1765 '>=': operator.ge,
1766 '=': operator.eq,
1767 '!=': operator.ne,
1768 }
1769 operator_rex = re.compile(r'''(?x)\s*
1770 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1771 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1772 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1773 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1774 m = operator_rex.fullmatch(filter_spec)
1775 if m:
1776 try:
1777 comparison_value = int(m.group('value'))
1778 except ValueError:
1779 comparison_value = parse_filesize(m.group('value'))
1780 if comparison_value is None:
1781 comparison_value = parse_filesize(m.group('value') + 'B')
1782 if comparison_value is None:
1783 raise ValueError(
1784 'Invalid value %r in format specification %r' % (
1785 m.group('value'), filter_spec))
1786 op = OPERATORS[m.group('op')]
1787
1788 if not m:
1789 STR_OPERATORS = {
1790 '=': operator.eq,
1791 '^=': lambda attr, value: attr.startswith(value),
1792 '$=': lambda attr, value: attr.endswith(value),
1793 '*=': lambda attr, value: value in attr,
1794 '~=': lambda attr, value: value.search(attr) is not None
1795 }
1796 str_operator_rex = re.compile(r'''(?x)\s*
1797 (?P<key>[a-zA-Z0-9._-]+)\s*
1798 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1799 (?P<quote>["'])?
1800 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1801 (?(quote)(?P=quote))\s*
1802 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1803 m = str_operator_rex.fullmatch(filter_spec)
1804 if m:
1805 if m.group('op') == '~=':
1806 comparison_value = re.compile(m.group('value'))
1807 else:
1808 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1809 str_op = STR_OPERATORS[m.group('op')]
1810 if m.group('negation'):
1811 op = lambda attr, value: not str_op(attr, value)
1812 else:
1813 op = str_op
1814
1815 if not m:
1816 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1817
1818 def _filter(f):
1819 actual_value = f.get(m.group('key'))
1820 if actual_value is None:
1821 return m.group('none_inclusive')
1822 return op(actual_value, comparison_value)
1823 return _filter
1824
1825 def _check_formats(self, formats):
1826 for f in formats:
1827 self.to_screen('[info] Testing format %s' % f['format_id'])
1828 path = self.get_output_path('temp')
1829 if not self._ensure_dir_exists(f'{path}/'):
1830 continue
1831 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1832 temp_file.close()
1833 try:
1834 success, _ = self.dl(temp_file.name, f, test=True)
1835 except (DownloadError, OSError, ValueError) + network_exceptions:
1836 success = False
1837 finally:
1838 if os.path.exists(temp_file.name):
1839 try:
1840 os.remove(temp_file.name)
1841 except OSError:
1842 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1843 if success:
1844 yield f
1845 else:
1846 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1847
1848 def _default_format_spec(self, info_dict, download=True):
1849
1850 def can_merge():
1851 merger = FFmpegMergerPP(self)
1852 return merger.available and merger.can_merge()
1853
1854 prefer_best = (
1855 not self.params.get('simulate')
1856 and download
1857 and (
1858 not can_merge()
1859 or info_dict.get('is_live') and not self.params.get('live_from_start')
1860 or self.outtmpl_dict['default'] == '-'))
1861 compat = (
1862 prefer_best
1863 or self.params.get('allow_multiple_audio_streams', False)
1864 or 'format-spec' in self.params['compat_opts'])
1865
1866 return (
1867 'best/bestvideo+bestaudio' if prefer_best
1868 else 'bestvideo*+bestaudio/best' if not compat
1869 else 'bestvideo+bestaudio/best')
1870
1871 def build_format_selector(self, format_spec):
1872 def syntax_error(note, start):
1873 message = (
1874 'Invalid format specification: '
1875 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1876 return SyntaxError(message)
1877
1878 PICKFIRST = 'PICKFIRST'
1879 MERGE = 'MERGE'
1880 SINGLE = 'SINGLE'
1881 GROUP = 'GROUP'
1882 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1883
1884 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1885 'video': self.params.get('allow_multiple_video_streams', False)}
1886
1887 check_formats = self.params.get('check_formats') == 'selected'
1888
1889 def _parse_filter(tokens):
1890 filter_parts = []
1891 for type, string, start, _, _ in tokens:
1892 if type == tokenize.OP and string == ']':
1893 return ''.join(filter_parts)
1894 else:
1895 filter_parts.append(string)
1896
1897 def _remove_unused_ops(tokens):
1898 # Remove operators that we don't use and join them with the surrounding strings
1899 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1900 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1901 last_string, last_start, last_end, last_line = None, None, None, None
1902 for type, string, start, end, line in tokens:
1903 if type == tokenize.OP and string == '[':
1904 if last_string:
1905 yield tokenize.NAME, last_string, last_start, last_end, last_line
1906 last_string = None
1907 yield type, string, start, end, line
1908 # everything inside brackets will be handled by _parse_filter
1909 for type, string, start, end, line in tokens:
1910 yield type, string, start, end, line
1911 if type == tokenize.OP and string == ']':
1912 break
1913 elif type == tokenize.OP and string in ALLOWED_OPS:
1914 if last_string:
1915 yield tokenize.NAME, last_string, last_start, last_end, last_line
1916 last_string = None
1917 yield type, string, start, end, line
1918 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1919 if not last_string:
1920 last_string = string
1921 last_start = start
1922 last_end = end
1923 else:
1924 last_string += string
1925 if last_string:
1926 yield tokenize.NAME, last_string, last_start, last_end, last_line
1927
1928 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1929 selectors = []
1930 current_selector = None
1931 for type, string, start, _, _ in tokens:
1932 # ENCODING is only defined in python 3.x
1933 if type == getattr(tokenize, 'ENCODING', None):
1934 continue
1935 elif type in [tokenize.NAME, tokenize.NUMBER]:
1936 current_selector = FormatSelector(SINGLE, string, [])
1937 elif type == tokenize.OP:
1938 if string == ')':
1939 if not inside_group:
1940 # ')' will be handled by the parentheses group
1941 tokens.restore_last_token()
1942 break
1943 elif inside_merge and string in ['/', ',']:
1944 tokens.restore_last_token()
1945 break
1946 elif inside_choice and string == ',':
1947 tokens.restore_last_token()
1948 break
1949 elif string == ',':
1950 if not current_selector:
1951 raise syntax_error('"," must follow a format selector', start)
1952 selectors.append(current_selector)
1953 current_selector = None
1954 elif string == '/':
1955 if not current_selector:
1956 raise syntax_error('"/" must follow a format selector', start)
1957 first_choice = current_selector
1958 second_choice = _parse_format_selection(tokens, inside_choice=True)
1959 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1960 elif string == '[':
1961 if not current_selector:
1962 current_selector = FormatSelector(SINGLE, 'best', [])
1963 format_filter = _parse_filter(tokens)
1964 current_selector.filters.append(format_filter)
1965 elif string == '(':
1966 if current_selector:
1967 raise syntax_error('Unexpected "("', start)
1968 group = _parse_format_selection(tokens, inside_group=True)
1969 current_selector = FormatSelector(GROUP, group, [])
1970 elif string == '+':
1971 if not current_selector:
1972 raise syntax_error('Unexpected "+"', start)
1973 selector_1 = current_selector
1974 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1975 if not selector_2:
1976 raise syntax_error('Expected a selector', start)
1977 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1978 else:
1979 raise syntax_error(f'Operator not recognized: "{string}"', start)
1980 elif type == tokenize.ENDMARKER:
1981 break
1982 if current_selector:
1983 selectors.append(current_selector)
1984 return selectors
1985
1986 def _merge(formats_pair):
1987 format_1, format_2 = formats_pair
1988
1989 formats_info = []
1990 formats_info.extend(format_1.get('requested_formats', (format_1,)))
1991 formats_info.extend(format_2.get('requested_formats', (format_2,)))
1992
1993 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
1994 get_no_more = {'video': False, 'audio': False}
1995 for (i, fmt_info) in enumerate(formats_info):
1996 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
1997 formats_info.pop(i)
1998 continue
1999 for aud_vid in ['audio', 'video']:
2000 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2001 if get_no_more[aud_vid]:
2002 formats_info.pop(i)
2003 break
2004 get_no_more[aud_vid] = True
2005
2006 if len(formats_info) == 1:
2007 return formats_info[0]
2008
2009 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2010 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2011
2012 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2013 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2014
2015 output_ext = self.params.get('merge_output_format')
2016 if not output_ext:
2017 if the_only_video:
2018 output_ext = the_only_video['ext']
2019 elif the_only_audio and not video_fmts:
2020 output_ext = the_only_audio['ext']
2021 else:
2022 output_ext = 'mkv'
2023
2024 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2025
2026 new_dict = {
2027 'requested_formats': formats_info,
2028 'format': '+'.join(filtered('format')),
2029 'format_id': '+'.join(filtered('format_id')),
2030 'ext': output_ext,
2031 'protocol': '+'.join(map(determine_protocol, formats_info)),
2032 'language': '+'.join(orderedSet(filtered('language'))) or None,
