]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[cleanup] Fix some typos (#4194)
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 import collections
2 import contextlib
3 import datetime
4 import errno
5 import fileinput
6 import functools
7 import io
8 import itertools
9 import json
10 import locale
11 import operator
12 import os
13 import platform
14 import random
15 import re
16 import shutil
17 import subprocess
18 import sys
19 import tempfile
20 import time
21 import tokenize
22 import traceback
23 import unicodedata
24 import urllib.request
25 from string import ascii_letters
26
27 from .cache import Cache
28 from .compat import HAS_LEGACY as compat_has_legacy
29 from .compat import compat_os_name, compat_shlex_quote
30 from .cookies import load_cookies
31 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
32 from .downloader.rtmp import rtmpdump_version
33 from .extractor import gen_extractor_classes, get_info_extractor
34 from .extractor.openload import PhantomJSwrapper
35 from .minicurses import format_text
36 from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
37 from .postprocessor import (
38 EmbedThumbnailPP,
39 FFmpegFixupDuplicateMoovPP,
40 FFmpegFixupDurationPP,
41 FFmpegFixupM3u8PP,
42 FFmpegFixupM4aPP,
43 FFmpegFixupStretchedPP,
44 FFmpegFixupTimestampPP,
45 FFmpegMergerPP,
46 FFmpegPostProcessor,
47 MoveFilesAfterDownloadPP,
48 get_postprocessor,
49 )
50 from .update import detect_variant
51 from .utils import (
52 DEFAULT_OUTTMPL,
53 IDENTITY,
54 LINK_TEMPLATES,
55 NO_DEFAULT,
56 NUMBER_RE,
57 OUTTMPL_TYPES,
58 POSTPROCESS_WHEN,
59 STR_FORMAT_RE_TMPL,
60 STR_FORMAT_TYPES,
61 ContentTooShortError,
62 DateRange,
63 DownloadCancelled,
64 DownloadError,
65 EntryNotInPlaylist,
66 ExistingVideoReached,
67 ExtractorError,
68 GeoRestrictedError,
69 HEADRequest,
70 ISO3166Utils,
71 LazyList,
72 MaxDownloadsReached,
73 Namespace,
74 PagedList,
75 PerRequestProxyHandler,
76 PlaylistEntries,
77 Popen,
78 PostProcessingError,
79 ReExtractInfo,
80 RejectedVideoReached,
81 SameFileError,
82 UnavailableVideoError,
83 YoutubeDLCookieProcessor,
84 YoutubeDLHandler,
85 YoutubeDLRedirectHandler,
86 age_restricted,
87 args_to_str,
88 date_from_str,
89 determine_ext,
90 determine_protocol,
91 encode_compat_str,
92 encodeFilename,
93 error_to_compat_str,
94 expand_path,
95 filter_dict,
96 float_or_none,
97 format_bytes,
98 format_decimal_suffix,
99 format_field,
100 formatSeconds,
101 get_domain,
102 int_or_none,
103 iri_to_uri,
104 join_nonempty,
105 locked_file,
106 make_dir,
107 make_HTTPS_handler,
108 merge_headers,
109 network_exceptions,
110 number_of_digits,
111 orderedSet,
112 parse_filesize,
113 platform_name,
114 preferredencoding,
115 prepend_extension,
116 register_socks_protocols,
117 remove_terminal_sequences,
118 render_table,
119 replace_extension,
120 sanitize_filename,
121 sanitize_path,
122 sanitize_url,
123 sanitized_Request,
124 std_headers,
125 str_or_none,
126 strftime_or_none,
127 subtitles_filename,
128 supports_terminal_sequences,
129 timetuple_from_msec,
130 to_high_limit_path,
131 traverse_obj,
132 try_get,
133 url_basename,
134 variadic,
135 version_tuple,
136 windows_enable_vt_mode,
137 write_json_file,
138 write_string,
139 )
140 from .version import RELEASE_GIT_HEAD, __version__
141
142 if compat_os_name == 'nt':
143 import ctypes
144
145
146 class YoutubeDL:
147 """YoutubeDL class.
148
149 YoutubeDL objects are the ones responsible of downloading the
150 actual video file and writing it to disk if the user has requested
151 it, among some other tasks. In most cases there should be one per
152 program. As, given a video URL, the downloader doesn't know how to
153 extract all the needed information, task that InfoExtractors do, it
154 has to pass the URL to one of them.
155
156 For this, YoutubeDL objects have a method that allows
157 InfoExtractors to be registered in a given order. When it is passed
158 a URL, the YoutubeDL object handles it to the first InfoExtractor it
159 finds that reports being able to handle it. The InfoExtractor extracts
160 all the information about the video or videos the URL refers to, and
161 YoutubeDL process the extracted information, possibly using a File
162 Downloader to download the video.
163
164 YoutubeDL objects accept a lot of parameters. In order not to saturate
165 the object constructor with arguments, it receives a dictionary of
166 options instead. These options are available through the params
167 attribute for the InfoExtractors to use. The YoutubeDL also
168 registers itself as the downloader in charge for the InfoExtractors
169 that are added to it, so this is a "mutual registration".
170
171 Available options:
172
173 username: Username for authentication purposes.
174 password: Password for authentication purposes.
175 videopassword: Password for accessing a video.
176 ap_mso: Adobe Pass multiple-system operator identifier.
177 ap_username: Multiple-system operator account username.
178 ap_password: Multiple-system operator account password.
179 usenetrc: Use netrc for authentication instead.
180 verbose: Print additional info to stdout.
181 quiet: Do not print messages to stdout.
182 no_warnings: Do not print out anything for warnings.
183 forceprint: A dict with keys WHEN mapped to a list of templates to
184 print to stdout. The allowed keys are video or any of the
185 items in utils.POSTPROCESS_WHEN.
186 For compatibility, a single list is also accepted
187 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
188 a list of tuples with (template, filename)
189 forcejson: Force printing info_dict as JSON.
190 dump_single_json: Force printing the info_dict of the whole playlist
191 (or video) as a single JSON line.
192 force_write_download_archive: Force writing download archive regardless
193 of 'skip_download' or 'simulate'.
194 simulate: Do not download the video files. If unset (or None),
195 simulate only if listsubtitles, listformats or list_thumbnails is used
196 format: Video format code. see "FORMAT SELECTION" for more details.
197 You can also pass a function. The function takes 'ctx' as
198 argument and returns the formats to download.
199 See "build_format_selector" for an implementation
200 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
201 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
202 extracting metadata even if the video is not actually
203 available for download (experimental)
204 format_sort: A list of fields by which to sort the video formats.
205 See "Sorting Formats" for more details.
206 format_sort_force: Force the given format_sort. see "Sorting Formats"
207 for more details.
208 prefer_free_formats: Whether to prefer video formats with free containers
209 over non-free ones of same quality.
210 allow_multiple_video_streams: Allow multiple video streams to be merged
211 into a single file
212 allow_multiple_audio_streams: Allow multiple audio streams to be merged
213 into a single file
214 check_formats Whether to test if the formats are downloadable.
215 Can be True (check all), False (check none),
216 'selected' (check selected formats),
217 or None (check only if requested by extractor)
218 paths: Dictionary of output paths. The allowed keys are 'home'
219 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
220 outtmpl: Dictionary of templates for output names. Allowed keys
221 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
222 For compatibility with youtube-dl, a single string can also be used
223 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
224 restrictfilenames: Do not allow "&" and spaces in file names
225 trim_file_name: Limit length of filename (extension excluded)
226 windowsfilenames: Force the filenames to be windows compatible
227 ignoreerrors: Do not stop on download/postprocessing errors.
228 Can be 'only_download' to ignore only download errors.
229 Default is 'only_download' for CLI, but False for API
230 skip_playlist_after_errors: Number of allowed failures until the rest of
231 the playlist is skipped
232 force_generic_extractor: Force downloader to use the generic extractor
233 overwrites: Overwrite all video and metadata files if True,
234 overwrite only non-video files if None
235 and don't overwrite any file if False
236 For compatibility with youtube-dl,
237 "nooverwrites" may also be used instead
238 playlist_items: Specific indices of playlist to download.
239 playlistrandom: Download playlist items in random order.
240 lazy_playlist: Process playlist entries as they are received.
241 matchtitle: Download only matching titles.
242 rejecttitle: Reject downloads for matching titles.
243 logger: Log messages to a logging.Logger instance.
244 logtostderr: Log messages to stderr instead of stdout.
245 consoletitle: Display progress in console window's titlebar.
246 writedescription: Write the video description to a .description file
247 writeinfojson: Write the video description to a .info.json file
248 clean_infojson: Remove private fields from the infojson
249 getcomments: Extract video comments. This will not be written to disk
250 unless writeinfojson is also given
251 writeannotations: Write the video annotations to a .annotations.xml file
252 writethumbnail: Write the thumbnail image to a file
253 allow_playlist_files: Whether to write playlists' description, infojson etc
254 also to disk when using the 'write*' options
255 write_all_thumbnails: Write all thumbnail formats to files
256 writelink: Write an internet shortcut file, depending on the
257 current platform (.url/.webloc/.desktop)
258 writeurllink: Write a Windows internet shortcut file (.url)
259 writewebloclink: Write a macOS internet shortcut file (.webloc)
260 writedesktoplink: Write a Linux internet shortcut file (.desktop)
261 writesubtitles: Write the video subtitles to a file
262 writeautomaticsub: Write the automatically generated subtitles to a file
263 listsubtitles: Lists all available subtitles for the video
264 subtitlesformat: The format code for subtitles
265 subtitleslangs: List of languages of the subtitles to download (can be regex).
266 The list may contain "all" to refer to all the available
267 subtitles. The language can be prefixed with a "-" to
268 exclude it from the requested languages. Eg: ['all', '-live_chat']
269 keepvideo: Keep the video file after post-processing
270 daterange: A DateRange object, download only if the upload_date is in the range.
271 skip_download: Skip the actual download of the video file
272 cachedir: Location of the cache files in the filesystem.
273 False to disable filesystem cache.
274 noplaylist: Download single video instead of a playlist if in doubt.
275 age_limit: An integer representing the user's age in years.
276 Unsuitable videos for the given age are skipped.
277 min_views: An integer representing the minimum view count the video
278 must have in order to not be skipped.
279 Videos without view count information are always
280 downloaded. None for no limit.
281 max_views: An integer representing the maximum view count.
282 Videos that are more popular than that are not
283 downloaded.
284 Videos without view count information are always
285 downloaded. None for no limit.
286 download_archive: File name of a file where all downloads are recorded.
287 Videos already present in the file are not downloaded
288 again.
289 break_on_existing: Stop the download process after attempting to download a
290 file that is in the archive.
291 break_on_reject: Stop the download process when encountering a video that
292 has been filtered out.
293 break_per_url: Whether break_on_reject and break_on_existing
294 should act on each input URL as opposed to for the entire queue
295 cookiefile: File name or text stream from where cookies should be read and dumped to
296 cookiesfrombrowser: A tuple containing the name of the browser, the profile
297 name/pathfrom where cookies are loaded, and the name of the
298 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
299 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
300 support RFC 5746 secure renegotiation
301 nocheckcertificate: Do not verify SSL certificates
302 client_certificate: Path to client certificate file in PEM format. May include the private key
303 client_certificate_key: Path to private key file for client certificate
304 client_certificate_password: Password for client certificate private key, if encrypted.
305 If not provided and the key is encrypted, yt-dlp will ask interactively
306 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
307 At the moment, this is only supported by YouTube.
308 http_headers: A dictionary of custom headers to be used for all requests
309 proxy: URL of the proxy server to use
310 geo_verification_proxy: URL of the proxy to use for IP address verification
311 on geo-restricted sites.
312 socket_timeout: Time to wait for unresponsive hosts, in seconds
313 bidi_workaround: Work around buggy terminals without bidirectional text
314 support, using fridibi
315 debug_printtraffic:Print out sent and received HTTP traffic
316 default_search: Prepend this string if an input url is not valid.
317 'auto' for elaborate guessing
318 encoding: Use this encoding instead of the system-specified.
319 extract_flat: Do not resolve URLs, return the immediate result.
320 Pass in 'in_playlist' to only show this behavior for
321 playlist items.
322 wait_for_video: If given, wait for scheduled streams to become available.
323 The value should be a tuple containing the range
324 (min_secs, max_secs) to wait between retries
325 postprocessors: A list of dictionaries, each with an entry
326 * key: The name of the postprocessor. See
327 yt_dlp/postprocessor/__init__.py for a list.
328 * when: When to run the postprocessor. Allowed values are
329 the entries of utils.POSTPROCESS_WHEN
330 Assumed to be 'post_process' if not given
331 progress_hooks: A list of functions that get called on download
332 progress, with a dictionary with the entries
333 * status: One of "downloading", "error", or "finished".
334 Check this first and ignore unknown values.
335 * info_dict: The extracted info_dict
336
337 If status is one of "downloading", or "finished", the
338 following properties may also be present:
339 * filename: The final filename (always present)
340 * tmpfilename: The filename we're currently writing to
341 * downloaded_bytes: Bytes on disk
342 * total_bytes: Size of the whole file, None if unknown
343 * total_bytes_estimate: Guess of the eventual file size,
344 None if unavailable.
345 * elapsed: The number of seconds since download started.
346 * eta: The estimated time in seconds, None if unknown
347 * speed: The download speed in bytes/second, None if
348 unknown
349 * fragment_index: The counter of the currently
350 downloaded video fragment.
351 * fragment_count: The number of fragments (= individual
352 files that will be merged)
353
354 Progress hooks are guaranteed to be called at least once
355 (with status "finished") if the download is successful.
356 postprocessor_hooks: A list of functions that get called on postprocessing
357 progress, with a dictionary with the entries
358 * status: One of "started", "processing", or "finished".
359 Check this first and ignore unknown values.
360 * postprocessor: Name of the postprocessor
361 * info_dict: The extracted info_dict
362
363 Progress hooks are guaranteed to be called at least twice
364 (with status "started" and "finished") if the processing is successful.
365 merge_output_format: Extension to use when merging formats.
366 final_ext: Expected final extension; used to detect when the file was
367 already downloaded and converted
368 fixup: Automatically correct known faults of the file.
369 One of:
370 - "never": do nothing
371 - "warn": only emit a warning
372 - "detect_or_warn": check whether we can do anything
373 about it, warn otherwise (default)
374 source_address: Client-side IP address to bind to.
375 sleep_interval_requests: Number of seconds to sleep between requests
376 during extraction
377 sleep_interval: Number of seconds to sleep before each download when
378 used alone or a lower bound of a range for randomized
379 sleep before each download (minimum possible number
380 of seconds to sleep) when used along with
381 max_sleep_interval.