2033 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2034 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2035 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2036 }
2037
2038 if the_only_video:
2039 new_dict.update({
2040 'width': the_only_video.get('width'),
2041 'height': the_only_video.get('height'),
2042 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2043 'fps': the_only_video.get('fps'),
2044 'dynamic_range': the_only_video.get('dynamic_range'),
2045 'vcodec': the_only_video.get('vcodec'),
2046 'vbr': the_only_video.get('vbr'),
2047 'stretched_ratio': the_only_video.get('stretched_ratio'),
2048 })
2049
2050 if the_only_audio:
2051 new_dict.update({
2052 'acodec': the_only_audio.get('acodec'),
2053 'abr': the_only_audio.get('abr'),
2054 'asr': the_only_audio.get('asr'),
2055 })
2056
2057 return new_dict
2058
2059 def _check_formats(formats):
2060 if not check_formats:
2061 yield from formats
2062 return
2063 yield from self._check_formats(formats)
2064
2065 def _build_selector_function(selector):
2066 if isinstance(selector, list): # ,
2067 fs = [_build_selector_function(s) for s in selector]
2068
2069 def selector_function(ctx):
2070 for f in fs:
2071 yield from f(ctx)
2072 return selector_function
2073
2074 elif selector.type == GROUP: # ()
2075 selector_function = _build_selector_function(selector.selector)
2076
2077 elif selector.type == PICKFIRST: # /
2078 fs = [_build_selector_function(s) for s in selector.selector]
2079
2080 def selector_function(ctx):
2081 for f in fs:
2082 picked_formats = list(f(ctx))
2083 if picked_formats:
2084 return picked_formats
2085 return []
2086
2087 elif selector.type == MERGE: # +
2088 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2089
2090 def selector_function(ctx):
2091 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2092 yield _merge(pair)
2093
2094 elif selector.type == SINGLE: # atom
2095 format_spec = selector.selector or 'best'
2096
2097 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2098 if format_spec == 'all':
2099 def selector_function(ctx):
2100 yield from _check_formats(ctx['formats'][::-1])
2101 elif format_spec == 'mergeall':
2102 def selector_function(ctx):
2103 formats = list(_check_formats(
2104 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2105 if not formats:
2106 return
2107 merged_format = formats[-1]
2108 for f in formats[-2::-1]:
2109 merged_format = _merge((merged_format, f))
2110 yield merged_format
2111
2112 else:
2113 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2114 mobj = re.match(
2115 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2116 format_spec)
2117 if mobj is not None:
2118 format_idx = int_or_none(mobj.group('n'), default=1)
2119 format_reverse = mobj.group('bw')[0] == 'b'
2120 format_type = (mobj.group('type') or [None])[0]
2121 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2122 format_modified = mobj.group('mod') is not None
2123
2124 format_fallback = not format_type and not format_modified # for b, w
2125 _filter_f = (
2126 (lambda f: f.get('%scodec' % format_type) != 'none')
2127 if format_type and format_modified # bv*, ba*, wv*, wa*
2128 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2129 if format_type # bv, ba, wv, wa
2130 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2131 if not format_modified # b, w
2132 else lambda f: True) # b*, w*
2133 filter_f = lambda f: _filter_f(f) and (
2134 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2135 else:
2136 if format_spec in self._format_selection_exts['audio']:
2137 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2138 elif format_spec in self._format_selection_exts['video']:
2139 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2140 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2141 elif format_spec in self._format_selection_exts['storyboards']:
2142 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2143 else:
2144 filter_f = lambda f: f.get('format_id') == format_spec # id
2145
2146 def selector_function(ctx):
2147 formats = list(ctx['formats'])
2148 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2149 if not matches:
2150 if format_fallback and ctx['incomplete_formats']:
2151 # for extractors with incomplete formats (audio only (soundcloud)
2152 # or video only (imgur)) best/worst will fallback to
2153 # best/worst {video,audio}-only format
2154 matches = formats
2155 elif seperate_fallback and not ctx['has_merged_format']:
2156 # for compatibility with youtube-dl when there is no pre-merged format
2157 matches = list(filter(seperate_fallback, formats))
2158 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2159 try:
2160 yield matches[format_idx - 1]
2161 except LazyList.IndexError:
2162 return
2163
2164 filters = [self._build_format_filter(f) for f in selector.filters]
2165
2166 def final_selector(ctx):
2167 ctx_copy = dict(ctx)
2168 for _filter in filters:
2169 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2170 return selector_function(ctx_copy)
2171 return final_selector
2172
2173 stream = io.BytesIO(format_spec.encode())
2174 try:
2175 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2176 except tokenize.TokenError:
2177 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2178
2179 class TokenIterator:
2180 def __init__(self, tokens):
2181 self.tokens = tokens
2182 self.counter = 0
2183
2184 def __iter__(self):
2185 return self
2186
2187 def __next__(self):
2188 if self.counter >= len(self.tokens):
2189 raise StopIteration()
2190 value = self.tokens[self.counter]
2191 self.counter += 1
2192 return value
2193
2194 next = __next__
2195
2196 def restore_last_token(self):
2197 self.counter -= 1
2198
2199 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2200 return _build_selector_function(parsed_selector)
2201
2202 def _calc_headers(self, info_dict):
2203 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2204
2205 cookies = self._calc_cookies(info_dict['url'])
2206 if cookies:
2207 res['Cookie'] = cookies
2208
2209 if 'X-Forwarded-For' not in res:
2210 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2211 if x_forwarded_for_ip:
2212 res['X-Forwarded-For'] = x_forwarded_for_ip
2213
2214 return res
2215
2216 def _calc_cookies(self, url):
2217 pr = sanitized_Request(url)
2218 self.cookiejar.add_cookie_header(pr)
2219 return pr.get_header('Cookie')
2220
2221 def _sort_thumbnails(self, thumbnails):
2222 thumbnails.sort(key=lambda t: (
2223 t.get('preference') if t.get('preference') is not None else -1,
2224 t.get('width') if t.get('width') is not None else -1,
2225 t.get('height') if t.get('height') is not None else -1,
2226 t.get('id') if t.get('id') is not None else '',
2227 t.get('url')))
2228
2229 def _sanitize_thumbnails(self, info_dict):
2230 thumbnails = info_dict.get('thumbnails')
2231 if thumbnails is None:
2232 thumbnail = info_dict.get('thumbnail')
2233 if thumbnail:
2234 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2235 if not thumbnails:
2236 return
2237
2238 def check_thumbnails(thumbnails):
2239 for t in thumbnails:
2240 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2241 try:
2242 self.urlopen(HEADRequest(t['url']))
2243 except network_exceptions as err:
2244 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2245 continue
2246 yield t
2247
2248 self._sort_thumbnails(thumbnails)
2249 for i, t in enumerate(thumbnails):
2250 if t.get('id') is None:
2251 t['id'] = '%d' % i
2252 if t.get('width') and t.get('height'):
2253 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2254 t['url'] = sanitize_url(t['url'])
2255
2256 if self.params.get('check_formats') is True:
2257 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2258 else:
2259 info_dict['thumbnails'] = thumbnails
2260
2261 def _fill_common_fields(self, info_dict, is_video=True):
2262 # TODO: move sanitization here
2263 if is_video:
2264 # playlists are allowed to lack "title"
2265 title = info_dict.get('title', NO_DEFAULT)
2266 if title is NO_DEFAULT:
2267 raise ExtractorError('Missing "title" field in extractor result',
2268 video_id=info_dict['id'], ie=info_dict['extractor'])
2269 info_dict['fulltitle'] = title
2270 if not title:
2271 if title == '':
2272 self.write_debug('Extractor gave empty title. Creating a generic title')
2273 else:
2274 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2275 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2276
2277 if info_dict.get('duration') is not None:
2278 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2279
2280 for ts_key, date_key in (
2281 ('timestamp', 'upload_date'),
2282 ('release_timestamp', 'release_date'),
2283 ('modified_timestamp', 'modified_date'),
2284 ):
2285 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2286 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2287 # see http://bugs.python.org/issue1646728)
2288 with contextlib.suppress(ValueError, OverflowError, OSError):
2289 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2290 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2291
2292 live_keys = ('is_live', 'was_live')
2293 live_status = info_dict.get('live_status')
2294 if live_status is None:
2295 for key in live_keys:
2296 if info_dict.get(key) is False:
2297 continue
2298 if info_dict.get(key):
2299 live_status = key
2300 break
2301 if all(info_dict.get(key) is False for key in live_keys):
2302 live_status = 'not_live'
2303 if live_status:
2304 info_dict['live_status'] = live_status
2305 for key in live_keys:
2306 if info_dict.get(key) is None:
2307 info_dict[key] = (live_status == key)
2308
2309 # Auto generate title fields corresponding to the *_number fields when missing
2310 # in order to always have clean titles. This is very common for TV series.