382 max_sleep_interval:Upper bound of a range for randomized sleep before each
383 download (maximum possible number of seconds to sleep).
384 Must only be used along with sleep_interval.
385 Actual sleep time will be a random float from range
386 [sleep_interval; max_sleep_interval].
387 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
388 listformats: Print an overview of available video formats and exit.
389 list_thumbnails: Print a table of all thumbnails and exit.
390 match_filter: A function that gets called for every video with the signature
391 (info_dict, *, incomplete: bool) -> Optional[str]
392 For backward compatibility with youtube-dl, the signature
393 (info_dict) -> Optional[str] is also allowed.
394 - If it returns a message, the video is ignored.
395 - If it returns None, the video is downloaded.
396 - If it returns utils.NO_DEFAULT, the user is interactively
397 asked whether to download the video.
398 match_filter_func in utils.py is one example for this.
399 no_color: Do not emit color codes in output.
400 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
401 HTTP header
402 geo_bypass_country:
403 Two-letter ISO 3166-2 country code that will be used for
404 explicit geographic restriction bypassing via faking
405 X-Forwarded-For HTTP header
406 geo_bypass_ip_block:
407 IP range in CIDR notation that will be used similarly to
408 geo_bypass_country
409 external_downloader: A dictionary of protocol keys and the executable of the
410 external downloader to use for it. The allowed protocols
411 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
412 Set the value to 'native' to use the native downloader
413 compat_opts: Compatibility options. See "Differences in default behavior".
414 The following options do not work when used through the API:
415 filename, abort-on-error, multistreams, no-live-chat, format-sort
416 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
417 Refer __init__.py for their implementation
418 progress_template: Dictionary of templates for progress outputs.
419 Allowed keys are 'download', 'postprocess',
420 'download-title' (console title) and 'postprocess-title'.
421 The template is mapped on a dictionary with keys 'progress' and 'info'
422 retry_sleep_functions: Dictionary of functions that takes the number of attempts
423 as argument and returns the time to sleep in seconds.
424 Allowed keys are 'http', 'fragment', 'file_access'
425 download_ranges: A function that gets called for every video with the signature
426 (info_dict, *, ydl) -> Iterable[Section].
427 Only the returned sections will be downloaded. Each Section contains:
428 * start_time: Start time of the section in seconds
429 * end_time: End time of the section in seconds
430 * title: Section title (Optional)
431 * index: Section number (Optional)
432
433 The following parameters are not used by YoutubeDL itself, they are used by
434 the downloader (see yt_dlp/downloader/common.py):
435 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
436 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
437 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
438 external_downloader_args, concurrent_fragment_downloads.
439
440 The following options are used by the post processors:
441 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
442 to the binary or its containing directory.
443 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
444 and a list of additional command-line arguments for the
445 postprocessor/executable. The dict can also have "PP+EXE" keys
446 which are used when the given exe is used by the given PP.
447 Use 'default' as the name for arguments to passed to all PP
448 For compatibility with youtube-dl, a single list of args
449 can also be used
450
451 The following options are used by the extractors:
452 extractor_retries: Number of times to retry for known errors
453 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
454 hls_split_discontinuity: Split HLS playlists to different formats at
455 discontinuities such as ad breaks (default: False)
456 extractor_args: A dictionary of arguments to be passed to the extractors.
457 See "EXTRACTOR ARGUMENTS" for details.
458 Eg: {'youtube': {'skip': ['dash', 'hls']}}
459 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
460
461 The following options are deprecated and may be removed in the future:
462
463 playliststart: - Use playlist_items
464 Playlist item to start at.
465 playlistend: - Use playlist_items
466 Playlist item to end at.
467 playlistreverse: - Use playlist_items
468 Download playlist items in reverse order.
469 forceurl: - Use forceprint
470 Force printing final URL.
471 forcetitle: - Use forceprint
472 Force printing title.
473 forceid: - Use forceprint
474 Force printing ID.
475 forcethumbnail: - Use forceprint
476 Force printing thumbnail URL.
477 forcedescription: - Use forceprint
478 Force printing description.
479 forcefilename: - Use forceprint
480 Force printing final filename.
481 forceduration: - Use forceprint
482 Force printing duration.
483 allsubtitles: - Use subtitleslangs = ['all']
484 Downloads all the subtitles of the video
485 (requires writesubtitles or writeautomaticsub)
486 include_ads: - Doesn't work
487 Download ads as well
488 call_home: - Not implemented
489 Boolean, true iff we are allowed to contact the
490 yt-dlp servers for debugging.
491 post_hooks: - Register a custom postprocessor
492 A list of functions that get called as the final step
493 for each video file, after all postprocessors have been
494 called. The filename will be passed as the only argument.
495 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
496 Use the native HLS downloader instead of ffmpeg/avconv
497 if True, otherwise use ffmpeg/avconv if False, otherwise
498 use downloader suggested by extractor if None.
499 prefer_ffmpeg: - avconv support is deprecated
500 If False, use avconv instead of ffmpeg if both are available,
501 otherwise prefer ffmpeg.
502 youtube_include_dash_manifest: - Use extractor_args
503 If True (default), DASH manifests and related
504 data will be downloaded and processed by extractor.
505 You can reduce network I/O by disabling it if you don't
506 care about DASH. (only for youtube)
507 youtube_include_hls_manifest: - Use extractor_args
508 If True (default), HLS manifests and related
509 data will be downloaded and processed by extractor.
510 You can reduce network I/O by disabling it if you don't
511 care about HLS. (only for youtube)
512 """
513
514 _NUMERIC_FIELDS = {
515 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
516 'timestamp', 'release_timestamp',
517 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
518 'average_rating', 'comment_count', 'age_limit',
519 'start_time', 'end_time',
520 'chapter_number', 'season_number', 'episode_number',
521 'track_number', 'disc_number', 'release_year',
522 }
523
524 _format_fields = {
525 # NB: Keep in sync with the docstring of extractor/common.py
526 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
527 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
528 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
529 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
530 'preference', 'language', 'language_preference', 'quality', 'source_preference',
531 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
532 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
533 }
534 _format_selection_exts = {
535 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
536 'video': {'mp4', 'flv', 'webm', '3gp'},
537 'storyboards': {'mhtml'},
538 }
539
540 def __init__(self, params=None, auto_init=True):
541 """Create a FileDownloader object with the given options.
542 @param auto_init Whether to load the default extractors and print header (if verbose).
543 Set to 'no_verbose_header' to not print the header
544 """
545 if params is None:
546 params = {}
547 self.params = params
548 self._ies = {}
549 self._ies_instances = {}
550 self._pps = {k: [] for k in POSTPROCESS_WHEN}
551 self._printed_messages = set()
552 self._first_webpage_request = True
553 self._post_hooks = []
554 self._progress_hooks = []
555 self._postprocessor_hooks = []
556 self._download_retcode = 0
557 self._num_downloads = 0
558 self._num_videos = 0
559 self._playlist_level = 0
560 self._playlist_urls = set()
561 self.cache = Cache(self)
562
563 windows_enable_vt_mode()
564 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
565 self._out_files = Namespace(
566 out=stdout,
567 error=sys.stderr,
568 screen=sys.stderr if self.params.get('quiet') else stdout,
569 console=None if compat_os_name == 'nt' else next(
570 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
571 )
572 self._allow_colors = Namespace(**{
573 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
574 for type_, stream in self._out_files.items_ if type_ != 'console'
575 })
576
577 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 6), (3, 7)
578 current_version = sys.version_info[:2]
579 if current_version < MIN_RECOMMENDED:
580 msg = 'Support for Python version %d.%d has been deprecated and will break in future versions of yt-dlp'
581 if current_version < MIN_SUPPORTED:
582 msg = 'Python version %d.%d is no longer supported'
583 self.deprecation_warning(
584 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
585
586 if self.params.get('allow_unplayable_formats'):
587 self.report_warning(
588 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
589 'This is a developer option intended for debugging. \n'
590 ' If you experience any issues while using this option, '
591 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
592
593 def check_deprecated(param, option, suggestion):
594 if self.params.get(param) is not None:
595 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
596 return True
597 return False
598
599 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
600 if self.params.get('geo_verification_proxy') is None:
601 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
602
603 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
604 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
605 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
606
607 for msg in self.params.get('_warnings', []):
608 self.report_warning(msg)
609 for msg in self.params.get('_deprecation_warnings', []):
610 self.deprecation_warning(msg)
611
612 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
613 if not compat_has_legacy:
614 self.params['compat_opts'].add('no-compat-legacy')
615 if 'list-formats' in self.params['compat_opts']:
616 self.params['listformats_table'] = False
617
618 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
619 # nooverwrites was unnecessarily changed to overwrites
620 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
621 # This ensures compatibility with both keys
622 self.params['overwrites'] = not self.params['nooverwrites']
623 elif self.params.get('overwrites') is None:
624 self.params.pop('overwrites', None)
625 else:
626 self.params['nooverwrites'] = not self.params['overwrites']
627
628 self.params.setdefault('forceprint', {})
629 self.params.setdefault('print_to_file', {})
630
631 # Compatibility with older syntax
632 if not isinstance(params['forceprint'], dict):
633 self.params['forceprint'] = {'video': params['forceprint']}
634
635 if self.params.get('bidi_workaround', False):
636 try:
637 import pty
638 master, slave = pty.openpty()
639 width = shutil.get_terminal_size().columns
640 width_args = [] if width is None else ['-w', str(width)]
641 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
642 try:
643 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
644 except OSError:
645 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
646 self._output_channel = os.fdopen(master, 'rb')
647 except OSError as ose:
648 if ose.errno == errno.ENOENT:
649 self.report_warning(
650 'Could not find fribidi executable, ignoring --bidi-workaround. '
651 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
652 else:
653 raise
654
655 if auto_init:
656 if auto_init != 'no_verbose_header':
657 self.print_debug_header()
658 self.add_default_info_extractors()
659
660 if (sys.platform != 'win32'
661 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
662 and not self.params.get('restrictfilenames', False)):
663 # Unicode filesystem API will throw errors (#1474, #13027)
664 self.report_warning(
665 'Assuming --restrict-filenames since file system encoding '
666 'cannot encode all characters. '
667 'Set the LC_ALL environment variable to fix this.')
668 self.params['restrictfilenames'] = True
669
670 self._parse_outtmpl()
671
672 # Creating format selector here allows us to catch syntax errors before the extraction
673 self.format_selector = (
674 self.params.get('format') if self.params.get('format') in (None, '-')
675 else self.params['format'] if callable(self.params['format'])
676 else self.build_format_selector(self.params['format']))
677
678 # Set http_headers defaults according to std_headers
679 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
680
681 hooks = {
682 'post_hooks': self.add_post_hook,
683 'progress_hooks': self.add_progress_hook,
684 'postprocessor_hooks': self.add_postprocessor_hook,
685 }
686 for opt, fn in hooks.items():
687 for ph in self.params.get(opt, []):
688 fn(ph)
689
690 for pp_def_raw in self.params.get('postprocessors', []):
691 pp_def = dict(pp_def_raw)
692 when = pp_def.pop('when', 'post_process')
693 self.add_post_processor(
694 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
695 when=when)
696
697 self._setup_opener()
698 register_socks_protocols()
699
700 def preload_download_archive(fn):
701 """Preload the archive, if any is specified"""
702 if fn is None:
703 return False
704 self.write_debug(f'Loading archive file {fn!r}')
705 try:
706 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
707 for line in archive_file:
708 self.archive.add(line.strip())
709 except OSError as ioe:
710 if ioe.errno != errno.ENOENT:
711 raise
712 return False
713 return True
714
715 self.archive = set()
716 preload_download_archive(self.params.get('download_archive'))
717
718 def warn_if_short_id(self, argv):
719 # short YouTube ID starting with dash?
720 idxs = [
721 i for i, a in enumerate(argv)
722 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
723 if idxs:
724 correct_argv = (
725 ['yt-dlp']
726 + [a for i, a in enumerate(argv) if i not in idxs]
727 + ['--'] + [argv[i] for i in idxs]
728 )
729 self.report_warning(
730 'Long argument string detected. '
731 'Use -- to separate parameters and URLs, like this:\n%s' %
732 args_to_str(correct_argv))
733
734 def add_info_extractor(self, ie):
735 """Add an InfoExtractor object to the end of the list."""
736 ie_key = ie.ie_key()
737 self._ies[ie_key] = ie
738 if not isinstance(ie, type):
739 self._ies_instances[ie_key] = ie
740 ie.set_downloader(self)
741
742 def _get_info_extractor_class(self, ie_key):
743 ie = self._ies.get(ie_key)
744 if ie is None:
745 ie = get_info_extractor(ie_key)
746 self.add_info_extractor(ie)
747 return ie
748
749 def get_info_extractor(self, ie_key):
750 """
751 Get an instance of an IE with name ie_key, it will try to get one from
752 the _ies list, if there's no instance it will create a new one and add
753 it to the extractor list.
754 """
755 ie = self._ies_instances.get(ie_key)
756 if ie is None:
757 ie = get_info_extractor(ie_key)()
758 self.add_info_extractor(ie)
759 return ie
760
761 def add_default_info_extractors(self):
762 """
763 Add the InfoExtractors returned by gen_extractors to the end of the list
764 """
765 for ie in gen_extractor_classes():
766 self.add_info_extractor(ie)
767
768 def add_post_processor(self, pp, when='post_process'):
769 """Add a PostProcessor object to the end of the chain."""