2311 for field in ('chapter', 'season', 'episode'):
2312 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2313 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2314
2315 def _raise_pending_errors(self, info):
2316 err = info.pop('__pending_error', None)
2317 if err:
2318 self.report_error(err, tb=False)
2319
2320 def process_video_result(self, info_dict, download=True):
2321 assert info_dict.get('_type', 'video') == 'video'
2322 self._num_videos += 1
2323
2324 if 'id' not in info_dict:
2325 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2326 elif not info_dict.get('id'):
2327 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2328
2329 def report_force_conversion(field, field_not, conversion):
2330 self.report_warning(
2331 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2332 % (field, field_not, conversion))
2333
2334 def sanitize_string_field(info, string_field):
2335 field = info.get(string_field)
2336 if field is None or isinstance(field, compat_str):
2337 return
2338 report_force_conversion(string_field, 'a string', 'string')
2339 info[string_field] = compat_str(field)
2340
2341 def sanitize_numeric_fields(info):
2342 for numeric_field in self._NUMERIC_FIELDS:
2343 field = info.get(numeric_field)
2344 if field is None or isinstance(field, (int, float)):
2345 continue
2346 report_force_conversion(numeric_field, 'numeric', 'int')
2347 info[numeric_field] = int_or_none(field)
2348
2349 sanitize_string_field(info_dict, 'id')
2350 sanitize_numeric_fields(info_dict)
2351 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2352 self.report_warning('"duration" field is negative, there is an error in extractor')
2353
2354 if 'playlist' not in info_dict:
2355 # It isn't part of a playlist
2356 info_dict['playlist'] = None
2357 info_dict['playlist_index'] = None
2358
2359 self._sanitize_thumbnails(info_dict)
2360
2361 thumbnail = info_dict.get('thumbnail')
2362 thumbnails = info_dict.get('thumbnails')
2363 if thumbnail:
2364 info_dict['thumbnail'] = sanitize_url(thumbnail)
2365 elif thumbnails:
2366 info_dict['thumbnail'] = thumbnails[-1]['url']
2367
2368 if info_dict.get('display_id') is None and 'id' in info_dict:
2369 info_dict['display_id'] = info_dict['id']
2370
2371 self._fill_common_fields(info_dict)
2372
2373 for cc_kind in ('subtitles', 'automatic_captions'):
2374 cc = info_dict.get(cc_kind)
2375 if cc:
2376 for _, subtitle in cc.items():
2377 for subtitle_format in subtitle:
2378 if subtitle_format.get('url'):
2379 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2380 if subtitle_format.get('ext') is None:
2381 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2382
2383 automatic_captions = info_dict.get('automatic_captions')
2384 subtitles = info_dict.get('subtitles')
2385
2386 info_dict['requested_subtitles'] = self.process_subtitles(
2387 info_dict['id'], subtitles, automatic_captions)
2388
2389 if info_dict.get('formats') is None:
2390 # There's only one format available
2391 formats = [info_dict]
2392 else:
2393 formats = info_dict['formats']
2394
2395 # or None ensures --clean-infojson removes it
2396 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2397 if not self.params.get('allow_unplayable_formats'):
2398 formats = [f for f in formats if not f.get('has_drm')]
2399 if info_dict['_has_drm'] and all(
2400 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2401 self.report_warning(
2402 'This video is DRM protected and only images are available for download. '
2403 'Use --list-formats to see them')
2404
2405 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2406 if not get_from_start:
2407 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2408 if info_dict.get('is_live') and formats:
2409 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2410 if get_from_start and not formats:
2411 self.raise_no_formats(info_dict, msg=(
2412 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2413 'If you want to download from the current time, use --no-live-from-start'))
2414
2415 if not formats:
2416 self.raise_no_formats(info_dict)
2417
2418 def is_wellformed(f):
2419 url = f.get('url')
2420 if not url:
2421 self.report_warning(
2422 '"url" field is missing or empty - skipping format, '
2423 'there is an error in extractor')
2424 return False
2425 if isinstance(url, bytes):
2426 sanitize_string_field(f, 'url')
2427 return True
2428
2429 # Filter out malformed formats for better extraction robustness
2430 formats = list(filter(is_wellformed, formats))
2431
2432 formats_dict = {}
2433
2434 # We check that all the formats have the format and format_id fields
2435 for i, format in enumerate(formats):
2436 sanitize_string_field(format, 'format_id')
2437 sanitize_numeric_fields(format)
2438 format['url'] = sanitize_url(format['url'])
2439 if not format.get('format_id'):
2440 format['format_id'] = compat_str(i)
2441 else:
2442 # Sanitize format_id from characters used in format selector expression
2443 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2444 format_id = format['format_id']
2445 if format_id not in formats_dict:
2446 formats_dict[format_id] = []
2447 formats_dict[format_id].append(format)
2448
2449 # Make sure all formats have unique format_id
2450 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2451 for format_id, ambiguous_formats in formats_dict.items():
2452 ambigious_id = len(ambiguous_formats) > 1
2453 for i, format in enumerate(ambiguous_formats):
2454 if ambigious_id:
2455 format['format_id'] = '%s-%d' % (format_id, i)
2456 if format.get('ext') is None:
2457 format['ext'] = determine_ext(format['url']).lower()
2458 # Ensure there is no conflict between id and ext in format selection
2459 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2460 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2461 format['format_id'] = 'f%s' % format['format_id']
2462
2463 for i, format in enumerate(formats):
2464 if format.get('format') is None:
2465 format['format'] = '{id} - {res}{note}'.format(
2466 id=format['format_id'],
2467 res=self.format_resolution(format),
2468 note=format_field(format, 'format_note', ' (%s)'),
2469 )
2470 if format.get('protocol') is None:
2471 format['protocol'] = determine_protocol(format)
2472 if format.get('resolution') is None:
2473 format['resolution'] = self.format_resolution(format, default=None)
2474 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2475 format['dynamic_range'] = 'SDR'
2476 if (info_dict.get('duration') and format.get('tbr')
2477 and not format.get('filesize') and not format.get('filesize_approx')):
2478 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2479
2480 # Add HTTP headers, so that external programs can use them from the
2481 # json output
2482 full_format_info = info_dict.copy()
2483 full_format_info.update(format)
2484 format['http_headers'] = self._calc_headers(full_format_info)
2485 # Remove private housekeeping stuff
2486 if '__x_forwarded_for_ip' in info_dict:
2487 del info_dict['__x_forwarded_for_ip']
2488
2489 if self.params.get('check_formats') is True:
2490 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2491
2492 if not formats or formats[0] is not info_dict:
2493 # only set the 'formats' fields if the original info_dict list them
2494 # otherwise we end up with a circular reference, the first (and unique)
2495 # element in the 'formats' field in info_dict is info_dict itself,
2496 # which can't be exported to json
2497 info_dict['formats'] = formats
2498
2499 info_dict, _ = self.pre_process(info_dict)
2500
2501 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2502 return info_dict
2503
2504 self.post_extract(info_dict)
2505 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2506
2507 # The pre-processors may have modified the formats
2508 formats = info_dict.get('formats', [info_dict])
2509
2510 list_only = self.params.get('simulate') is None and (
2511 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2512 interactive_format_selection = not list_only and self.format_selector == '-'
2513 if self.params.get('list_thumbnails'):
2514 self.list_thumbnails(info_dict)
2515 if self.params.get('listsubtitles'):
2516 if 'automatic_captions' in info_dict:
2517 self.list_subtitles(
2518 info_dict['id'], automatic_captions, 'automatic captions')
2519 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2520 if self.params.get('listformats') or interactive_format_selection:
2521 self.list_formats(info_dict)
2522 if list_only:
2523 # Without this printing, -F --print-json will not work
2524 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2525 return info_dict
2526
2527 format_selector = self.format_selector
2528 if format_selector is None:
2529 req_format = self._default_format_spec(info_dict, download=download)
2530 self.write_debug('Default format spec: %s' % req_format)
2531 format_selector = self.build_format_selector(req_format)
2532
2533 while True:
2534 if interactive_format_selection:
2535 req_format = input(
2536 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2537 try:
2538 format_selector = self.build_format_selector(req_format)
2539 except SyntaxError as err:
2540 self.report_error(err, tb=False, is_error=False)
2541 continue
2542
2543 formats_to_download = list(format_selector({
2544 'formats': formats,
2545 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2546 'incomplete_formats': (
2547 # All formats are video-only or
2548 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2549 # all formats are audio-only
2550 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2551 }))
2552 if interactive_format_selection and not formats_to_download:
2553 self.report_error('Requested format is not available', tb=False, is_error=False)
2554 continue
2555 break
2556
2557 if not formats_to_download:
2558 if not self.params.get('ignore_no_formats_error'):
2559 raise ExtractorError(
2560 'Requested format is not available. Use --list-formats for a list of available formats',
2561 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2562 self.report_warning('Requested format is not available')
2563 # Process what we can, even without any available formats.