770 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
771 self._pps[when].append(pp)
772 pp.set_downloader(self)
773
774 def add_post_hook(self, ph):
775 """Add the post hook"""
776 self._post_hooks.append(ph)
777
778 def add_progress_hook(self, ph):
779 """Add the download progress hook"""
780 self._progress_hooks.append(ph)
781
782 def add_postprocessor_hook(self, ph):
783 """Add the postprocessing progress hook"""
784 self._postprocessor_hooks.append(ph)
785 for pps in self._pps.values():
786 for pp in pps:
787 pp.add_progress_hook(ph)
788
789 def _bidi_workaround(self, message):
790 if not hasattr(self, '_output_channel'):
791 return message
792
793 assert hasattr(self, '_output_process')
794 assert isinstance(message, str)
795 line_count = message.count('\n') + 1
796 self._output_process.stdin.write((message + '\n').encode())
797 self._output_process.stdin.flush()
798 res = ''.join(self._output_channel.readline().decode()
799 for _ in range(line_count))
800 return res[:-len('\n')]
801
802 def _write_string(self, message, out=None, only_once=False):
803 if only_once:
804 if message in self._printed_messages:
805 return
806 self._printed_messages.add(message)
807 write_string(message, out=out, encoding=self.params.get('encoding'))
808
809 def to_stdout(self, message, skip_eol=False, quiet=None):
810 """Print message to stdout"""
811 if quiet is not None:
812 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
813 if skip_eol is not False:
814 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. Use "YoutubeDL.to_screen" instead')
815 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
816
817 def to_screen(self, message, skip_eol=False, quiet=None):
818 """Print message to screen if not in quiet mode"""
819 if self.params.get('logger'):
820 self.params['logger'].debug(message)
821 return
822 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
823 return
824 self._write_string(
825 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
826 self._out_files.screen)
827
828 def to_stderr(self, message, only_once=False):
829 """Print message to stderr"""
830 assert isinstance(message, str)
831 if self.params.get('logger'):
832 self.params['logger'].error(message)
833 else:
834 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
835
836 def _send_console_code(self, code):
837 if compat_os_name == 'nt' or not self._out_files.console:
838 return
839 self._write_string(code, self._out_files.console)
840
841 def to_console_title(self, message):
842 if not self.params.get('consoletitle', False):
843 return
844 message = remove_terminal_sequences(message)
845 if compat_os_name == 'nt':
846 if ctypes.windll.kernel32.GetConsoleWindow():
847 # c_wchar_p() might not be necessary if `message` is
848 # already of type unicode()
849 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
850 else:
851 self._send_console_code(f'\033]0;{message}\007')
852
853 def save_console_title(self):
854 if not self.params.get('consoletitle') or self.params.get('simulate'):
855 return
856 self._send_console_code('\033[22;0t') # Save the title on stack
857
858 def restore_console_title(self):
859 if not self.params.get('consoletitle') or self.params.get('simulate'):
860 return
861 self._send_console_code('\033[23;0t') # Restore the title from stack
862
863 def __enter__(self):
864 self.save_console_title()
865 return self
866
867 def __exit__(self, *args):
868 self.restore_console_title()
869
870 if self.params.get('cookiefile') is not None:
871 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
872
873 def trouble(self, message=None, tb=None, is_error=True):
874 """Determine action to take when a download problem appears.
875
876 Depending on if the downloader has been configured to ignore
877 download errors or not, this method may throw an exception or
878 not when errors are found, after printing the message.
879
880 @param tb If given, is additional traceback information
881 @param is_error Whether to raise error according to ignorerrors
882 """
883 if message is not None:
884 self.to_stderr(message)
885 if self.params.get('verbose'):
886 if tb is None:
887 if sys.exc_info()[0]: # if .trouble has been called from an except block
888 tb = ''
889 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
890 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
891 tb += encode_compat_str(traceback.format_exc())
892 else:
893 tb_data = traceback.format_list(traceback.extract_stack())
894 tb = ''.join(tb_data)
895 if tb:
896 self.to_stderr(tb)
897 if not is_error:
898 return
899 if not self.params.get('ignoreerrors'):
900 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
901 exc_info = sys.exc_info()[1].exc_info
902 else:
903 exc_info = sys.exc_info()
904 raise DownloadError(message, exc_info)
905 self._download_retcode = 1
906
907 Styles = Namespace(
908 HEADERS='yellow',
909 EMPHASIS='light blue',
910 FILENAME='green',
911 ID='green',
912 DELIM='blue',
913 ERROR='red',
914 WARNING='yellow',
915 SUPPRESS='light black',
916 )
917
918 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
919 text = str(text)
920 if test_encoding:
921 original_text = text
922 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
923 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
924 text = text.encode(encoding, 'ignore').decode(encoding)
925 if fallback is not None and text != original_text:
926 text = fallback
927 return format_text(text, f) if allow_colors else text if fallback is None else fallback
928
929 def _format_out(self, *args, **kwargs):
930 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
931
932 def _format_screen(self, *args, **kwargs):
933 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
934
935 def _format_err(self, *args, **kwargs):
936 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
937
938 def report_warning(self, message, only_once=False):
939 '''
940 Print the message to stderr, it will be prefixed with 'WARNING:'
941 If stderr is a tty file the 'WARNING:' will be colored
942 '''
943 if self.params.get('logger') is not None:
944 self.params['logger'].warning(message)
945 else:
946 if self.params.get('no_warnings'):
947 return
948 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
949
950 def deprecation_warning(self, message):
951 if self.params.get('logger') is not None:
952 self.params['logger'].warning(f'DeprecationWarning: {message}')
953 else:
954 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
955
956 def report_error(self, message, *args, **kwargs):
957 '''
958 Do the same as trouble, but prefixes the message with 'ERROR:', colored
959 in red if stderr is a tty file.
960 '''
961 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
962
963 def write_debug(self, message, only_once=False):
964 '''Log debug message or Print message to stderr'''
965 if not self.params.get('verbose', False):
966 return
967 message = f'[debug] {message}'
968 if self.params.get('logger'):
969 self.params['logger'].debug(message)
970 else:
971 self.to_stderr(message, only_once)
972
973 def report_file_already_downloaded(self, file_name):
974 """Report file has already been fully downloaded."""
975 try:
976 self.to_screen('[download] %s has already been downloaded' % file_name)
977 except UnicodeEncodeError:
978 self.to_screen('[download] The file has already been downloaded')
979
980 def report_file_delete(self, file_name):
981 """Report that existing file will be deleted."""
982 try:
983 self.to_screen('Deleting existing file %s' % file_name)
984 except UnicodeEncodeError:
985 self.to_screen('Deleting existing file')
986
987 def raise_no_formats(self, info, forced=False, *, msg=None):
988 has_drm = info.get('_has_drm')
989 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
990 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
991 if forced or not ignored:
992 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
993 expected=has_drm or ignored or expected)
994 else:
995 self.report_warning(msg)
996
997 def parse_outtmpl(self):
998 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
999 self._parse_outtmpl()
1000 return self.params['outtmpl']
1001
1002 def _parse_outtmpl(self):
1003 sanitize = IDENTITY
1004 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1005 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1006
1007 outtmpl = self.params.setdefault('outtmpl', {})
1008 if not isinstance(outtmpl, dict):
1009 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1010 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1011
1012 def get_output_path(self, dir_type='', filename=None):
1013 paths = self.params.get('paths', {})
1014 assert isinstance(paths, dict)
1015 path = os.path.join(
1016 expand_path(paths.get('home', '').strip()),
1017 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1018 filename or '')
1019 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1020
1021 @staticmethod
1022 def _outtmpl_expandpath(outtmpl):
1023 # expand_path translates '%%' into '%' and '$$' into '$'
1024 # correspondingly that is not what we want since we need to keep
1025 # '%%' intact for template dict substitution step. Working around
1026 # with boundary-alike separator hack.
1027 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1028 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1029
1030 # outtmpl should be expand_path'ed before template dict substitution
1031 # because meta fields may contain env variables we don't want to
1032 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1033 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1034 return expand_path(outtmpl).replace(sep, '')
1035
1036 @staticmethod
1037 def escape_outtmpl(outtmpl):
1038 ''' Escape any remaining strings like %s, %abc% etc. '''
1039 return re.sub(
1040 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1041 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1042 outtmpl)
1043
1044 @classmethod
1045 def validate_outtmpl(cls, outtmpl):
1046 ''' @return None or Exception object '''
1047 outtmpl = re.sub(
1048 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1049 lambda mobj: f'{mobj.group(0)[:-1]}s',
1050 cls._outtmpl_expandpath(outtmpl))
1051 try:
1052 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1053 return None
1054 except ValueError as err:
1055 return err
1056
1057 @staticmethod
1058 def _copy_infodict(info_dict):
1059 info_dict = dict(info_dict)
1060 info_dict.pop('__postprocessors', None)
1061 info_dict.pop('__pending_error', None)
1062 return info_dict
1063
1064 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1065 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1066 @param sanitize Whether to sanitize the output as a filename.
1067 For backward compatibility, a function can also be passed
1068 """
1069
1070 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1071
1072 info_dict = self._copy_infodict(info_dict)
1073 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1074 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1075 if info_dict.get('duration', None) is not None
1076 else None)
1077 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1078 info_dict['video_autonumber'] = self._num_videos
1079 if info_dict.get('resolution') is None:
1080 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1081
1082 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1083 # of %(field)s to %(field)0Nd for backward compatibility
1084 field_size_compat_map = {
1085 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1086 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1087 'autonumber': self.params.get('autonumber_size') or 5,
1088 }
1089
1090 TMPL_DICT = {}
1091 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1092 MATH_FUNCTIONS = {
1093 '+': float.__add__,
1094 '-': float.__sub__,
1095 }
1096 # Field is of the form key1.key2...
1097 # where keys (except first) can be string, int or slice
1098 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1099 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1100 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1101 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1102 (?P<negate>-)?
1103 (?P<fields>{FIELD_RE})
1104 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1105 (?:>(?P<strf_format>.+?))?
1106 (?P<remaining>
1107 (?P<alternate>(?<!\\),[^|&)]+)?
1108 (?:&(?P<replacement>.*?))?
1109 (?:\|(?P<default>.*?))?
1110 )$''')
1111
1112 def _traverse_infodict(k):
1113 k = k.split('.')
1114 if k[0] == '':
1115 k.pop(0)
1116 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1117
1118 def get_value(mdict):
1119 # Object traversal
1120 value = _traverse_infodict(mdict['fields'])
1121 # Negative
1122 if mdict['negate']:
1123 value = float_or_none(value)
1124 if value is not None:
1125 value *= -1
1126 # Do maths
1127 offset_key = mdict['maths']
1128 if offset_key:
1129 value = float_or_none(value)
1130 operator = None
1131 while offset_key:
1132 item = re.match(
1133 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1134 offset_key).group(0)
1135 offset_key = offset_key[len(item):]
1136 if operator is None:
1137 operator = MATH_FUNCTIONS[item]
1138 continue
1139 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1140 offset = float_or_none(item)
1141 if offset is None:
1142 offset = float_or_none(_traverse_infodict(item))
1143 try:
1144 value = operator(value, multiplier * offset)
1145 except (TypeError, ZeroDivisionError):
1146 return None
1147 operator = None
1148 # Datetime formatting
1149 if mdict['strf_format']:
1150 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1151
1152 return value
1153
1154 na = self.params.get('outtmpl_na_placeholder', 'NA')
1155
1156 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1157 return sanitize_filename(str(value), restricted=restricted, is_id=(
1158 bool(re.search(r'(^|[_.])id(\.|$)', key))
1159 if 'filename-sanitization' in self.params['compat_opts']
1160 else NO_DEFAULT))
1161
1162 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1163 sanitize = bool(sanitize)
1164
1165 def _dumpjson_default(obj):
1166 if isinstance(obj, (set, LazyList)):
1167 return list(obj)
1168 return repr(obj)
1169
1170 def create_key(outer_mobj):
1171 if not outer_mobj.group('has_key'):
1172 return outer_mobj.group(0)
1173 key = outer_mobj.group('key')
1174 mobj = re.match(INTERNAL_FORMAT_RE, key)
1175 initial_field = mobj.group('fields') if mobj else ''
1176 value, replacement, default = None, None, na
1177 while mobj:
1178 mobj = mobj.groupdict()
1179 default = mobj['default'] if mobj['default'] is not None else default
1180 value = get_value(mobj)
1181 replacement = mobj['replacement']
1182 if value is None and mobj['alternate']:
1183 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1184 else:
1185 break
1186
1187 fmt = outer_mobj.group('format')
1188 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1189 fmt = f'0{field_size_compat_map[key]:d}d'
1190
1191 value = default if value is None else value if replacement is None else replacement
1192
1193 flags = outer_mobj.group('conversion') or ''
1194 str_fmt = f'{fmt[:-1]}s'
1195 if fmt[-1] == 'l': # list
1196 delim = '\n' if '#' in flags else ', '
1197 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1198 elif fmt[-1] == 'j': # json
1199 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1200 elif fmt[-1] == 'q': # quoted
1201 value = map(str, variadic(value) if '#' in flags else [value])
1202 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1203 elif fmt[-1] == 'B': # bytes
1204 value = f'%{str_fmt}'.encode() % str(value).encode()
1205 value, fmt = value.decode('utf-8', 'ignore'), 's'
1206 elif fmt[-1] == 'U': # unicode normalized
1207 value, fmt = unicodedata.normalize(
1208 # "+" = compatibility equivalence, "#" = NFD
1209 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1210 value), str_fmt
1211 elif fmt[-1] == 'D': # decimal suffix
1212 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1213 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1214 factor=1024 if '#' in flags else 1000)
1215 elif fmt[-1] == 'S': # filename sanitization
1216 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1217 elif fmt[-1] == 'c':
1218 if value:
1219 value = str(value)[0]
1220 else:
1221 fmt = str_fmt
1222 elif fmt[-1] not in 'rs': # numeric
1223 value = float_or_none(value)
1224 if value is None:
1225 value, fmt = default, 's'
1226
1227 if sanitize:
1228 if fmt[-1] == 'r':
1229 # If value is an object, sanitize might convert it to a string
1230 # So we convert it to repr first
1231 value, fmt = repr(value), str_fmt
1232 if fmt[-1] in 'csr':
1233 value = sanitizer(initial_field, value)
1234
1235 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1236 TMPL_DICT[key] = value
1237 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1238
1239 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1240
1241 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1242 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1243 return self.escape_outtmpl(outtmpl) % info_dict
1244
1245 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1246 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1247 if outtmpl is None:
1248 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1249 try:
1250 outtmpl = self._outtmpl_expandpath(outtmpl)
1251 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1252 if not filename:
1253 return None
1254
1255 if tmpl_type in ('', 'temp'):
1256 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1257 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1258 filename = replace_extension(filename, ext, final_ext)
1259 elif tmpl_type:
1260 force_ext = OUTTMPL_TYPES[tmpl_type]
1261 if force_ext:
1262 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1263
1264 # https://github.com/blackjack4494/youtube-dlc/issues/85
1265 trim_file_name = self.params.get('trim_file_name', False)
1266 if trim_file_name:
1267 no_ext, *ext = filename.rsplit('.', 2)
1268 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1269
1270 return filename
1271 except ValueError as err:
1272 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1273 return None
1274
1275 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1276 """Generate the output filename"""
1277 if outtmpl:
1278 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1279 dir_type = None
1280 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1281 if not filename and dir_type not in ('', 'temp'):
1282 return ''
1283
1284 if warn:
1285 if not self.params.get('paths'):
1286 pass
1287 elif filename == '-':
1288 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1289 elif os.path.isabs(filename):
1290 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1291 if filename == '-' or not filename:
1292 return filename
1293
1294 return self.get_output_path(dir_type, filename)
1295
1296 def _match_entry(self, info_dict, incomplete=False, silent=False):
1297 """ Returns None if the file should be downloaded """
1298
1299 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1300
1301 def check_filter():
1302 if 'title' in info_dict:
1303 # This can happen when we're just evaluating the playlist
1304 title = info_dict['title']
1305 matchtitle = self.params.get('matchtitle', False)
1306 if matchtitle:
1307 if not re.search(matchtitle, title, re.IGNORECASE):
1308 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1309 rejecttitle = self.params.get('rejecttitle', False)
1310 if rejecttitle:
1311 if re.search(rejecttitle, title, re.IGNORECASE):
1312 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1313 date = info_dict.get('upload_date')
1314 if date is not None:
1315 dateRange = self.params.get('daterange', DateRange())
1316 if date not in dateRange:
1317 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1318 view_count = info_dict.get('view_count')
1319 if view_count is not None:
1320 min_views = self.params.get('min_views')
1321 if min_views is not None and view_count < min_views:
1322 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1323 max_views = self.params.get('max_views')
1324 if max_views is not None and view_count > max_views:
1325 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1326 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1327 return 'Skipping "%s" because it is age restricted' % video_title
1328
1329 match_filter = self.params.get('match_filter')
1330 if match_filter is not None:
1331 try:
1332 ret = match_filter(info_dict, incomplete=incomplete)
1333 except TypeError:
1334 # For backward compatibility
1335 ret = None if incomplete else match_filter(info_dict)
1336 if ret is NO_DEFAULT:
1337 while True:
1338 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1339 reply = input(self._format_screen(
1340 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1341 if reply in {'y', ''}:
1342 return None
1343 elif reply == 'n':
1344 return f'Skipping {video_title}'
1345 elif ret is not None:
1346 return ret
1347 return None
1348
1349 if self.in_download_archive(info_dict):
1350 reason = '%s has already been recorded in the archive' % video_title
1351 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1352 else:
1353 reason = check_filter()
1354 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1355 if reason is not None:
1356 if not silent:
1357 self.to_screen('[download] ' + reason)
1358 if self.params.get(break_opt, False):
1359 raise break_err()
1360 return reason
1361
1362 @staticmethod
1363 def add_extra_info(info_dict, extra_info):
1364 '''Set the keys from extra_info in info dict if they are missing'''
1365 for key, value in extra_info.items():
1366 info_dict.setdefault(key, value)
1367
1368 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1369 process=True, force_generic_extractor=False):
1370 """