2564 formats_to_download = [{}]
2565
2566 requested_ranges = self.params.get('download_ranges')
2567 if requested_ranges:
2568 requested_ranges = tuple(requested_ranges(info_dict, self))
2569
2570 best_format, downloaded_formats = formats_to_download[-1], []
2571 if download:
2572 if best_format:
2573 def to_screen(*msg):
2574 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2575
2576 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2577 (f['format_id'] for f in formats_to_download))
2578 if requested_ranges:
2579 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2580 (f'{int(c["start_time"])}-{int(c["end_time"])}' for c in requested_ranges))
2581 max_downloads_reached = False
2582
2583 for fmt, chapter in itertools.product(formats_to_download, requested_ranges or [{}]):
2584 new_info = self._copy_infodict(info_dict)
2585 new_info.update(fmt)
2586 if chapter:
2587 new_info.update({
2588 'section_start': chapter.get('start_time'),
2589 'section_end': chapter.get('end_time', 0),
2590 'section_title': chapter.get('title'),
2591 'section_number': chapter.get('index'),
2592 })
2593 downloaded_formats.append(new_info)
2594 try:
2595 self.process_info(new_info)
2596 except MaxDownloadsReached:
2597 max_downloads_reached = True
2598 self._raise_pending_errors(new_info)
2599 # Remove copied info
2600 for key, val in tuple(new_info.items()):
2601 if info_dict.get(key) == val:
2602 new_info.pop(key)
2603 if max_downloads_reached:
2604 break
2605
2606 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2607 assert write_archive.issubset({True, False, 'ignore'})
2608 if True in write_archive and False not in write_archive:
2609 self.record_download_archive(info_dict)
2610
2611 info_dict['requested_downloads'] = downloaded_formats
2612 info_dict = self.run_all_pps('after_video', info_dict)
2613 if max_downloads_reached:
2614 raise MaxDownloadsReached()
2615
2616 # We update the info dict with the selected best quality format (backwards compatibility)
2617 info_dict.update(best_format)
2618 return info_dict
2619
2620 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2621 """Select the requested subtitles and their format"""
2622 available_subs, normal_sub_langs = {}, []
2623 if normal_subtitles and self.params.get('writesubtitles'):
2624 available_subs.update(normal_subtitles)
2625 normal_sub_langs = tuple(normal_subtitles.keys())
2626 if automatic_captions and self.params.get('writeautomaticsub'):
2627 for lang, cap_info in automatic_captions.items():
2628 if lang not in available_subs:
2629 available_subs[lang] = cap_info
2630
2631 if (not self.params.get('writesubtitles') and not
2632 self.params.get('writeautomaticsub') or not
2633 available_subs):
2634 return None
2635
2636 all_sub_langs = tuple(available_subs.keys())
2637 if self.params.get('allsubtitles', False):
2638 requested_langs = all_sub_langs
2639 elif self.params.get('subtitleslangs', False):
2640 # A list is used so that the order of languages will be the same as
2641 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2642 requested_langs = []
2643 for lang_re in self.params.get('subtitleslangs'):
2644 discard = lang_re[0] == '-'
2645 if discard:
2646 lang_re = lang_re[1:]
2647 if lang_re == 'all':
2648 if discard:
2649 requested_langs = []
2650 else:
2651 requested_langs.extend(all_sub_langs)
2652 continue
2653 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2654 if discard:
2655 for lang in current_langs:
2656 while lang in requested_langs:
2657 requested_langs.remove(lang)
2658 else:
2659 requested_langs.extend(current_langs)
2660 requested_langs = orderedSet(requested_langs)
2661 elif normal_sub_langs:
2662 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2663 else:
2664 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2665 if requested_langs:
2666 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2667
2668 formats_query = self.params.get('subtitlesformat', 'best')
2669 formats_preference = formats_query.split('/') if formats_query else []
2670 subs = {}
2671 for lang in requested_langs:
2672 formats = available_subs.get(lang)
2673 if formats is None:
2674 self.report_warning(f'{lang} subtitles not available for {video_id}')
2675 continue
2676 for ext in formats_preference:
2677 if ext == 'best':
2678 f = formats[-1]
2679 break
2680 matches = list(filter(lambda f: f['ext'] == ext, formats))
2681 if matches:
2682 f = matches[-1]
2683 break
2684 else:
2685 f = formats[-1]
2686 self.report_warning(
2687 'No subtitle format found matching "%s" for language %s, '
2688 'using %s' % (formats_query, lang, f['ext']))
2689 subs[lang] = f
2690 return subs
2691
2692 def _forceprint(self, key, info_dict):
2693 if info_dict is None:
2694 return
2695 info_copy = info_dict.copy()
2696 info_copy['formats_table'] = self.render_formats_table(info_dict)
2697 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2698 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2699 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2700
2701 def format_tmpl(tmpl):
2702 mobj = re.match(r'\w+(=?)$', tmpl)
2703 if mobj and mobj.group(1):
2704 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2705 elif mobj:
2706 return f'%({tmpl})s'
2707 return tmpl
2708
2709 for tmpl in self.params['forceprint'].get(key, []):
2710 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2711
2712 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2713 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2714 tmpl = format_tmpl(tmpl)
2715 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2716 if self._ensure_dir_exists(filename):
2717 with open(filename, 'a', encoding='utf-8') as f:
2718 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2719
2720 def __forced_printings(self, info_dict, filename, incomplete):
2721 def print_mandatory(field, actual_field=None):
2722 if actual_field is None:
2723 actual_field = field
2724 if (self.params.get('force%s' % field, False)
2725 and (not incomplete or info_dict.get(actual_field) is not None)):
2726 self.to_stdout(info_dict[actual_field])
2727
2728 def print_optional(field):
2729 if (self.params.get('force%s' % field, False)
2730 and info_dict.get(field) is not None):
2731 self.to_stdout(info_dict[field])
2732
2733 info_dict = info_dict.copy()
2734 if filename is not None:
2735 info_dict['filename'] = filename
2736 if info_dict.get('requested_formats') is not None:
2737 # For RTMP URLs, also include the playpath
2738 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2739 elif info_dict.get('url'):
2740 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2741
2742 if (self.params.get('forcejson')
2743 or self.params['forceprint'].get('video')
2744 or self.params['print_to_file'].get('video')):
2745 self.post_extract(info_dict)
2746 self._forceprint('video', info_dict)
2747
2748 print_mandatory('title')
2749 print_mandatory('id')
2750 print_mandatory('url', 'urls')
2751 print_optional('thumbnail')
2752 print_optional('description')
2753 print_optional('filename')
2754 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2755 self.to_stdout(formatSeconds(info_dict['duration']))
2756 print_mandatory('format')
2757
2758 if self.params.get('forcejson'):
2759 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2760
2761 def dl(self, name, info, subtitle=False, test=False):
2762 if not info.get('url'):
2763 self.raise_no_formats(info, True)
2764
2765 if test:
2766 verbose = self.params.get('verbose')
2767 params = {
2768 'test': True,
2769 'quiet': self.params.get('quiet') or not verbose,
2770 'verbose': verbose,
2771 'noprogress': not verbose,
2772 'nopart': True,
2773 'skip_unavailable_fragments': False,
2774 'keep_fragments': False,
2775 'overwrites': True,
2776 '_no_ytdl_file': True,
2777 }
2778 else:
2779 params = self.params
2780 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2781 if not test:
2782 for ph in self._progress_hooks:
2783 fd.add_progress_hook(ph)
2784 urls = '", "'.join(
2785 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2786 for f in info.get('requested_formats', []) or [info])
2787 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2788
2789 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2790 # But it may contain objects that are not deep-copyable
2791 new_info = self._copy_infodict(info)
2792 if new_info.get('http_headers') is None:
2793 new_info['http_headers'] = self._calc_headers(new_info)
2794 return fd.download(name, new_info, subtitle)
2795
2796 def existing_file(self, filepaths, *, default_overwrite=True):
2797 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2798 if existing_files and not self.params.get('overwrites', default_overwrite):
2799 return existing_files[0]
2800
2801 for file in existing_files:
2802 self.report_file_delete(file)
2803 os.remove(file)
2804 return None
2805
2806 def process_info(self, info_dict):
2807 """Process a single resolved IE result. (Modifies it in-place)"""
2808
2809 assert info_dict.get('_type', 'video') == 'video'
2810 original_infodict = info_dict
2811
2812 if 'format' not in info_dict and 'ext' in info_dict:
2813 info_dict['format'] = info_dict['ext']
2814
2815 # This is mostly just for backward compatibility of process_info
2816 # As a side-effect, this allows for format-specific filters
2817 if self._match_entry(info_dict) is not None:
2818 info_dict['__write_download_archive'] = 'ignore'
2819 return
2820
2821 # Does nothing under normal operation - for backward compatibility of process_info
2822 self.post_extract(info_dict)
2823 self._num_downloads += 1
2824
2825 # info_dict['_filename'] needs to be set for backward compatibility
2826 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2827 temp_filename = self.prepare_filename(info_dict, 'temp')
2828 files_to_move = {}
2829
2830 # Forced printings
2831 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2832
2833 def check_max_downloads():
2834 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
2835 raise MaxDownloadsReached()
2836
2837 if self.params.get('simulate'):
2838 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2839 check_max_downloads()
2840 return
2841
2842 if full_filename is None:
2843 return
2844 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2845 return
2846 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2847 return
2848
2849 if self._write_description('video', info_dict,
2850 self.prepare_filename(info_dict, 'description')) is None:
2851 return
2852
2853 sub_files = self._write_subtitles(info_dict, temp_filename)
2854 if sub_files is None:
2855 return
2856 files_to_move.update(dict(sub_files))
2857
2858 thumb_files = self._write_thumbnails(
2859 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2860 if thumb_files is None:
2861 return
2862 files_to_move.update(dict(thumb_files))
2863
2864 infofn = self.prepare_filename(info_dict, 'infojson')
2865 _infojson_written = self._write_info_json('video', info_dict, infofn)
2866 if _infojson_written:
2867 info_dict['infojson_filename'] = infofn
2868 # For backward compatibility, even though it was a private field
2869 info_dict['__infojson_filename'] = infofn
2870 elif _infojson_written is None:
2871 return
2872
2873 # Note: Annotations are deprecated
2874 annofn = None
2875 if self.params.get('writeannotations', False):
2876 annofn = self.prepare_filename(info_dict, 'annotation')
2877 if annofn:
2878 if not self._ensure_dir_exists(encodeFilename(annofn)):
2879 return
2880 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2881 self.to_screen('[info] Video annotations are already present')
2882 elif not info_dict.get('annotations'):
2883 self.report_warning('There are no annotations to write.')