1371 Return a list with a dictionary for each video extracted.
1372
1373 Arguments:
1374 url -- URL to extract
1375
1376 Keyword arguments:
1377 download -- whether to download videos during extraction
1378 ie_key -- extractor key hint
1379 extra_info -- dictionary containing the extra values to add to each result
1380 process -- whether to resolve all unresolved references (URLs, playlist items),
1381 must be True for download to work.
1382 force_generic_extractor -- force using the generic extractor
1383 """
1384
1385 if extra_info is None:
1386 extra_info = {}
1387
1388 if not ie_key and force_generic_extractor:
1389 ie_key = 'Generic'
1390
1391 if ie_key:
1392 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1393 else:
1394 ies = self._ies
1395
1396 for ie_key, ie in ies.items():
1397 if not ie.suitable(url):
1398 continue
1399
1400 if not ie.working():
1401 self.report_warning('The program functionality for this site has been marked as broken, '
1402 'and will probably not work.')
1403
1404 temp_id = ie.get_temp_id(url)
1405 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1406 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1407 if self.params.get('break_on_existing', False):
1408 raise ExistingVideoReached()
1409 break
1410 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1411 else:
1412 self.report_error('no suitable InfoExtractor for URL %s' % url)
1413
1414 def _handle_extraction_exceptions(func):
1415 @functools.wraps(func)
1416 def wrapper(self, *args, **kwargs):
1417 while True:
1418 try:
1419 return func(self, *args, **kwargs)
1420 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1421 raise
1422 except ReExtractInfo as e:
1423 if e.expected:
1424 self.to_screen(f'{e}; Re-extracting data')
1425 else:
1426 self.to_stderr('\r')
1427 self.report_warning(f'{e}; Re-extracting data')
1428 continue
1429 except GeoRestrictedError as e:
1430 msg = e.msg
1431 if e.countries:
1432 msg += '\nThis video is available in %s.' % ', '.join(
1433 map(ISO3166Utils.short2full, e.countries))
1434 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1435 self.report_error(msg)
1436 except ExtractorError as e: # An error we somewhat expected
1437 self.report_error(str(e), e.format_traceback())
1438 except Exception as e:
1439 if self.params.get('ignoreerrors'):
1440 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1441 else:
1442 raise
1443 break
1444 return wrapper
1445
1446 def _wait_for_video(self, ie_result):
1447 if (not self.params.get('wait_for_video')
1448 or ie_result.get('_type', 'video') != 'video'
1449 or ie_result.get('formats') or ie_result.get('url')):
1450 return
1451
1452 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1453 last_msg = ''
1454
1455 def progress(msg):
1456 nonlocal last_msg
1457 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1458 last_msg = msg
1459
1460 min_wait, max_wait = self.params.get('wait_for_video')
1461 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1462 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1463 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1464 self.report_warning('Release time of video is not known')
1465 elif (diff or 0) <= 0:
1466 self.report_warning('Video should already be available according to extracted info')
1467 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1468 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1469
1470 wait_till = time.time() + diff
1471 try:
1472 while True:
1473 diff = wait_till - time.time()
1474 if diff <= 0:
1475 progress('')
1476 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1477 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1478 time.sleep(1)
1479 except KeyboardInterrupt:
1480 progress('')
1481 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1482 except BaseException as e:
1483 if not isinstance(e, ReExtractInfo):
1484 self.to_screen('')
1485 raise
1486
1487 @_handle_extraction_exceptions
1488 def __extract_info(self, url, ie, download, extra_info, process):
1489 ie_result = ie.extract(url)
1490 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1491 return
1492 if isinstance(ie_result, list):
1493 # Backwards compatibility: old IE result format
1494 ie_result = {
1495 '_type': 'compat_list',
1496 'entries': ie_result,
1497 }
1498 if extra_info.get('original_url'):
1499 ie_result.setdefault('original_url', extra_info['original_url'])
1500 self.add_default_extra_info(ie_result, ie, url)
1501 if process:
1502 self._wait_for_video(ie_result)
1503 return self.process_ie_result(ie_result, download, extra_info)
1504 else:
1505 return ie_result
1506
1507 def add_default_extra_info(self, ie_result, ie, url):
1508 if url is not None:
1509 self.add_extra_info(ie_result, {
1510 'webpage_url': url,
1511 'original_url': url,
1512 })
1513 webpage_url = ie_result.get('webpage_url')
1514 if webpage_url:
1515 self.add_extra_info(ie_result, {
1516 'webpage_url_basename': url_basename(webpage_url),
1517 'webpage_url_domain': get_domain(webpage_url),
1518 })
1519 if ie is not None:
1520 self.add_extra_info(ie_result, {
1521 'extractor': ie.IE_NAME,
1522 'extractor_key': ie.ie_key(),
1523 })
1524
1525 def process_ie_result(self, ie_result, download=True, extra_info=None):
1526 """
1527 Take the result of the ie(may be modified) and resolve all unresolved
1528 references (URLs, playlist items).
1529
1530 It will also download the videos if 'download'.
1531 Returns the resolved ie_result.
1532 """
1533 if extra_info is None:
1534 extra_info = {}
1535 result_type = ie_result.get('_type', 'video')
1536
1537 if result_type in ('url', 'url_transparent'):
1538 ie_result['url'] = sanitize_url(ie_result['url'])
1539 if ie_result.get('original_url'):
1540 extra_info.setdefault('original_url', ie_result['original_url'])
1541
1542 extract_flat = self.params.get('extract_flat', False)
1543 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1544 or extract_flat is True):
1545 info_copy = ie_result.copy()
1546 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1547 if ie and not ie_result.get('id'):
1548 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1549 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1550 self.add_extra_info(info_copy, extra_info)
1551 info_copy, _ = self.pre_process(info_copy)
1552 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1553 self._raise_pending_errors(info_copy)
1554 if self.params.get('force_write_download_archive', False):
1555 self.record_download_archive(info_copy)
1556 return ie_result
1557
1558 if result_type == 'video':
1559 self.add_extra_info(ie_result, extra_info)
1560 ie_result = self.process_video_result(ie_result, download=download)
1561 self._raise_pending_errors(ie_result)
1562 additional_urls = (ie_result or {}).get('additional_urls')
1563 if additional_urls:
1564 # TODO: Improve MetadataParserPP to allow setting a list
1565 if isinstance(additional_urls, str):
1566 additional_urls = [additional_urls]
1567 self.to_screen(
1568 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1569 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1570 ie_result['additional_entries'] = [
1571 self.extract_info(
1572 url, download, extra_info=extra_info,
1573 force_generic_extractor=self.params.get('force_generic_extractor'))
1574 for url in additional_urls
1575 ]
1576 return ie_result
1577 elif result_type == 'url':
1578 # We have to add extra_info to the results because it may be
1579 # contained in a playlist
1580 return self.extract_info(
1581 ie_result['url'], download,
1582 ie_key=ie_result.get('ie_key'),
1583 extra_info=extra_info)
1584 elif result_type == 'url_transparent':
1585 # Use the information from the embedding page
1586 info = self.extract_info(
1587 ie_result['url'], ie_key=ie_result.get('ie_key'),
1588 extra_info=extra_info, download=False, process=False)
1589
1590 # extract_info may return None when ignoreerrors is enabled and
1591 # extraction failed with an error, don't crash and return early
1592 # in this case
1593 if not info:
1594 return info
1595
1596 exempted_fields = {'_type', 'url', 'ie_key'}
1597 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1598 # For video clips, the id etc of the clip extractor should be used
1599 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1600
1601 new_result = info.copy()
1602 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1603
1604 # Extracted info may not be a video result (i.e.
1605 # info.get('_type', 'video') != video) but rather an url or
1606 # url_transparent. In such cases outer metadata (from ie_result)
1607 # should be propagated to inner one (info). For this to happen
1608 # _type of info should be overridden with url_transparent. This
1609 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1610 if new_result.get('_type') == 'url':
1611 new_result['_type'] = 'url_transparent'
1612
1613 return self.process_ie_result(
1614 new_result, download=download, extra_info=extra_info)
1615 elif result_type in ('playlist', 'multi_video'):
1616 # Protect from infinite recursion due to recursively nested playlists
1617 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1618 webpage_url = ie_result['webpage_url']
1619 if webpage_url in self._playlist_urls:
1620 self.to_screen(
1621 '[download] Skipping already downloaded playlist: %s'
1622 % ie_result.get('title') or ie_result.get('id'))
1623 return
1624
1625 self._playlist_level += 1
1626 self._playlist_urls.add(webpage_url)
1627 self._fill_common_fields(ie_result, False)
1628 self._sanitize_thumbnails(ie_result)
1629 try:
1630 return self.__process_playlist(ie_result, download)
1631 finally:
1632 self._playlist_level -= 1
1633 if not self._playlist_level:
1634 self._playlist_urls.clear()
1635 elif result_type == 'compat_list':
1636 self.report_warning(
1637 'Extractor %s returned a compat_list result. '
1638 'It needs to be updated.' % ie_result.get('extractor'))
1639
1640 def _fixup(r):
1641 self.add_extra_info(r, {
1642 'extractor': ie_result['extractor'],
1643 'webpage_url': ie_result['webpage_url'],
1644 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1645 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1646 'extractor_key': ie_result['extractor_key'],
1647 })
1648 return r
1649 ie_result['entries'] = [
1650 self.process_ie_result(_fixup(r), download, extra_info)
1651 for r in ie_result['entries']
1652 ]
1653 return ie_result
1654 else:
1655 raise Exception('Invalid result type: %s' % result_type)
1656
1657 def _ensure_dir_exists(self, path):
1658 return make_dir(path, self.report_error)
1659
1660 @staticmethod
1661 def _playlist_infodict(ie_result, **kwargs):
1662 return {
1663 **ie_result,
1664 'playlist': ie_result.get('title') or ie_result.get('id'),
1665 'playlist_id': ie_result.get('id'),
1666 'playlist_title': ie_result.get('title'),
1667 'playlist_uploader': ie_result.get('uploader'),
1668 'playlist_uploader_id': ie_result.get('uploader_id'),
1669 'playlist_index': 0,
1670 **kwargs,
1671 }
1672
1673 def __process_playlist(self, ie_result, download):
1674 """Process each entry in the playlist"""
1675 title = ie_result.get('title') or ie_result.get('id') or '<Untitled>'
1676 self.to_screen(f'[download] Downloading playlist: {title}')
1677
1678 all_entries = PlaylistEntries(self, ie_result)
1679 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1680
1681 lazy = self.params.get('lazy_playlist')
1682 if lazy:
1683 resolved_entries, n_entries = [], 'N/A'
1684 ie_result['requested_entries'], ie_result['entries'] = None, None
1685 else:
1686 entries = resolved_entries = list(entries)
1687 n_entries = len(resolved_entries)
1688 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1689 if not ie_result.get('playlist_count'):
1690 # Better to do this after potentially exhausting entries
1691 ie_result['playlist_count'] = all_entries.get_full_count()
1692
1693 _infojson_written = False
1694 write_playlist_files = self.params.get('allow_playlist_files', True)
1695 if write_playlist_files and self.params.get('list_thumbnails'):
1696 self.list_thumbnails(ie_result)
1697 if write_playlist_files and not self.params.get('simulate'):
1698 ie_copy = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1699 _infojson_written = self._write_info_json(
1700 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1701 if _infojson_written is None:
1702 return
1703 if self._write_description('playlist', ie_result,
1704 self.prepare_filename(ie_copy, 'pl_description')) is None:
1705 return
1706 # TODO: This should be passed to ThumbnailsConvertor if necessary
1707 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1708
1709 if lazy:
1710 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1711 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1712 elif self.params.get('playlistreverse'):
1713 entries.reverse()
1714 elif self.params.get('playlistrandom'):
1715 random.shuffle(entries)
1716
1717 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} videos'
1718 f'{format_field(ie_result, "playlist_count", " of %s")}')
1719
1720 failures = 0
1721 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1722 for i, (playlist_index, entry) in enumerate(entries):
1723 if lazy:
1724 resolved_entries.append((playlist_index, entry))
1725
1726 # TODO: Add auto-generated fields
1727 if not entry or self._match_entry(entry, incomplete=True) is not None:
1728 continue
1729
1730 self.to_screen('[download] Downloading video %s of %s' % (
1731 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1732
1733 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1734 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1735 playlist_index = ie_result['requested_entries'][i]
1736
1737 entry_result = self.__process_iterable_entry(entry, download, {
1738 'n_entries': int_or_none(n_entries),
1739 '__last_playlist_index': max(ie_result['requested_entries'] or (0, 0)),
1740 'playlist_count': ie_result.get('playlist_count'),
1741 'playlist_index': playlist_index,
1742 'playlist_autonumber': i + 1,
1743 'playlist': title,
1744 'playlist_id': ie_result.get('id'),
1745 'playlist_title': ie_result.get('title'),
1746 'playlist_uploader': ie_result.get('uploader'),
1747 'playlist_uploader_id': ie_result.get('uploader_id'),
1748 'extractor': ie_result['extractor'],
1749 'webpage_url': ie_result['webpage_url'],
1750 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1751 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1752 'extractor_key': ie_result['extractor_key'],
1753 })
1754 if not entry_result:
1755 failures += 1
1756 if failures >= max_failures:
1757 self.report_error(
1758 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1759 break
1760 resolved_entries[i] = (playlist_index, entry_result)
1761
1762 # Update with processed data
1763 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1764
1765 # Write the updated info to json
1766 if _infojson_written is True and self._write_info_json(
1767 'updated playlist', ie_result,
1768 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1769 return
1770
1771 ie_result = self.run_all_pps('playlist', ie_result)
1772 self.to_screen(f'[download] Finished downloading playlist: {title}')
1773 return ie_result
1774
1775 @_handle_extraction_exceptions
1776 def __process_iterable_entry(self, entry, download, extra_info):
1777 return self.process_ie_result(
1778 entry, download=download, extra_info=extra_info)
1779
1780 def _build_format_filter(self, filter_spec):
1781 " Returns a function to filter the formats according to the filter_spec "
1782
1783 OPERATORS = {
1784 '<': operator.lt,
1785 '<=': operator.le,
1786 '>': operator.gt,
1787 '>=': operator.ge,
1788 '=': operator.eq,
1789 '!=': operator.ne,
1790 }
1791 operator_rex = re.compile(r'''(?x)\s*
1792 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1793 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1794 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1795 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1796 m = operator_rex.fullmatch(filter_spec)
1797 if m:
1798 try:
1799 comparison_value = int(m.group('value'))
1800 except ValueError:
1801 comparison_value = parse_filesize(m.group('value'))
1802 if comparison_value is None:
1803 comparison_value = parse_filesize(m.group('value') + 'B')
1804 if comparison_value is None:
1805 raise ValueError(
1806 'Invalid value %r in format specification %r' % (
1807 m.group('value'), filter_spec))
1808 op = OPERATORS[m.group('op')]
1809
1810 if not m:
1811 STR_OPERATORS = {
1812 '=': operator.eq,
1813 '^=': lambda attr, value: attr.startswith(value),
1814 '$=': lambda attr, value: attr.endswith(value),
1815 '*=': lambda attr, value: value in attr,
1816 '~=': lambda attr, value: value.search(attr) is not None
1817 }
1818 str_operator_rex = re.compile(r'''(?x)\s*
1819 (?P<key>[a-zA-Z0-9._-]+)\s*
1820 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1821 (?P<quote>["'])?