2884 else:
2885 try:
2886 self.to_screen('[info] Writing video annotations to: ' + annofn)
2887 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2888 annofile.write(info_dict['annotations'])
2889 except (KeyError, TypeError):
2890 self.report_warning('There are no annotations to write.')
2891 except OSError:
2892 self.report_error('Cannot write annotations file: ' + annofn)
2893 return
2894
2895 # Write internet shortcut files
2896 def _write_link_file(link_type):
2897 url = try_get(info_dict['webpage_url'], iri_to_uri)
2898 if not url:
2899 self.report_warning(
2900 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2901 return True
2902 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2903 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2904 return False
2905 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2906 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2907 return True
2908 try:
2909 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2910 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2911 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2912 template_vars = {'url': url}
2913 if link_type == 'desktop':
2914 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2915 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2916 except OSError:
2917 self.report_error(f'Cannot write internet shortcut {linkfn}')
2918 return False
2919 return True
2920
2921 write_links = {
2922 'url': self.params.get('writeurllink'),
2923 'webloc': self.params.get('writewebloclink'),
2924 'desktop': self.params.get('writedesktoplink'),
2925 }
2926 if self.params.get('writelink'):
2927 link_type = ('webloc' if sys.platform == 'darwin'
2928 else 'desktop' if sys.platform.startswith('linux')
2929 else 'url')
2930 write_links[link_type] = True
2931
2932 if any(should_write and not _write_link_file(link_type)
2933 for link_type, should_write in write_links.items()):
2934 return
2935
2936 def replace_info_dict(new_info):
2937 nonlocal info_dict
2938 if new_info == info_dict:
2939 return
2940 info_dict.clear()
2941 info_dict.update(new_info)
2942
2943 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2944 replace_info_dict(new_info)
2945
2946 if self.params.get('skip_download'):
2947 info_dict['filepath'] = temp_filename
2948 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2949 info_dict['__files_to_move'] = files_to_move
2950 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2951 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2952 else:
2953 # Download
2954 info_dict.setdefault('__postprocessors', [])
2955 try:
2956
2957 def existing_video_file(*filepaths):
2958 ext = info_dict.get('ext')
2959 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
2960 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
2961 default_overwrite=False)
2962 if file:
2963 info_dict['ext'] = os.path.splitext(file)[1][1:]
2964 return file
2965
2966 success = True
2967 merger, fd = FFmpegMergerPP(self), None
2968 if info_dict.get('protocol') or info_dict.get('url'):
2969 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
2970 if fd is not FFmpegFD and (
2971 info_dict.get('section_start') or info_dict.get('section_end')):
2972 msg = ('This format cannot be partially downloaded' if merger.available
2973 else 'You have requested downloading the video partially, but ffmpeg is not installed')
2974 self.report_error(f'{msg}. Aborting')
2975 return
2976
2977 if info_dict.get('requested_formats') is not None:
2978
2979 def compatible_formats(formats):
2980 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2981 video_formats = [format for format in formats if format.get('vcodec') != 'none']
2982 audio_formats = [format for format in formats if format.get('acodec') != 'none']
2983 if len(video_formats) > 2 or len(audio_formats) > 2:
2984 return False
2985
2986 # Check extension
2987 exts = {format.get('ext') for format in formats}
2988 COMPATIBLE_EXTS = (
2989 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
2990 {'webm'},
2991 )
2992 for ext_sets in COMPATIBLE_EXTS:
2993 if ext_sets.issuperset(exts):
2994 return True
2995 # TODO: Check acodec/vcodec
2996 return False
2997
2998 requested_formats = info_dict['requested_formats']
2999 old_ext = info_dict['ext']
3000 if self.params.get('merge_output_format') is None:
3001 if not compatible_formats(requested_formats):
3002 info_dict['ext'] = 'mkv'
3003 self.report_warning(
3004 'Requested formats are incompatible for merge and will be merged into mkv')
3005 if (info_dict['ext'] == 'webm'
3006 and info_dict.get('thumbnails')
3007 # check with type instead of pp_key, __name__, or isinstance
3008 # since we dont want any custom PPs to trigger this
3009 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3010 info_dict['ext'] = 'mkv'
3011 self.report_warning(
3012 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3013 new_ext = info_dict['ext']
3014
3015 def correct_ext(filename, ext=new_ext):
3016 if filename == '-':
3017 return filename
3018 filename_real_ext = os.path.splitext(filename)[1][1:]
3019 filename_wo_ext = (
3020 os.path.splitext(filename)[0]
3021 if filename_real_ext in (old_ext, new_ext)
3022 else filename)
3023 return f'{filename_wo_ext}.{ext}'
3024
3025 # Ensure filename always has a correct extension for successful merge
3026 full_filename = correct_ext(full_filename)
3027 temp_filename = correct_ext(temp_filename)
3028 dl_filename = existing_video_file(full_filename, temp_filename)
3029 info_dict['__real_download'] = False
3030
3031 downloaded = []
3032 if dl_filename is not None:
3033 self.report_file_already_downloaded(dl_filename)
3034 elif fd:
3035 for f in requested_formats if fd != FFmpegFD else []:
3036 f['filepath'] = fname = prepend_extension(
3037 correct_ext(temp_filename, info_dict['ext']),
3038 'f%s' % f['format_id'], info_dict['ext'])
3039 downloaded.append(fname)
3040 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3041 success, real_download = self.dl(temp_filename, info_dict)
3042 info_dict['__real_download'] = real_download
3043 else:
3044 if self.params.get('allow_unplayable_formats'):
3045 self.report_warning(
3046 'You have requested merging of multiple formats '
3047 'while also allowing unplayable formats to be downloaded. '
3048 'The formats won\'t be merged to prevent data corruption.')
3049 elif not merger.available:
3050 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3051 if not self.params.get('ignoreerrors'):
3052 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3053 return
3054 self.report_warning(f'{msg}. The formats won\'t be merged')
3055
3056 if temp_filename == '-':
3057 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3058 else 'but the formats are incompatible for simultaneous download' if merger.available
3059 else 'but ffmpeg is not installed')
3060 self.report_warning(
3061 f'You have requested downloading multiple formats to stdout {reason}. '
3062 'The formats will be streamed one after the other')
3063 fname = temp_filename
3064 for f in requested_formats:
3065 new_info = dict(info_dict)
3066 del new_info['requested_formats']
3067 new_info.update(f)
3068 if temp_filename != '-':
3069 fname = prepend_extension(
3070 correct_ext(temp_filename, new_info['ext']),
3071 'f%s' % f['format_id'], new_info['ext'])
3072 if not self._ensure_dir_exists(fname):
3073 return
3074 f['filepath'] = fname
3075 downloaded.append(fname)
3076 partial_success, real_download = self.dl(fname, new_info)
3077 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3078 success = success and partial_success
3079
3080 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3081 info_dict['__postprocessors'].append(merger)
3082 info_dict['__files_to_merge'] = downloaded
3083 # Even if there were no downloads, it is being merged only now
3084 info_dict['__real_download'] = True
3085 else:
3086 for file in downloaded:
3087 files_to_move[file] = None
3088 else:
3089 # Just a single file
3090 dl_filename = existing_video_file(full_filename, temp_filename)
3091 if dl_filename is None or dl_filename == temp_filename:
3092 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3093 # So we should try to resume the download
3094 success, real_download = self.dl(temp_filename, info_dict)
3095 info_dict['__real_download'] = real_download
3096 else:
3097 self.report_file_already_downloaded(dl_filename)
3098
3099 dl_filename = dl_filename or temp_filename
3100 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3101
3102 except network_exceptions as err:
3103 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3104 return
3105 except OSError as err:
3106 raise UnavailableVideoError(err)
3107 except (ContentTooShortError, ) as err:
3108 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3109 return
3110
3111 self._raise_pending_errors(info_dict)
3112 if success and full_filename != '-':
3113
3114 def fixup():
3115 do_fixup = True
3116 fixup_policy = self.params.get('fixup')
3117 vid = info_dict['id']
3118
3119 if fixup_policy in ('ignore', 'never'):
3120 return
3121 elif fixup_policy == 'warn':
3122 do_fixup = 'warn'
3123 elif fixup_policy != 'force':
3124 assert fixup_policy in ('detect_or_warn', None)
3125 if not info_dict.get('__real_download'):
3126 do_fixup = False
3127
3128 def ffmpeg_fixup(cndn, msg, cls):
3129 if not (do_fixup and cndn):
3130 return
3131 elif do_fixup == 'warn':
3132 self.report_warning(f'{vid}: {msg}')
3133 return
3134 pp = cls(self)
3135 if pp.available:
3136 info_dict['__postprocessors'].append(pp)
3137 else:
3138 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3139
3140 stretched_ratio = info_dict.get('stretched_ratio')
3141 ffmpeg_fixup(
3142 stretched_ratio not in (1, None),
3143 f'Non-uniform pixel ratio {stretched_ratio}',
3144 FFmpegFixupStretchedPP)
3145
3146 ffmpeg_fixup(
3147 (info_dict.get('requested_formats') is None
3148 and info_dict.get('container') == 'm4a_dash'
3149 and info_dict.get('ext') == 'm4a'),
3150 'writing DASH m4a. Only some players support this container',
3151 FFmpegFixupM4aPP)
3152
3153 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3154 downloader = downloader.FD_NAME if downloader else None
3155
3156 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3157 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3158 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3159 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3160 FFmpegFixupM3u8PP)
3161 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3162 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3163
3164 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3165 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3166
3167 fixup()
3168 try:
3169 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3170 except PostProcessingError as err:
3171 self.report_error('Postprocessing: %s' % str(err))
3172 return
3173 try:
3174 for ph in self._post_hooks:
3175 ph(info_dict['filepath'])
3176 except Exception as err:
3177 self.report_error('post hooks: %s' % str(err))
3178 return
3179 info_dict['__write_download_archive'] = True
3180
3181 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3182 if self.params.get('force_write_download_archive'):
3183 info_dict['__write_download_archive'] = True
3184 check_max_downloads()
3185
3186 def __download_wrapper(self, func):
3187 @functools.wraps(func)
3188 def wrapper(*args, **kwargs):
3189 try:
3190 res = func(*args, **kwargs)
3191 except UnavailableVideoError as e:
3192 self.report_error(e)
3193 except DownloadCancelled as e:
3194 self.to_screen(f'[info] {e}')
3195 if not self.params.get('break_per_url'):
3196 raise
3197 else:
3198 if self.params.get('dump_single_json', False):
3199 self.post_extract(res)
3200 self.to_stdout(json.dumps(self.sanitize_info(res)))
3201 return wrapper
3202
3203 def download(self, url_list):
3204 """Download a given list of URLs."""
3205 url_list = variadic(url_list) # Passing a single URL is a common mistake
3206 outtmpl = self.outtmpl_dict['default']
3207 if (len(url_list) > 1
3208 and outtmpl != '-'
3209 and '%' not in outtmpl
3210 and self.params.get('max_downloads') != 1):
3211 raise SameFileError(outtmpl)
3212
3213 for url in url_list:
3214 self.__download_wrapper(self.extract_info)(
3215 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3216
3217 return self._download_retcode
3218
3219 def download_with_info_file(self, info_filename):
3220 with contextlib.closing(fileinput.FileInput(
3221 [info_filename], mode='r',
3222 openhook=fileinput.hook_encoded('utf-8'))) as f:
3223 # FileInput doesn't have a read method, we can't call json.load
3224 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3225 try:
3226 self.__download_wrapper(self.process_ie_result)(info, download=True)
3227 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3228 if not isinstance(e, EntryNotInPlaylist):
3229 self.to_stderr('\r')
3230 webpage_url = info.get('webpage_url')
3231 if webpage_url is not None:
3232 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3233 return self.download([webpage_url])
3234 else:
3235 raise
3236 return self._download_retcode
3237
3238 @staticmethod
3239 def sanitize_info(info_dict, remove_private_keys=False):
3240 ''' Sanitize the infodict for converting to json '''
3241 if info_dict is None:
3242 return info_dict
3243 info_dict.setdefault('epoch', int(time.time()))
3244 info_dict.setdefault('_type', 'video')
3245
3246 if remove_private_keys:
3247 reject = lambda k, v: v is None or k.startswith('__') or k in {
3248 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3249 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3250 }
3251 else:
3252 reject = lambda k, v: False
3253
3254 def filter_fn(obj):
3255 if isinstance(obj, dict):
3256 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3257 elif isinstance(obj, (list, tuple, set, LazyList)):
3258 return list(map(filter_fn, obj))
3259 elif obj is None or isinstance(obj, (str, int, float, bool)):
3260 return obj
3261 else:
3262 return repr(obj)
3263
3264 return filter_fn(info_dict)
3265
3266 @staticmethod
3267 def filter_requested_info(info_dict, actually_filter=True):
3268 ''' Alias of sanitize_info for backward compatibility '''
3269 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3270
3271 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3272 for filename in set(filter(None, files_to_delete)):
3273 if msg:
3274 self.to_screen(msg % filename)
3275 try:
3276 os.remove(filename)
3277 except OSError:
3278 self.report_warning(f'Unable to delete file {filename}')
3279 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3280 del info['__files_to_move'][filename]
3281
3282 @staticmethod
3283 def post_extract(info_dict):
3284 def actual_post_extract(info_dict):
3285 if info_dict.get('_type') in ('playlist', 'multi_video'):
3286 for video_dict in info_dict.get('entries', {}):
3287 actual_post_extract(video_dict or {})
3288 return
3289
3290 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3291 info_dict.update(post_extractor())
3292
3293 actual_post_extract(info_dict or {})
3294
3295 def run_pp(self, pp, infodict):
3296 files_to_delete = []
3297 if '__files_to_move' not in infodict:
3298 infodict['__files_to_move'] = {}
3299 try:
3300 files_to_delete, infodict = pp.run(infodict)
3301 except PostProcessingError as e:
3302 # Must be True and not 'only_download'
3303 if self.params.get('ignoreerrors') is True:
3304 self.report_error(e)
3305 return infodict
3306 raise
3307
3308 if not files_to_delete:
3309 return infodict
3310 if self.params.get('keepvideo', False):
3311 for f in files_to_delete:
3312 infodict['__files_to_move'].setdefault(f, '')
3313 else:
3314 self._delete_downloaded_files(
3315 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3316 return infodict
3317
3318 def run_all_pps(self, key, info, *, additional_pps=None):
3319 self._forceprint(key, info)
3320 for pp in (additional_pps or []) + self._pps[key]:
3321 info = self.run_pp(pp, info)
3322 return info
3323
3324 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3325 info = dict(ie_info)
3326 info['__files_to_move'] = files_to_move or {}
3327 try:
3328 info = self.run_all_pps(key, info)
3329 except PostProcessingError as err:
3330 msg = f'Preprocessing: {err}'
3331 info.setdefault('__pending_error', msg)
3332 self.report_error(msg, is_error=False)
3333 return info, info.pop('__files_to_move', None)
3334
3335 def post_process(self, filename, info, files_to_move=None):
3336 """Run all the postprocessors on the given file."""