1822 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1823 (?(quote)(?P=quote))\s*
1824 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1825 m = str_operator_rex.fullmatch(filter_spec)
1826 if m:
1827 if m.group('op') == '~=':
1828 comparison_value = re.compile(m.group('value'))
1829 else:
1830 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1831 str_op = STR_OPERATORS[m.group('op')]
1832 if m.group('negation'):
1833 op = lambda attr, value: not str_op(attr, value)
1834 else:
1835 op = str_op
1836
1837 if not m:
1838 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1839
1840 def _filter(f):
1841 actual_value = f.get(m.group('key'))
1842 if actual_value is None:
1843 return m.group('none_inclusive')
1844 return op(actual_value, comparison_value)
1845 return _filter
1846
1847 def _check_formats(self, formats):
1848 for f in formats:
1849 self.to_screen('[info] Testing format %s' % f['format_id'])
1850 path = self.get_output_path('temp')
1851 if not self._ensure_dir_exists(f'{path}/'):
1852 continue
1853 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1854 temp_file.close()
1855 try:
1856 success, _ = self.dl(temp_file.name, f, test=True)
1857 except (DownloadError, OSError, ValueError) + network_exceptions:
1858 success = False
1859 finally:
1860 if os.path.exists(temp_file.name):
1861 try:
1862 os.remove(temp_file.name)
1863 except OSError:
1864 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1865 if success:
1866 yield f
1867 else:
1868 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1869
1870 def _default_format_spec(self, info_dict, download=True):
1871
1872 def can_merge():
1873 merger = FFmpegMergerPP(self)
1874 return merger.available and merger.can_merge()
1875
1876 prefer_best = (
1877 not self.params.get('simulate')
1878 and download
1879 and (
1880 not can_merge()
1881 or info_dict.get('is_live') and not self.params.get('live_from_start')
1882 or self.params['outtmpl']['default'] == '-'))
1883 compat = (
1884 prefer_best
1885 or self.params.get('allow_multiple_audio_streams', False)
1886 or 'format-spec' in self.params['compat_opts'])
1887
1888 return (
1889 'best/bestvideo+bestaudio' if prefer_best
1890 else 'bestvideo*+bestaudio/best' if not compat
1891 else 'bestvideo+bestaudio/best')
1892
1893 def build_format_selector(self, format_spec):
1894 def syntax_error(note, start):
1895 message = (
1896 'Invalid format specification: '
1897 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1898 return SyntaxError(message)
1899
1900 PICKFIRST = 'PICKFIRST'
1901 MERGE = 'MERGE'
1902 SINGLE = 'SINGLE'
1903 GROUP = 'GROUP'
1904 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1905
1906 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1907 'video': self.params.get('allow_multiple_video_streams', False)}
1908
1909 check_formats = self.params.get('check_formats') == 'selected'
1910
1911 def _parse_filter(tokens):
1912 filter_parts = []
1913 for type, string, start, _, _ in tokens:
1914 if type == tokenize.OP and string == ']':
1915 return ''.join(filter_parts)
1916 else:
1917 filter_parts.append(string)
1918
1919 def _remove_unused_ops(tokens):
1920 # Remove operators that we don't use and join them with the surrounding strings
1921 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1922 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1923 last_string, last_start, last_end, last_line = None, None, None, None
1924 for type, string, start, end, line in tokens:
1925 if type == tokenize.OP and string == '[':
1926 if last_string:
1927 yield tokenize.NAME, last_string, last_start, last_end, last_line
1928 last_string = None
1929 yield type, string, start, end, line
1930 # everything inside brackets will be handled by _parse_filter
1931 for type, string, start, end, line in tokens:
1932 yield type, string, start, end, line
1933 if type == tokenize.OP and string == ']':
1934 break
1935 elif type == tokenize.OP and string in ALLOWED_OPS:
1936 if last_string:
1937 yield tokenize.NAME, last_string, last_start, last_end, last_line
1938 last_string = None
1939 yield type, string, start, end, line
1940 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1941 if not last_string:
1942 last_string = string
1943 last_start = start
1944 last_end = end
1945 else:
1946 last_string += string
1947 if last_string:
1948 yield tokenize.NAME, last_string, last_start, last_end, last_line
1949
1950 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1951 selectors = []
1952 current_selector = None
1953 for type, string, start, _, _ in tokens:
1954 # ENCODING is only defined in python 3.x
1955 if type == getattr(tokenize, 'ENCODING', None):
1956 continue
1957 elif type in [tokenize.NAME, tokenize.NUMBER]:
1958 current_selector = FormatSelector(SINGLE, string, [])
1959 elif type == tokenize.OP:
1960 if string == ')':
1961 if not inside_group:
1962 # ')' will be handled by the parentheses group
1963 tokens.restore_last_token()
1964 break
1965 elif inside_merge and string in ['/', ',']:
1966 tokens.restore_last_token()
1967 break
1968 elif inside_choice and string == ',':
1969 tokens.restore_last_token()
1970 break
1971 elif string == ',':
1972 if not current_selector:
1973 raise syntax_error('"," must follow a format selector', start)
1974 selectors.append(current_selector)
1975 current_selector = None
1976 elif string == '/':
1977 if not current_selector:
1978 raise syntax_error('"/" must follow a format selector', start)
1979 first_choice = current_selector
1980 second_choice = _parse_format_selection(tokens, inside_choice=True)
1981 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1982 elif string == '[':
1983 if not current_selector:
1984 current_selector = FormatSelector(SINGLE, 'best', [])
1985 format_filter = _parse_filter(tokens)
1986 current_selector.filters.append(format_filter)
1987 elif string == '(':
1988 if current_selector:
1989 raise syntax_error('Unexpected "("', start)
1990 group = _parse_format_selection(tokens, inside_group=True)
1991 current_selector = FormatSelector(GROUP, group, [])
1992 elif string == '+':
1993 if not current_selector:
1994 raise syntax_error('Unexpected "+"', start)
1995 selector_1 = current_selector
1996 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1997 if not selector_2:
1998 raise syntax_error('Expected a selector', start)
1999 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2000 else:
2001 raise syntax_error(f'Operator not recognized: "{string}"', start)
2002 elif type == tokenize.ENDMARKER:
2003 break
2004 if current_selector:
2005 selectors.append(current_selector)
2006 return selectors
2007
2008 def _merge(formats_pair):
2009 format_1, format_2 = formats_pair
2010
2011 formats_info = []
2012 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2013 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2014
2015 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2016 get_no_more = {'video': False, 'audio': False}
2017 for (i, fmt_info) in enumerate(formats_info):
2018 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2019 formats_info.pop(i)
2020 continue
2021 for aud_vid in ['audio', 'video']:
2022 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2023 if get_no_more[aud_vid]:
2024 formats_info.pop(i)
2025 break
2026 get_no_more[aud_vid] = True
2027
2028 if len(formats_info) == 1:
2029 return formats_info[0]
2030
2031 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2032 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2033
2034 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2035 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2036
2037 output_ext = self.params.get('merge_output_format')
2038 if not output_ext:
2039 if the_only_video:
2040 output_ext = the_only_video['ext']
2041 elif the_only_audio and not video_fmts:
2042 output_ext = the_only_audio['ext']
2043 else:
2044 output_ext = 'mkv'
2045
2046 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2047
2048 new_dict = {
2049 'requested_formats': formats_info,
2050 'format': '+'.join(filtered('format')),
2051 'format_id': '+'.join(filtered('format_id')),
2052 'ext': output_ext,
2053 'protocol': '+'.join(map(determine_protocol, formats_info)),
2054 'language': '+'.join(orderedSet(filtered('language'))) or None,
2055 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2056 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2057 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2058 }
2059
2060 if the_only_video:
2061 new_dict.update({
2062 'width': the_only_video.get('width'),
2063 'height': the_only_video.get('height'),
2064 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2065 'fps': the_only_video.get('fps'),
2066 'dynamic_range': the_only_video.get('dynamic_range'),
2067 'vcodec': the_only_video.get('vcodec'),
2068 'vbr': the_only_video.get('vbr'),
2069 'stretched_ratio': the_only_video.get('stretched_ratio'),
2070 })
2071
2072 if the_only_audio:
2073 new_dict.update({
2074 'acodec': the_only_audio.get('acodec'),
2075 'abr': the_only_audio.get('abr'),
2076 'asr': the_only_audio.get('asr'),
2077 })
2078
2079 return new_dict
2080
2081 def _check_formats(formats):
2082 if not check_formats:
2083 yield from formats
2084 return
2085 yield from self._check_formats(formats)
2086
2087 def _build_selector_function(selector):
2088 if isinstance(selector, list): # ,
2089 fs = [_build_selector_function(s) for s in selector]
2090
2091 def selector_function(ctx):
2092 for f in fs:
2093 yield from f(ctx)
2094 return selector_function
2095
2096 elif selector.type == GROUP: # ()
2097 selector_function = _build_selector_function(selector.selector)
2098
2099 elif selector.type == PICKFIRST: # /
2100 fs = [_build_selector_function(s) for s in selector.selector]
2101
2102 def selector_function(ctx):
2103 for f in fs:
2104 picked_formats = list(f(ctx))
2105 if picked_formats:
2106 return picked_formats
2107 return []
2108
2109 elif selector.type == MERGE: # +
2110 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2111
2112 def selector_function(ctx):
2113 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2114 yield _merge(pair)
2115
2116 elif selector.type == SINGLE: # atom
2117 format_spec = selector.selector or 'best'
2118
2119 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2120 if format_spec == 'all':
2121 def selector_function(ctx):
2122 yield from _check_formats(ctx['formats'][::-1])
2123 elif format_spec == 'mergeall':
2124 def selector_function(ctx):
2125 formats = list(_check_formats(
2126 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2127 if not formats:
2128 return
2129 merged_format = formats[-1]
2130 for f in formats[-2::-1]:
2131 merged_format = _merge((merged_format, f))
2132 yield merged_format
2133
2134 else:
2135 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2136 mobj = re.match(
2137 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2138 format_spec)
2139 if mobj is not None:
2140 format_idx = int_or_none(mobj.group('n'), default=1)
2141 format_reverse = mobj.group('bw')[0] == 'b'
2142 format_type = (mobj.group('type') or [None])[0]
2143 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2144 format_modified = mobj.group('mod') is not None
2145
2146 format_fallback = not format_type and not format_modified # for b, w
2147 _filter_f = (
2148 (lambda f: f.get('%scodec' % format_type) != 'none')
2149 if format_type and format_modified # bv*, ba*, wv*, wa*
2150 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2151 if format_type # bv, ba, wv, wa
2152 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2153 if not format_modified # b, w
2154 else lambda f: True) # b*, w*
2155 filter_f = lambda f: _filter_f(f) and (
2156 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2157 else:
2158 if format_spec in self._format_selection_exts['audio']:
2159 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2160 elif format_spec in self._format_selection_exts['video']:
2161 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2162 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2163 elif format_spec in self._format_selection_exts['storyboards']:
2164 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2165 else:
2166 filter_f = lambda f: f.get('format_id') == format_spec # id
2167
2168 def selector_function(ctx):
2169 formats = list(ctx['formats'])
2170 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2171 if not matches:
2172 if format_fallback and ctx['incomplete_formats']:
2173 # for extractors with incomplete formats (audio only (soundcloud)
2174 # or video only (imgur)) best/worst will fallback to
2175 # best/worst {video,audio}-only format
2176 matches = formats
2177 elif seperate_fallback and not ctx['has_merged_format']:
2178 # for compatibility with youtube-dl when there is no pre-merged format
2179 matches = list(filter(seperate_fallback, formats))
2180 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2181 try:
2182 yield matches[format_idx - 1]
2183 except LazyList.IndexError:
2184 return
2185
2186 filters = [self._build_format_filter(f) for f in selector.filters]
2187
2188 def final_selector(ctx):
2189 ctx_copy = dict(ctx)
2190 for _filter in filters:
2191 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2192 return selector_function(ctx_copy)
2193 return final_selector
2194
2195 stream = io.BytesIO(format_spec.encode())
2196 try:
2197 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2198 except tokenize.TokenError:
2199 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2200
2201 class TokenIterator:
2202 def __init__(self, tokens):
2203 self.tokens = tokens
2204 self.counter = 0
2205
2206 def __iter__(self):
2207 return self
2208
2209 def __next__(self):
2210 if self.counter >= len(self.tokens):
2211 raise StopIteration()
2212 value = self.tokens[self.counter]
2213 self.counter += 1
2214 return value
2215
2216 next = __next__
2217
2218 def restore_last_token(self):
2219 self.counter -= 1
2220
2221 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2222 return _build_selector_function(parsed_selector)
2223
2224 def _calc_headers(self, info_dict):
2225 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2226
2227 cookies = self._calc_cookies(info_dict['url'])
2228 if cookies:
2229 res['Cookie'] = cookies
2230
2231 if 'X-Forwarded-For' not in res:
2232 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2233 if x_forwarded_for_ip:
2234 res['X-Forwarded-For'] = x_forwarded_for_ip
2235
2236 return res
2237
2238 def _calc_cookies(self, url):
2239 pr = sanitized_Request(url)
2240 self.cookiejar.add_cookie_header(pr)
2241 return pr.get_header('Cookie')
2242
2243 def _sort_thumbnails(self, thumbnails):
2244 thumbnails.sort(key=lambda t: (
2245 t.get('preference') if t.get('preference') is not None else -1,
2246 t.get('width') if t.get('width') is not None else -1,
2247 t.get('height') if t.get('height') is not None else -1,
2248 t.get('id') if t.get('id') is not None else '',
2249 t.get('url')))
2250
2251 def _sanitize_thumbnails(self, info_dict):
2252 thumbnails = info_dict.get('thumbnails')
2253 if thumbnails is None:
2254 thumbnail = info_dict.get('thumbnail')
2255 if thumbnail:
2256 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2257 if not thumbnails:
2258 return
2259
2260 def check_thumbnails(thumbnails):
2261 for t in thumbnails:
2262 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2263 try:
2264 self.urlopen(HEADRequest(t['url']))
2265 except network_exceptions as err:
2266 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2267 continue
2268 yield t
2269
2270 self._sort_thumbnails(thumbnails)
2271 for i, t in enumerate(thumbnails):
2272 if t.get('id') is None:
2273 t['id'] = '%d' % i
2274 if t.get('width') and t.get('height'):
2275 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2276 t['url'] = sanitize_url(t['url'])
2277
2278 if self.params.get('check_formats') is True:
2279 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2280 else:
2281 info_dict['thumbnails'] = thumbnails
2282
2283 def _fill_common_fields(self, info_dict, is_video=True):
2284 # TODO: move sanitization here
2285 if is_video:
2286 # playlists are allowed to lack "title"
2287 title = info_dict.get('title', NO_DEFAULT)
2288 if title is NO_DEFAULT:
2289 raise ExtractorError('Missing "title" field in extractor result',
2290 video_id=info_dict['id'], ie=info_dict['extractor'])
2291 info_dict['fulltitle'] = title
2292 if not title:
2293 if title == '':
2294 self.write_debug('Extractor gave empty title. Creating a generic title')
2295 else:
2296 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2297 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2298
2299 if info_dict.get('duration') is not None:
2300 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2301
2302 for ts_key, date_key in (
2303 ('timestamp', 'upload_date'),
2304 ('release_timestamp', 'release_date'),
2305 ('modified_timestamp', 'modified_date'),
2306 ):
2307 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2308 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2309 # see http://bugs.python.org/issue1646728)
2310 with contextlib.suppress(ValueError, OverflowError, OSError):
2311 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2312 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2313
2314 live_keys = ('is_live', 'was_live')
2315 live_status = info_dict.get('live_status')
2316 if live_status is None:
2317 for key in live_keys:
2318 if info_dict.get(key) is False:
2319 continue
2320 if info_dict.get(key):
2321 live_status = key
2322 break
2323 if all(info_dict.get(key) is False for key in live_keys):
2324 live_status = 'not_live'
2325 if live_status:
2326 info_dict['live_status'] = live_status
2327 for key in live_keys:
2328 if info_dict.get(key) is None:
2329 info_dict[key] = (live_status == key)
2330
2331 # Auto generate title fields corresponding to the *_number fields when missing
2332 # in order to always have clean titles. This is very common for TV series.