3337 info['filepath'] = filename
3338 info['__files_to_move'] = files_to_move or {}
3339 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3340 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3341 del info['__files_to_move']
3342 return self.run_all_pps('after_move', info)
3343
3344 def _make_archive_id(self, info_dict):
3345 video_id = info_dict.get('id')
3346 if not video_id:
3347 return
3348 # Future-proof against any change in case
3349 # and backwards compatibility with prior versions
3350 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3351 if extractor is None:
3352 url = str_or_none(info_dict.get('url'))
3353 if not url:
3354 return
3355 # Try to find matching extractor for the URL and take its ie_key
3356 for ie_key, ie in self._ies.items():
3357 if ie.suitable(url):
3358 extractor = ie_key
3359 break
3360 else:
3361 return
3362 return f'{extractor.lower()} {video_id}'
3363
3364 def in_download_archive(self, info_dict):
3365 fn = self.params.get('download_archive')
3366 if fn is None:
3367 return False
3368
3369 vid_id = self._make_archive_id(info_dict)
3370 if not vid_id:
3371 return False # Incomplete video information
3372
3373 return vid_id in self.archive
3374
3375 def record_download_archive(self, info_dict):
3376 fn = self.params.get('download_archive')
3377 if fn is None:
3378 return
3379 vid_id = self._make_archive_id(info_dict)
3380 assert vid_id
3381 self.write_debug(f'Adding to archive: {vid_id}')
3382 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3383 archive_file.write(vid_id + '\n')
3384 self.archive.add(vid_id)
3385
3386 @staticmethod
3387 def format_resolution(format, default='unknown'):
3388 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3389 return 'audio only'
3390 if format.get('resolution') is not None:
3391 return format['resolution']
3392 if format.get('width') and format.get('height'):
3393 return '%dx%d' % (format['width'], format['height'])
3394 elif format.get('height'):
3395 return '%sp' % format['height']
3396 elif format.get('width'):
3397 return '%dx?' % format['width']
3398 return default
3399
3400 def _list_format_headers(self, *headers):
3401 if self.params.get('listformats_table', True) is not False:
3402 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3403 return headers
3404
3405 def _format_note(self, fdict):
3406 res = ''
3407 if fdict.get('ext') in ['f4f', 'f4m']:
3408 res += '(unsupported)'
3409 if fdict.get('language'):
3410 if res:
3411 res += ' '
3412 res += '[%s]' % fdict['language']
3413 if fdict.get('format_note') is not None:
3414 if res:
3415 res += ' '
3416 res += fdict['format_note']
3417 if fdict.get('tbr') is not None:
3418 if res:
3419 res += ', '
3420 res += '%4dk' % fdict['tbr']
3421 if fdict.get('container') is not None:
3422 if res:
3423 res += ', '
3424 res += '%s container' % fdict['container']
3425 if (fdict.get('vcodec') is not None
3426 and fdict.get('vcodec') != 'none'):
3427 if res:
3428 res += ', '
3429 res += fdict['vcodec']
3430 if fdict.get('vbr') is not None:
3431 res += '@'
3432 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3433 res += 'video@'
3434 if fdict.get('vbr') is not None:
3435 res += '%4dk' % fdict['vbr']
3436 if fdict.get('fps') is not None:
3437 if res:
3438 res += ', '
3439 res += '%sfps' % fdict['fps']
3440 if fdict.get('acodec') is not None:
3441 if res:
3442 res += ', '
3443 if fdict['acodec'] == 'none':
3444 res += 'video only'
3445 else:
3446 res += '%-5s' % fdict['acodec']
3447 elif fdict.get('abr') is not None:
3448 if res:
3449 res += ', '
3450 res += 'audio'
3451 if fdict.get('abr') is not None:
3452 res += '@%3dk' % fdict['abr']
3453 if fdict.get('asr') is not None:
3454 res += ' (%5dHz)' % fdict['asr']
3455 if fdict.get('filesize') is not None:
3456 if res:
3457 res += ', '
3458 res += format_bytes(fdict['filesize'])
3459 elif fdict.get('filesize_approx') is not None:
3460 if res:
3461 res += ', '
3462 res += '~' + format_bytes(fdict['filesize_approx'])
3463 return res
3464
3465 def render_formats_table(self, info_dict):
3466 if not info_dict.get('formats') and not info_dict.get('url'):
3467 return None
3468
3469 formats = info_dict.get('formats', [info_dict])
3470 if not self.params.get('listformats_table', True) is not False:
3471 table = [
3472 [
3473 format_field(f, 'format_id'),
3474 format_field(f, 'ext'),
3475 self.format_resolution(f),
3476 self._format_note(f)
3477 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3478 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3479
3480 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3481 table = [
3482 [
3483 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3484 format_field(f, 'ext'),
3485 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3486 format_field(f, 'fps', '\t%d'),
3487 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3488 delim,
3489 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3490 format_field(f, 'tbr', '\t%dk'),
3491 shorten_protocol_name(f.get('protocol', '')),
3492 delim,
3493 format_field(f, 'vcodec', default='unknown').replace(
3494 'none', 'images' if f.get('acodec') == 'none'
3495 else self._format_out('audio only', self.Styles.SUPPRESS)),
3496 format_field(f, 'vbr', '\t%dk'),
3497 format_field(f, 'acodec', default='unknown').replace(
3498 'none', '' if f.get('vcodec') == 'none'
3499 else self._format_out('video only', self.Styles.SUPPRESS)),
3500 format_field(f, 'abr', '\t%dk'),
3501 format_field(f, 'asr', '\t%dHz'),
3502 join_nonempty(
3503 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3504 format_field(f, 'language', '[%s]'),
3505 join_nonempty(format_field(f, 'format_note'),
3506 format_field(f, 'container', ignore=(None, f.get('ext'))),
3507 delim=', '),
3508 delim=' '),
3509 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3510 header_line = self._list_format_headers(
3511 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3512 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3513
3514 return render_table(
3515 header_line, table, hide_empty=True,
3516 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3517
3518 def render_thumbnails_table(self, info_dict):
3519 thumbnails = list(info_dict.get('thumbnails') or [])
3520 if not thumbnails:
3521 return None
3522 return render_table(
3523 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3524 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3525
3526 def render_subtitles_table(self, video_id, subtitles):
3527 def _row(lang, formats):
3528 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3529 if len(set(names)) == 1:
3530 names = [] if names[0] == 'unknown' else names[:1]
3531 return [lang, ', '.join(names), ', '.join(exts)]
3532
3533 if not subtitles:
3534 return None
3535 return render_table(
3536 self._list_format_headers('Language', 'Name', 'Formats'),
3537 [_row(lang, formats) for lang, formats in subtitles.items()],
3538 hide_empty=True)
3539
3540 def __list_table(self, video_id, name, func, *args):
3541 table = func(*args)
3542 if not table:
3543 self.to_screen(f'{video_id} has no {name}')
3544 return
3545 self.to_screen(f'[info] Available {name} for {video_id}:')
3546 self.to_stdout(table)
3547
3548 def list_formats(self, info_dict):
3549 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3550
3551 def list_thumbnails(self, info_dict):
3552 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3553
3554 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3555 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3556
3557 def urlopen(self, req):
3558 """ Start an HTTP download """
3559 if isinstance(req, str):
3560 req = sanitized_Request(req)
3561 return self._opener.open(req, timeout=self._socket_timeout)
3562
3563 def print_debug_header(self):
3564 if not self.params.get('verbose'):
3565 return
3566
3567 # These imports can be slow. So import them only as needed
3568 from .extractor.extractors import _LAZY_LOADER
3569 from .extractor.extractors import _PLUGIN_CLASSES as plugin_extractors
3570
3571 def get_encoding(stream):
3572 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3573 if not supports_terminal_sequences(stream):
3574 from .utils import WINDOWS_VT_MODE # Must be imported locally
3575 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3576 return ret
3577
3578 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3579 locale.getpreferredencoding(),
3580 sys.getfilesystemencoding(),
3581 self.get_encoding(),
3582 ', '.join(
3583 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3584 if stream is not None and key != 'console')
3585 )
3586
3587 logger = self.params.get('logger')
3588 if logger:
3589 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3590 write_debug(encoding_str)
3591 else:
3592 write_string(f'[debug] {encoding_str}\n', encoding=None)
3593 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3594
3595 source = detect_variant()
3596 write_debug(join_nonempty(
3597 'yt-dlp version', __version__,
3598 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3599 '' if source == 'unknown' else f'({source})',
3600 delim=' '))
3601 if not _LAZY_LOADER:
3602 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3603 write_debug('Lazy loading extractors is forcibly disabled')
3604 else:
3605 write_debug('Lazy loading extractors is disabled')
3606 if plugin_extractors or plugin_postprocessors:
3607 write_debug('Plugins: %s' % [
3608 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3609 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3610 if self.params['compat_opts']:
3611 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3612
3613 if source == 'source':
3614 try:
3615 stdout, _, _ = Popen.run(
3616 ['git', 'rev-parse', '--short', 'HEAD'],
3617 text=True, cwd=os.path.dirname(os.path.abspath(__file__)),
3618 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
3619 if re.fullmatch('[0-9a-f]+', stdout.strip()):
3620 write_debug(f'Git HEAD: {stdout.strip()}')
3621 except Exception:
3622 with contextlib.suppress(Exception):