2333 for field in ('chapter', 'season', 'episode'):
2334 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2335 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2336
2337 def _raise_pending_errors(self, info):
2338 err = info.pop('__pending_error', None)
2339 if err:
2340 self.report_error(err, tb=False)
2341
2342 def process_video_result(self, info_dict, download=True):
2343 assert info_dict.get('_type', 'video') == 'video'
2344 self._num_videos += 1
2345
2346 if 'id' not in info_dict:
2347 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2348 elif not info_dict.get('id'):
2349 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2350
2351 def report_force_conversion(field, field_not, conversion):
2352 self.report_warning(
2353 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2354 % (field, field_not, conversion))
2355
2356 def sanitize_string_field(info, string_field):
2357 field = info.get(string_field)
2358 if field is None or isinstance(field, str):
2359 return
2360 report_force_conversion(string_field, 'a string', 'string')
2361 info[string_field] = str(field)
2362
2363 def sanitize_numeric_fields(info):
2364 for numeric_field in self._NUMERIC_FIELDS:
2365 field = info.get(numeric_field)
2366 if field is None or isinstance(field, (int, float)):
2367 continue
2368 report_force_conversion(numeric_field, 'numeric', 'int')
2369 info[numeric_field] = int_or_none(field)
2370
2371 sanitize_string_field(info_dict, 'id')
2372 sanitize_numeric_fields(info_dict)
2373 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2374 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2375 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2376 self.report_warning('"duration" field is negative, there is an error in extractor')
2377
2378 chapters = info_dict.get('chapters') or []
2379 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2380 for prev, current, next_ in zip(
2381 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)):
2382 if current.get('start_time') is None:
2383 current['start_time'] = prev.get('end_time')
2384 if not current.get('end_time'):
2385 current['end_time'] = next_.get('start_time')
2386
2387 if 'playlist' not in info_dict:
2388 # It isn't part of a playlist
2389 info_dict['playlist'] = None
2390 info_dict['playlist_index'] = None
2391
2392 self._sanitize_thumbnails(info_dict)
2393
2394 thumbnail = info_dict.get('thumbnail')
2395 thumbnails = info_dict.get('thumbnails')
2396 if thumbnail:
2397 info_dict['thumbnail'] = sanitize_url(thumbnail)
2398 elif thumbnails:
2399 info_dict['thumbnail'] = thumbnails[-1]['url']
2400
2401 if info_dict.get('display_id') is None and 'id' in info_dict:
2402 info_dict['display_id'] = info_dict['id']
2403
2404 self._fill_common_fields(info_dict)
2405
2406 for cc_kind in ('subtitles', 'automatic_captions'):
2407 cc = info_dict.get(cc_kind)
2408 if cc:
2409 for _, subtitle in cc.items():
2410 for subtitle_format in subtitle:
2411 if subtitle_format.get('url'):
2412 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2413 if subtitle_format.get('ext') is None:
2414 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2415
2416 automatic_captions = info_dict.get('automatic_captions')
2417 subtitles = info_dict.get('subtitles')
2418
2419 info_dict['requested_subtitles'] = self.process_subtitles(
2420 info_dict['id'], subtitles, automatic_captions)
2421
2422 if info_dict.get('formats') is None:
2423 # There's only one format available
2424 formats = [info_dict]
2425 else:
2426 formats = info_dict['formats']
2427
2428 # or None ensures --clean-infojson removes it
2429 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2430 if not self.params.get('allow_unplayable_formats'):
2431 formats = [f for f in formats if not f.get('has_drm')]
2432 if info_dict['_has_drm'] and all(
2433 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2434 self.report_warning(
2435 'This video is DRM protected and only images are available for download. '
2436 'Use --list-formats to see them')
2437
2438 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2439 if not get_from_start:
2440 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2441 if info_dict.get('is_live') and formats:
2442 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2443 if get_from_start and not formats:
2444 self.raise_no_formats(info_dict, msg=(
2445 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2446 'If you want to download from the current time, use --no-live-from-start'))
2447
2448 if not formats:
2449 self.raise_no_formats(info_dict)
2450
2451 def is_wellformed(f):
2452 url = f.get('url')
2453 if not url:
2454 self.report_warning(
2455 '"url" field is missing or empty - skipping format, '
2456 'there is an error in extractor')
2457 return False
2458 if isinstance(url, bytes):
2459 sanitize_string_field(f, 'url')
2460 return True
2461
2462 # Filter out malformed formats for better extraction robustness
2463 formats = list(filter(is_wellformed, formats))
2464
2465 formats_dict = {}
2466
2467 # We check that all the formats have the format and format_id fields
2468 for i, format in enumerate(formats):
2469 sanitize_string_field(format, 'format_id')
2470 sanitize_numeric_fields(format)
2471 format['url'] = sanitize_url(format['url'])
2472 if not format.get('format_id'):
2473 format['format_id'] = str(i)
2474 else:
2475 # Sanitize format_id from characters used in format selector expression
2476 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2477 format_id = format['format_id']
2478 if format_id not in formats_dict:
2479 formats_dict[format_id] = []
2480 formats_dict[format_id].append(format)
2481
2482 # Make sure all formats have unique format_id
2483 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2484 for format_id, ambiguous_formats in formats_dict.items():
2485 ambigious_id = len(ambiguous_formats) > 1
2486 for i, format in enumerate(ambiguous_formats):
2487 if ambigious_id:
2488 format['format_id'] = '%s-%d' % (format_id, i)
2489 if format.get('ext') is None:
2490 format['ext'] = determine_ext(format['url']).lower()
2491 # Ensure there is no conflict between id and ext in format selection
2492 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2493 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2494 format['format_id'] = 'f%s' % format['format_id']
2495
2496 for i, format in enumerate(formats):
2497 if format.get('format') is None:
2498 format['format'] = '{id} - {res}{note}'.format(
2499 id=format['format_id'],
2500 res=self.format_resolution(format),
2501 note=format_field(format, 'format_note', ' (%s)'),
2502 )
2503 if format.get('protocol') is None:
2504 format['protocol'] = determine_protocol(format)
2505 if format.get('resolution') is None:
2506 format['resolution'] = self.format_resolution(format, default=None)
2507 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2508 format['dynamic_range'] = 'SDR'
2509 if (info_dict.get('duration') and format.get('tbr')
2510 and not format.get('filesize') and not format.get('filesize_approx')):
2511 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2512
2513 # Add HTTP headers, so that external programs can use them from the
2514 # json output
2515 full_format_info = info_dict.copy()
2516 full_format_info.update(format)
2517 format['http_headers'] = self._calc_headers(full_format_info)
2518 # Remove private housekeeping stuff
2519 if '__x_forwarded_for_ip' in info_dict:
2520 del info_dict['__x_forwarded_for_ip']
2521
2522 if self.params.get('check_formats') is True:
2523 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2524
2525 if not formats or formats[0] is not info_dict:
2526 # only set the 'formats' fields if the original info_dict list them
2527 # otherwise we end up with a circular reference, the first (and unique)
2528 # element in the 'formats' field in info_dict is info_dict itself,
2529 # which can't be exported to json
2530 info_dict['formats'] = formats
2531
2532 info_dict, _ = self.pre_process(info_dict)
2533
2534 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2535 return info_dict
2536
2537 self.post_extract(info_dict)
2538 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2539
2540 # The pre-processors may have modified the formats
2541 formats = info_dict.get('formats', [info_dict])
2542
2543 list_only = self.params.get('simulate') is None and (
2544 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2545 interactive_format_selection = not list_only and self.format_selector == '-'
2546 if self.params.get('list_thumbnails'):
2547 self.list_thumbnails(info_dict)
2548 if self.params.get('listsubtitles'):
2549 if 'automatic_captions' in info_dict:
2550 self.list_subtitles(
2551 info_dict['id'], automatic_captions, 'automatic captions')
2552 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2553 if self.params.get('listformats') or interactive_format_selection:
2554 self.list_formats(info_dict)
2555 if list_only:
2556 # Without this printing, -F --print-json will not work
2557 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2558 return info_dict
2559
2560 format_selector = self.format_selector
2561 if format_selector is None:
2562 req_format = self._default_format_spec(info_dict, download=download)
2563 self.write_debug('Default format spec: %s' % req_format)
2564 format_selector = self.build_format_selector(req_format)
2565
2566 while True:
2567 if interactive_format_selection:
2568 req_format = input(
2569 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2570 try:
2571 format_selector = self.build_format_selector(req_format)
2572 except SyntaxError as err:
2573 self.report_error(err, tb=False, is_error=False)
2574 continue
2575
2576 formats_to_download = list(format_selector({
2577 'formats': formats,
2578 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2579 'incomplete_formats': (
2580 # All formats are video-only or
2581 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2582 # all formats are audio-only
2583 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2584 }))
2585 if interactive_format_selection and not formats_to_download:
2586 self.report_error('Requested format is not available', tb=False, is_error=False)
2587 continue
2588 break
2589
2590 if not formats_to_download:
2591 if not self.params.get('ignore_no_formats_error'):
2592 raise ExtractorError(
2593 'Requested format is not available. Use --list-formats for a list of available formats',
2594 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2595 self.report_warning('Requested format is not available')
2596 # Process what we can, even without any available formats.
2597 formats_to_download = [{}]
2598
2599 requested_ranges = self.params.get('download_ranges')
2600 if requested_ranges:
2601 requested_ranges = tuple(requested_ranges(info_dict, self))
2602
2603 best_format, downloaded_formats = formats_to_download[-1], []
2604 if download:
2605 if best_format:
2606 def to_screen(*msg):
2607 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2608
2609 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2610 (f['format_id'] for f in formats_to_download))
2611 if requested_ranges:
2612 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2613 (f'{int(c["start_time"])}-{int(c["end_time"])}' for c in requested_ranges))
2614 max_downloads_reached = False
2615
2616 for fmt, chapter in itertools.product(formats_to_download, requested_ranges or [{}]):
2617 new_info = self._copy_infodict(info_dict)
2618 new_info.update(fmt)
2619 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2620 if chapter or offset:
2621 new_info.update({
2622 'section_start': offset + chapter.get('start_time', 0),
2623 'section_end': offset + min(chapter.get('end_time', duration), duration),
2624 'section_title': chapter.get('title'),
2625 'section_number': chapter.get('index'),
2626 })
2627 downloaded_formats.append(new_info)
2628 try:
2629 self.process_info(new_info)
2630 except MaxDownloadsReached:
2631 max_downloads_reached = True
2632 self._raise_pending_errors(new_info)
2633 # Remove copied info
2634 for key, val in tuple(new_info.items()):
2635 if info_dict.get(key) == val:
2636 new_info.pop(key)
2637 if max_downloads_reached:
2638 break
2639
2640 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2641 assert write_archive.issubset({True, False, 'ignore'})
2642 if True in write_archive and False not in write_archive:
2643 self.record_download_archive(info_dict)
2644
2645 info_dict['requested_downloads'] = downloaded_formats
2646 info_dict = self.run_all_pps('after_video', info_dict)
2647 if max_downloads_reached:
2648 raise MaxDownloadsReached()
2649
2650 # We update the info dict with the selected best quality format (backwards compatibility)
2651 info_dict.update(best_format)
2652 return info_dict
2653
2654 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2655 """Select the requested subtitles and their format"""
2656 available_subs, normal_sub_langs = {}, []
2657 if normal_subtitles and self.params.get('writesubtitles'):
2658 available_subs.update(normal_subtitles)
2659 normal_sub_langs = tuple(normal_subtitles.keys())
2660 if automatic_captions and self.params.get('writeautomaticsub'):
2661 for lang, cap_info in automatic_captions.items():
2662 if lang not in available_subs:
2663 available_subs[lang] = cap_info
2664
2665 if (not self.params.get('writesubtitles') and not
2666 self.params.get('writeautomaticsub') or not
2667 available_subs):
2668 return None
2669
2670 all_sub_langs = tuple(available_subs.keys())
2671 if self.params.get('allsubtitles', False):
2672 requested_langs = all_sub_langs
2673 elif self.params.get('subtitleslangs', False):
2674 # A list is used so that the order of languages will be the same as
2675 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2676 requested_langs = []
2677 for lang_re in self.params.get('subtitleslangs'):
2678 discard = lang_re[0] == '-'
2679 if discard:
2680 lang_re = lang_re[1:]
2681 if lang_re == 'all':
2682 if discard:
2683 requested_langs = []
2684 else:
2685 requested_langs.extend(all_sub_langs)
2686 continue
2687 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2688 if discard:
2689 for lang in current_langs:
2690 while lang in requested_langs:
2691 requested_langs.remove(lang)
2692 else:
2693 requested_langs.extend(current_langs)
2694 requested_langs = orderedSet(requested_langs)
2695 elif normal_sub_langs:
2696 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2697 else:
2698 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2699 if requested_langs:
2700 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2701
2702 formats_query = self.params.get('subtitlesformat', 'best')
2703 formats_preference = formats_query.split('/') if formats_query else []
2704 subs = {}
2705 for lang in requested_langs:
2706 formats = available_subs.get(lang)
2707 if formats is None:
2708 self.report_warning(f'{lang} subtitles not available for {video_id}')
2709 continue
2710 for ext in formats_preference:
2711 if ext == 'best':
2712 f = formats[-1]
2713 break
2714 matches = list(filter(lambda f: f['ext'] == ext, formats))
2715 if matches:
2716 f = matches[-1]
2717 break
2718 else:
2719 f = formats[-1]
2720 self.report_warning(
2721 'No subtitle format found matching "%s" for language %s, '
2722 'using %s' % (formats_query, lang, f['ext']))
2723 subs[lang] = f
2724 return subs
2725
2726 def _forceprint(self, key, info_dict):
2727 if info_dict is None:
2728 return
2729 info_copy = info_dict.copy()
2730 info_copy['formats_table'] = self.render_formats_table(info_dict)
2731 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2732 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2733 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2734
2735 def format_tmpl(tmpl):
2736 mobj = re.match(r'\w+(=?)$', tmpl)
2737 if mobj and mobj.group(1):
2738 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2739 elif mobj:
2740 return f'%({tmpl})s'
2741 return tmpl
2742
2743 for tmpl in self.params['forceprint'].get(key, []):
2744 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2745
2746 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2747 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2748 tmpl = format_tmpl(tmpl)
2749 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2750 if self._ensure_dir_exists(filename):
2751 with open(filename, 'a', encoding='utf-8') as f:
2752 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2753
2754 def __forced_printings(self, info_dict, filename, incomplete):
2755 def print_mandatory(field, actual_field=None):
2756 if actual_field is None:
2757 actual_field = field
2758 if (self.params.get('force%s' % field, False)
2759 and (not incomplete or info_dict.get(actual_field) is not None)):
2760 self.to_stdout(info_dict[actual_field])
2761
2762 def print_optional(field):
2763 if (self.params.get('force%s' % field, False)
2764 and info_dict.get(field) is not None):
2765 self.to_stdout(info_dict[field])
2766
2767 info_dict = info_dict.copy()
2768 if filename is not None:
2769 info_dict['filename'] = filename
2770 if info_dict.get('requested_formats') is not None:
2771 # For RTMP URLs, also include the playpath
2772 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2773 elif info_dict.get('url'):
2774 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2775
2776 if (self.params.get('forcejson')
2777 or self.params['forceprint'].get('video')
2778 or self.params['print_to_file'].get('video')):
2779 self.post_extract(info_dict)
2780 self._forceprint('video', info_dict)
2781
2782 print_mandatory('title')
2783 print_mandatory('id')
2784 print_mandatory('url', 'urls')
2785 print_optional('thumbnail')
2786 print_optional('description')
2787 print_optional('filename')
2788 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2789 self.to_stdout(formatSeconds(info_dict['duration']))
2790 print_mandatory('format')
2791
2792 if self.params.get('forcejson'):
2793 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2794
2795 def dl(self, name, info, subtitle=False, test=False):
2796 if not info.get('url'):
2797 self.raise_no_formats(info, True)
2798
2799 if test:
2800 verbose = self.params.get('verbose')
2801 params = {
2802 'test': True,
2803 'quiet': self.params.get('quiet') or not verbose,
2804 'verbose': verbose,
2805 'noprogress': not verbose,
2806 'nopart': True,
2807 'skip_unavailable_fragments': False,
2808 'keep_fragments': False,
2809 'overwrites': True,
2810 '_no_ytdl_file': True,
2811 }
2812 else:
2813 params = self.params
2814 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2815 if not test:
2816 for ph in self._progress_hooks:
2817 fd.add_progress_hook(ph)
2818 urls = '", "'.join(
2819 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2820 for f in info.get('requested_formats', []) or [info])
2821 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2822
2823 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2824 # But it may contain objects that are not deep-copyable
2825 new_info = self._copy_infodict(info)
2826 if new_info.get('http_headers') is None:
2827 new_info['http_headers'] = self._calc_headers(new_info)
2828 return fd.download(name, new_info, subtitle)
2829
2830 def existing_file(self, filepaths, *, default_overwrite=True):
2831 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2832 if existing_files and not self.params.get('overwrites', default_overwrite):
2833 return existing_files[0]
2834
2835 for file in existing_files:
2836 self.report_file_delete(file)
2837 os.remove(file)
2838 return None
2839
2840 def process_info(self, info_dict):
2841 """Process a single resolved IE result. (Modifies it in-place)"""
2842
2843 assert info_dict.get('_type', 'video') == 'video'
2844 original_infodict = info_dict
2845
2846 if 'format' not in info_dict and 'ext' in info_dict:
2847 info_dict['format'] = info_dict['ext']
2848
2849 # This is mostly just for backward compatibility of process_info
2850 # As a side-effect, this allows for format-specific filters
2851 if self._match_entry(info_dict) is not None:
2852 info_dict['__write_download_archive'] = 'ignore'
2853 return
2854
2855 # Does nothing under normal operation - for backward compatibility of process_info
2856 self.post_extract(info_dict)
2857 self._num_downloads += 1
2858
2859 # info_dict['_filename'] needs to be set for backward compatibility
2860 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2861 temp_filename = self.prepare_filename(info_dict, 'temp')
2862 files_to_move = {}
2863
2864 # Forced printings
2865 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2866
2867 def check_max_downloads():
2868 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
2869 raise MaxDownloadsReached()
2870
2871 if self.params.get('simulate'):
2872 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2873 check_max_downloads()
2874 return
2875
2876 if full_filename is None:
2877 return
2878 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2879 return
2880 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2881 return
2882
2883 if self._write_description('video', info_dict,
2884 self.prepare_filename(info_dict, 'description')) is None:
2885 return
2886
2887 sub_files = self._write_subtitles(info_dict, temp_filename)
2888 if sub_files is None:
2889 return
2890 files_to_move.update(dict(sub_files))
2891
2892 thumb_files = self._write_thumbnails(
2893 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2894 if thumb_files is None:
2895 return
2896 files_to_move.update(dict(thumb_files))
2897
2898 infofn = self.prepare_filename(info_dict, 'infojson')
2899 _infojson_written = self._write_info_json('video', info_dict, infofn)
2900 if _infojson_written:
2901 info_dict['infojson_filename'] = infofn
2902 # For backward compatibility, even though it was a private field
2903 info_dict['__infojson_filename'] = infofn
2904 elif _infojson_written is None:
2905 return
2906
2907 # Note: Annotations are deprecated
2908 annofn = None
2909 if self.params.get('writeannotations', False):
2910 annofn = self.prepare_filename(info_dict, 'annotation')
2911 if annofn:
2912 if not self._ensure_dir_exists(encodeFilename(annofn)):
2913 return
2914 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2915 self.to_screen('[info] Video annotations are already present')
2916 elif not info_dict.get('annotations'):
2917 self.report_warning('There are no annotations to write.')
2918 else:
2919 try:
2920 self.to_screen('[info] Writing video annotations to: ' + annofn)
2921 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2922 annofile.write(info_dict['annotations'])
2923 except (KeyError, TypeError):
2924 self.report_warning('There are no annotations to write.')
2925 except OSError:
2926 self.report_error('Cannot write annotations file: ' + annofn)
2927 return
2928
2929 # Write internet shortcut files
2930 def _write_link_file(link_type):
2931 url = try_get(info_dict['webpage_url'], iri_to_uri)
2932 if not url:
2933 self.report_warning(
2934 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2935 return True
2936 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2937 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2938 return False
2939 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2940 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2941 return True
2942 try:
2943 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2944 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2945 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2946 template_vars = {'url': url}
2947 if link_type == 'desktop':
2948 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2949 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2950 except OSError:
2951 self.report_error(f'Cannot write internet shortcut {linkfn}')
2952 return False
2953 return True
2954
2955 write_links = {
2956 'url': self.params.get('writeurllink'),
2957 'webloc': self.params.get('writewebloclink'),
2958 'desktop': self.params.get('writedesktoplink'),
2959 }
2960 if self.params.get('writelink'):
2961 link_type = ('webloc' if sys.platform == 'darwin'
2962 else 'desktop' if sys.platform.startswith('linux')
2963 else 'url')
2964 write_links[link_type] = True
2965
2966 if any(should_write and not _write_link_file(link_type)
2967 for link_type, should_write in write_links.items()):
2968 return
2969
2970 def replace_info_dict(new_info):
2971 nonlocal info_dict
2972 if new_info == info_dict:
2973 return
2974 info_dict.clear()
2975 info_dict.update(new_info)
2976
2977 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2978 replace_info_dict(new_info)
2979
2980 if self.params.get('skip_download'):
2981 info_dict['filepath'] = temp_filename
2982 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2983 info_dict['__files_to_move'] = files_to_move
2984 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2985 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2986 else:
2987 # Download
2988 info_dict.setdefault('__postprocessors', [])
2989 try:
2990
2991 def existing_video_file(*filepaths):
2992 ext = info_dict.get('ext')
2993 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
2994 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
2995 default_overwrite=False)
2996 if file:
2997 info_dict['ext'] = os.path.splitext(file)[1][1:]
2998 return file
2999
3000 fd, success = None, True
3001 if info_dict.get('protocol') or info_dict.get('url'):
3002 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3003 if fd is not FFmpegFD and (
3004 info_dict.get('section_start') or info_dict.get('section_end')):
3005 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3006 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3007 self.report_error(f'{msg}. Aborting')
3008 return
3009
3010 if info_dict.get('requested_formats') is not None:
3011
3012 def compatible_formats(formats):
3013 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3014 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3015 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3016 if len(video_formats) > 2 or len(audio_formats) > 2:
3017 return False
3018
3019 # Check extension
3020 exts = {format.get('ext') for format in formats}
3021 COMPATIBLE_EXTS = (
3022 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3023 {'webm'},
3024 )
3025 for ext_sets in COMPATIBLE_EXTS:
3026 if ext_sets.issuperset(exts):
3027 return True
3028 # TODO: Check acodec/vcodec
3029 return False
3030
3031 requested_formats = info_dict['requested_formats']
3032 old_ext = info_dict['ext']
3033 if self.params.get('merge_output_format') is None:
3034 if not compatible_formats(requested_formats):
3035 info_dict['ext'] = 'mkv'
3036 self.report_warning(
3037 'Requested formats are incompatible for merge and will be merged into mkv')
3038 if (info_dict['ext'] == 'webm'
3039 and info_dict.get('thumbnails')
3040 # check with type instead of pp_key, __name__, or isinstance
3041 # since we dont want any custom PPs to trigger this
3042 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3043 info_dict['ext'] = 'mkv'
3044 self.report_warning(
3045 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3046 new_ext = info_dict['ext']
3047
3048 def correct_ext(filename, ext=new_ext):
3049 if filename == '-':
3050 return filename
3051 filename_real_ext = os.path.splitext(filename)[1][1:]
3052 filename_wo_ext = (
3053 os.path.splitext(filename)[0]
3054 if filename_real_ext in (old_ext, new_ext)
3055 else filename)
3056 return f'{filename_wo_ext}.{ext}'
3057
3058 # Ensure filename always has a correct extension for successful merge
3059 full_filename = correct_ext(full_filename)
3060 temp_filename = correct_ext(temp_filename)
3061 dl_filename = existing_video_file(full_filename, temp_filename)
3062 info_dict['__real_download'] = False
3063
3064 merger = FFmpegMergerPP(self)
3065 downloaded = []
3066 if dl_filename is not None:
3067 self.report_file_already_downloaded(dl_filename)
3068 elif fd:
3069 for f in requested_formats if fd != FFmpegFD else []:
3070 f['filepath'] = fname = prepend_extension(
3071 correct_ext(temp_filename, info_dict['ext']),
3072 'f%s' % f['format_id'], info_dict['ext'])
3073 downloaded.append(fname)
3074 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3075 success, real_download = self.dl(temp_filename, info_dict)
3076 info_dict['__real_download'] = real_download
3077 else:
3078 if self.params.get('allow_unplayable_formats'):
3079 self.report_warning(
3080 'You have requested merging of multiple formats '
3081 'while also allowing unplayable formats to be downloaded. '
3082 'The formats won\'t be merged to prevent data corruption.')