3623 sys.exc_clear()
3624
3625 def python_implementation():
3626 impl_name = platform.python_implementation()
3627 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3628 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3629 return impl_name
3630
3631 write_debug('Python version %s (%s %s) - %s' % (
3632 platform.python_version(),
3633 python_implementation(),
3634 platform.architecture()[0],
3635 platform_name()))
3636
3637 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3638 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3639 if ffmpeg_features:
3640 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3641
3642 exe_versions['rtmpdump'] = rtmpdump_version()
3643 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3644 exe_str = ', '.join(
3645 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3646 ) or 'none'
3647 write_debug('exe versions: %s' % exe_str)
3648
3649 from .compat.compat_utils import get_package_info
3650 from .dependencies import available_dependencies
3651
3652 write_debug('Optional libraries: %s' % (', '.join(sorted({
3653 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3654 })) or 'none'))
3655
3656 self._setup_opener()
3657 proxy_map = {}
3658 for handler in self._opener.handlers:
3659 if hasattr(handler, 'proxies'):
3660 proxy_map.update(handler.proxies)
3661 write_debug(f'Proxy map: {proxy_map}')
3662
3663 # Not implemented
3664 if False and self.params.get('call_home'):
3665 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3666 write_debug('Public IP address: %s' % ipaddr)
3667 latest_version = self.urlopen(
3668 'https://yt-dl.org/latest/version').read().decode()
3669 if version_tuple(latest_version) > version_tuple(__version__):
3670 self.report_warning(
3671 'You are using an outdated version (newest version: %s)! '
3672 'See https://yt-dl.org/update if you need help updating.' %
3673 latest_version)
3674
3675 def _setup_opener(self):
3676 if hasattr(self, '_opener'):
3677 return
3678 timeout_val = self.params.get('socket_timeout')
3679 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3680
3681 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3682 opts_cookiefile = self.params.get('cookiefile')
3683 opts_proxy = self.params.get('proxy')
3684
3685 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3686
3687 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3688 if opts_proxy is not None:
3689 if opts_proxy == '':
3690 proxies = {}
3691 else:
3692 proxies = {'http': opts_proxy, 'https': opts_proxy}
3693 else:
3694 proxies = compat_urllib_request.getproxies()
3695 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3696 if 'http' in proxies and 'https' not in proxies:
3697 proxies['https'] = proxies['http']
3698 proxy_handler = PerRequestProxyHandler(proxies)
3699
3700 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3701 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3702 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3703 redirect_handler = YoutubeDLRedirectHandler()
3704 data_handler = urllib.request.DataHandler()
3705
3706 # When passing our own FileHandler instance, build_opener won't add the
3707 # default FileHandler and allows us to disable the file protocol, which
3708 # can be used for malicious purposes (see
3709 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3710 file_handler = compat_urllib_request.FileHandler()
3711
3712 def file_open(*args, **kwargs):
3713 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3714 file_handler.file_open = file_open
3715
3716 opener = compat_urllib_request.build_opener(
3717 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3718
3719 # Delete the default user-agent header, which would otherwise apply in
3720 # cases where our custom HTTP handler doesn't come into play
3721 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3722 opener.addheaders = []
3723 self._opener = opener
3724
3725 def encode(self, s):
3726 if isinstance(s, bytes):
3727 return s # Already encoded
3728
3729 try:
3730 return s.encode(self.get_encoding())
3731 except UnicodeEncodeError as err:
3732 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3733 raise
3734
3735 def get_encoding(self):
3736 encoding = self.params.get('encoding')
3737 if encoding is None:
3738 encoding = preferredencoding()
3739 return encoding
3740
3741 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3742 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3743 if overwrite is None:
3744 overwrite = self.params.get('overwrites', True)
3745 if not self.params.get('writeinfojson'):
3746 return False
3747 elif not infofn:
3748 self.write_debug(f'Skipping writing {label} infojson')
3749 return False
3750 elif not self._ensure_dir_exists(infofn):
3751 return None
3752 elif not overwrite and os.path.exists(infofn):
3753 self.to_screen(f'[info] {label.title()} metadata is already present')
3754 return 'exists'
3755
3756 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3757 try:
3758 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3759 return True
3760 except OSError:
3761 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3762 return None
3763
3764 def _write_description(self, label, ie_result, descfn):
3765 ''' Write description and returns True = written, False = skip, None = error '''
3766 if not self.params.get('writedescription'):
3767 return False
3768 elif not descfn:
3769 self.write_debug(f'Skipping writing {label} description')
3770 return False
3771 elif not self._ensure_dir_exists(descfn):
3772 return None
3773 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3774 self.to_screen(f'[info] {label.title()} description is already present')
3775 elif ie_result.get('description') is None:
3776 self.report_warning(f'There\'s no {label} description to write')
3777 return False
3778 else:
3779 try:
3780 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3781 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3782 descfile.write(ie_result['description'])
3783 except OSError:
3784 self.report_error(f'Cannot write {label} description file {descfn}')
3785 return None
3786 return True
3787
3788 def _write_subtitles(self, info_dict, filename):
3789 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3790 ret = []
3791 subtitles = info_dict.get('requested_subtitles')
3792 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3793 # subtitles download errors are already managed as troubles in relevant IE
3794 # that way it will silently go on when used with unsupporting IE
3795 return ret
3796
3797 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3798 if not sub_filename_base:
3799 self.to_screen('[info] Skipping writing video subtitles')
3800 return ret
3801 for sub_lang, sub_info in subtitles.items():
3802 sub_format = sub_info['ext']
3803 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3804 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3805 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3806 if existing_sub:
3807 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3808 sub_info['filepath'] = existing_sub
3809 ret.append((existing_sub, sub_filename_final))
3810 continue
3811
3812 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3813 if sub_info.get('data') is not None:
3814 try:
3815 # Use newline='' to prevent conversion of newline characters
3816 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3817 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3818 subfile.write(sub_info['data'])
3819 sub_info['filepath'] = sub_filename
3820 ret.append((sub_filename, sub_filename_final))
3821 continue
3822 except OSError:
3823 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3824 return None
3825
3826 try:
3827 sub_copy = sub_info.copy()
3828 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3829 self.dl(sub_filename, sub_copy, subtitle=True)
3830 sub_info['filepath'] = sub_filename
3831 ret.append((sub_filename, sub_filename_final))
3832 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3833 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3834 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3835 if not self.params.get('ignoreerrors'):
3836 self.report_error(msg)
3837 raise DownloadError(msg)
3838 self.report_warning(msg)
3839 return ret
3840
3841 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3842 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3843 write_all = self.params.get('write_all_thumbnails', False)
3844 thumbnails, ret = [], []
3845 if write_all or self.params.get('writethumbnail', False):
3846 thumbnails = info_dict.get('thumbnails') or []
3847 multiple = write_all and len(thumbnails) > 1
3848
3849 if thumb_filename_base is None:
3850 thumb_filename_base = filename
3851 if thumbnails and not thumb_filename_base:
3852 self.write_debug(f'Skipping writing {label} thumbnail')
3853 return ret
3854
3855 for idx, t in list(enumerate(thumbnails))[::-1]:
3856 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3857 thumb_display_id = f'{label} thumbnail {t["id"]}'
3858 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3859 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3860
3861 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3862 if existing_thumb:
3863 self.to_screen('[info] %s is already present' % (
3864 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3865 t['filepath'] = existing_thumb
3866 ret.append((existing_thumb, thumb_filename_final))
3867 else:
3868 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3869 try:
3870 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3871 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3872 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3873 shutil.copyfileobj(uf, thumbf)
3874 ret.append((thumb_filename, thumb_filename_final))
3875 t['filepath'] = thumb_filename
3876 except network_exceptions as err:
3877 thumbnails.pop(idx)
3878 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3879 if ret and not write_all:
3880 break
3881 return ret