3083 elif not merger.available:
3084 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3085 if not self.params.get('ignoreerrors'):
3086 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3087 return
3088 self.report_warning(f'{msg}. The formats won\'t be merged')
3089
3090 if temp_filename == '-':
3091 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3092 else 'but the formats are incompatible for simultaneous download' if merger.available
3093 else 'but ffmpeg is not installed')
3094 self.report_warning(
3095 f'You have requested downloading multiple formats to stdout {reason}. '
3096 'The formats will be streamed one after the other')
3097 fname = temp_filename
3098 for f in requested_formats:
3099 new_info = dict(info_dict)
3100 del new_info['requested_formats']
3101 new_info.update(f)
3102 if temp_filename != '-':
3103 fname = prepend_extension(
3104 correct_ext(temp_filename, new_info['ext']),
3105 'f%s' % f['format_id'], new_info['ext'])
3106 if not self._ensure_dir_exists(fname):
3107 return
3108 f['filepath'] = fname
3109 downloaded.append(fname)
3110 partial_success, real_download = self.dl(fname, new_info)
3111 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3112 success = success and partial_success
3113
3114 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3115 info_dict['__postprocessors'].append(merger)
3116 info_dict['__files_to_merge'] = downloaded
3117 # Even if there were no downloads, it is being merged only now
3118 info_dict['__real_download'] = True
3119 else:
3120 for file in downloaded:
3121 files_to_move[file] = None
3122 else:
3123 # Just a single file
3124 dl_filename = existing_video_file(full_filename, temp_filename)
3125 if dl_filename is None or dl_filename == temp_filename:
3126 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3127 # So we should try to resume the download
3128 success, real_download = self.dl(temp_filename, info_dict)
3129 info_dict['__real_download'] = real_download
3130 else:
3131 self.report_file_already_downloaded(dl_filename)
3132
3133 dl_filename = dl_filename or temp_filename
3134 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3135
3136 except network_exceptions as err:
3137 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3138 return
3139 except OSError as err:
3140 raise UnavailableVideoError(err)
3141 except (ContentTooShortError, ) as err:
3142 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3143 return
3144
3145 self._raise_pending_errors(info_dict)
3146 if success and full_filename != '-':
3147
3148 def fixup():
3149 do_fixup = True
3150 fixup_policy = self.params.get('fixup')
3151 vid = info_dict['id']
3152
3153 if fixup_policy in ('ignore', 'never'):
3154 return
3155 elif fixup_policy == 'warn':
3156 do_fixup = 'warn'
3157 elif fixup_policy != 'force':
3158 assert fixup_policy in ('detect_or_warn', None)
3159 if not info_dict.get('__real_download'):
3160 do_fixup = False
3161
3162 def ffmpeg_fixup(cndn, msg, cls):
3163 if not (do_fixup and cndn):
3164 return
3165 elif do_fixup == 'warn':
3166 self.report_warning(f'{vid}: {msg}')
3167 return
3168 pp = cls(self)
3169 if pp.available:
3170 info_dict['__postprocessors'].append(pp)
3171 else:
3172 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3173
3174 stretched_ratio = info_dict.get('stretched_ratio')
3175 ffmpeg_fixup(
3176 stretched_ratio not in (1, None),
3177 f'Non-uniform pixel ratio {stretched_ratio}',
3178 FFmpegFixupStretchedPP)
3179
3180 ffmpeg_fixup(
3181 (info_dict.get('requested_formats') is None
3182 and info_dict.get('container') == 'm4a_dash'
3183 and info_dict.get('ext') == 'm4a'),
3184 'writing DASH m4a. Only some players support this container',
3185 FFmpegFixupM4aPP)
3186
3187 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3188 downloader = downloader.FD_NAME if downloader else None
3189
3190 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3191 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3192 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3193 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3194 FFmpegFixupM3u8PP)
3195 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3196 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3197
3198 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3199 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3200
3201 fixup()
3202 try:
3203 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3204 except PostProcessingError as err:
3205 self.report_error('Postprocessing: %s' % str(err))
3206 return
3207 try:
3208 for ph in self._post_hooks:
3209 ph(info_dict['filepath'])
3210 except Exception as err:
3211 self.report_error('post hooks: %s' % str(err))
3212 return
3213 info_dict['__write_download_archive'] = True
3214
3215 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3216 if self.params.get('force_write_download_archive'):
3217 info_dict['__write_download_archive'] = True
3218 check_max_downloads()
3219
3220 def __download_wrapper(self, func):
3221 @functools.wraps(func)
3222 def wrapper(*args, **kwargs):
3223 try:
3224 res = func(*args, **kwargs)
3225 except UnavailableVideoError as e:
3226 self.report_error(e)
3227 except DownloadCancelled as e:
3228 self.to_screen(f'[info] {e}')
3229 if not self.params.get('break_per_url'):
3230 raise
3231 else:
3232 if self.params.get('dump_single_json', False):
3233 self.post_extract(res)
3234 self.to_stdout(json.dumps(self.sanitize_info(res)))
3235 return wrapper
3236
3237 def download(self, url_list):
3238 """Download a given list of URLs."""
3239 url_list = variadic(url_list) # Passing a single URL is a common mistake
3240 outtmpl = self.params['outtmpl']['default']
3241 if (len(url_list) > 1
3242 and outtmpl != '-'
3243 and '%' not in outtmpl
3244 and self.params.get('max_downloads') != 1):
3245 raise SameFileError(outtmpl)
3246
3247 for url in url_list:
3248 self.__download_wrapper(self.extract_info)(
3249 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3250
3251 return self._download_retcode
3252
3253 def download_with_info_file(self, info_filename):
3254 with contextlib.closing(fileinput.FileInput(
3255 [info_filename], mode='r',
3256 openhook=fileinput.hook_encoded('utf-8'))) as f:
3257 # FileInput doesn't have a read method, we can't call json.load
3258 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3259 try:
3260 self.__download_wrapper(self.process_ie_result)(info, download=True)
3261 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3262 if not isinstance(e, EntryNotInPlaylist):
3263 self.to_stderr('\r')
3264 webpage_url = info.get('webpage_url')
3265 if webpage_url is not None:
3266 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3267 return self.download([webpage_url])
3268 else:
3269 raise
3270 return self._download_retcode
3271
3272 @staticmethod
3273 def sanitize_info(info_dict, remove_private_keys=False):
3274 ''' Sanitize the infodict for converting to json '''
3275 if info_dict is None:
3276 return info_dict
3277 info_dict.setdefault('epoch', int(time.time()))
3278 info_dict.setdefault('_type', 'video')
3279
3280 if remove_private_keys:
3281 reject = lambda k, v: v is None or k.startswith('__') or k in {
3282 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3283 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3284 }
3285 else:
3286 reject = lambda k, v: False
3287
3288 def filter_fn(obj):
3289 if isinstance(obj, dict):
3290 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3291 elif isinstance(obj, (list, tuple, set, LazyList)):
3292 return list(map(filter_fn, obj))
3293 elif obj is None or isinstance(obj, (str, int, float, bool)):
3294 return obj
3295 else:
3296 return repr(obj)
3297
3298 return filter_fn(info_dict)
3299
3300 @staticmethod
3301 def filter_requested_info(info_dict, actually_filter=True):
3302 ''' Alias of sanitize_info for backward compatibility '''
3303 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3304
3305 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3306 for filename in set(filter(None, files_to_delete)):
3307 if msg:
3308 self.to_screen(msg % filename)
3309 try:
3310 os.remove(filename)
3311 except OSError:
3312 self.report_warning(f'Unable to delete file {filename}')
3313 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3314 del info['__files_to_move'][filename]
3315
3316 @staticmethod
3317 def post_extract(info_dict):
3318 def actual_post_extract(info_dict):
3319 if info_dict.get('_type') in ('playlist', 'multi_video'):
3320 for video_dict in info_dict.get('entries', {}):
3321 actual_post_extract(video_dict or {})
3322 return
3323
3324 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3325 info_dict.update(post_extractor())
3326
3327 actual_post_extract(info_dict or {})
3328
3329 def run_pp(self, pp, infodict):
3330 files_to_delete = []
3331 if '__files_to_move' not in infodict:
3332 infodict['__files_to_move'] = {}
3333 try:
3334 files_to_delete, infodict = pp.run(infodict)
3335 except PostProcessingError as e:
3336 # Must be True and not 'only_download'
3337 if self.params.get('ignoreerrors') is True:
3338 self.report_error(e)
3339 return infodict
3340 raise
3341
3342 if not files_to_delete:
3343 return infodict
3344 if self.params.get('keepvideo', False):
3345 for f in files_to_delete:
3346 infodict['__files_to_move'].setdefault(f, '')
3347 else:
3348 self._delete_downloaded_files(
3349 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3350 return infodict
3351
3352 def run_all_pps(self, key, info, *, additional_pps=None):
3353 self._forceprint(key, info)
3354 for pp in (additional_pps or []) + self._pps[key]:
3355 info = self.run_pp(pp, info)
3356 return info
3357
3358 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3359 info = dict(ie_info)
3360 info['__files_to_move'] = files_to_move or {}
3361 try:
3362 info = self.run_all_pps(key, info)
3363 except PostProcessingError as err:
3364 msg = f'Preprocessing: {err}'
3365 info.setdefault('__pending_error', msg)
3366 self.report_error(msg, is_error=False)
3367 return info, info.pop('__files_to_move', None)
3368
3369 def post_process(self, filename, info, files_to_move=None):
3370 """Run all the postprocessors on the given file."""
3371 info['filepath'] = filename
3372 info['__files_to_move'] = files_to_move or {}
3373 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3374 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3375 del info['__files_to_move']
3376 return self.run_all_pps('after_move', info)
3377
3378 def _make_archive_id(self, info_dict):
3379 video_id = info_dict.get('id')
3380 if not video_id:
3381 return
3382 # Future-proof against any change in case
3383 # and backwards compatibility with prior versions
3384 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3385 if extractor is None:
3386 url = str_or_none(info_dict.get('url'))
3387 if not url:
3388 return
3389 # Try to find matching extractor for the URL and take its ie_key
3390 for ie_key, ie in self._ies.items():
3391 if ie.suitable(url):
3392 extractor = ie_key
3393 break
3394 else:
3395 return
3396 return f'{extractor.lower()} {video_id}'
3397
3398 def in_download_archive(self, info_dict):
3399 fn = self.params.get('download_archive')
3400 if fn is None:
3401 return False
3402
3403 vid_id = self._make_archive_id(info_dict)
3404 if not vid_id:
3405 return False # Incomplete video information
3406
3407 return vid_id in self.archive
3408
3409 def record_download_archive(self, info_dict):
3410 fn = self.params.get('download_archive')
3411 if fn is None:
3412 return
3413 vid_id = self._make_archive_id(info_dict)
3414 assert vid_id
3415 self.write_debug(f'Adding to archive: {vid_id}')
3416 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3417 archive_file.write(vid_id + '\n')
3418 self.archive.add(vid_id)
3419
3420 @staticmethod
3421 def format_resolution(format, default='unknown'):
3422 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3423 return 'audio only'
3424 if format.get('resolution') is not None:
3425 return format['resolution']
3426 if format.get('width') and format.get('height'):
3427 return '%dx%d' % (format['width'], format['height'])
3428 elif format.get('height'):
3429 return '%sp' % format['height']
3430 elif format.get('width'):
3431 return '%dx?' % format['width']
3432 return default
3433
3434 def _list_format_headers(self, *headers):
3435 if self.params.get('listformats_table', True) is not False:
3436 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3437 return headers
3438
3439 def _format_note(self, fdict):
3440 res = ''
3441 if fdict.get('ext') in ['f4f', 'f4m']:
3442 res += '(unsupported)'
3443 if fdict.get('language'):
3444 if res:
3445 res += ' '
3446 res += '[%s]' % fdict['language']
3447 if fdict.get('format_note') is not None:
3448 if res:
3449 res += ' '
3450 res += fdict['format_note']
3451 if fdict.get('tbr') is not None:
3452 if res:
3453 res += ', '
3454 res += '%4dk' % fdict['tbr']
3455 if fdict.get('container') is not None:
3456 if res:
3457 res += ', '
3458 res += '%s container' % fdict['container']
3459 if (fdict.get('vcodec') is not None
3460 and fdict.get('vcodec') != 'none'):
3461 if res:
3462 res += ', '
3463 res += fdict['vcodec']
3464 if fdict.get('vbr') is not None:
3465 res += '@'
3466 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3467 res += 'video@'
3468 if fdict.get('vbr') is not None:
3469 res += '%4dk' % fdict['vbr']
3470 if fdict.get('fps') is not None:
3471 if res:
3472 res += ', '
3473 res += '%sfps' % fdict['fps']
3474 if fdict.get('acodec') is not None:
3475 if res:
3476 res += ', '
3477 if fdict['acodec'] == 'none':
3478 res += 'video only'
3479 else:
3480 res += '%-5s' % fdict['acodec']
3481 elif fdict.get('abr') is not None:
3482 if res:
3483 res += ', '
3484 res += 'audio'
3485 if fdict.get('abr') is not None:
3486 res += '@%3dk' % fdict['abr']
3487 if fdict.get('asr') is not None:
3488 res += ' (%5dHz)' % fdict['asr']
3489 if fdict.get('filesize') is not None:
3490 if res:
3491 res += ', '
3492 res += format_bytes(fdict['filesize'])
3493 elif fdict.get('filesize_approx') is not None:
3494 if res:
3495 res += ', '
3496 res += '~' + format_bytes(fdict['filesize_approx'])
3497 return res
3498
3499 def render_formats_table(self, info_dict):
3500 if not info_dict.get('formats') and not info_dict.get('url'):
3501 return None
3502
3503 formats = info_dict.get('formats', [info_dict])
3504 if not self.params.get('listformats_table', True) is not False:
3505 table = [
3506 [
3507 format_field(f, 'format_id'),
3508 format_field(f, 'ext'),
3509 self.format_resolution(f),
3510 self._format_note(f)
3511 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3512 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3513
3514 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3515 table = [
3516 [
3517 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3518 format_field(f, 'ext'),
3519 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3520 format_field(f, 'fps', '\t%d'),
3521 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3522 delim,
3523 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3524 format_field(f, 'tbr', '\t%dk'),
3525 shorten_protocol_name(f.get('protocol', '')),
3526 delim,
3527 format_field(f, 'vcodec', default='unknown').replace(
3528 'none', 'images' if f.get('acodec') == 'none'
3529 else self._format_out('audio only', self.Styles.SUPPRESS)),
3530 format_field(f, 'vbr', '\t%dk'),
3531 format_field(f, 'acodec', default='unknown').replace(
3532 'none', '' if f.get('vcodec') == 'none'
3533 else self._format_out('video only', self.Styles.SUPPRESS)),
3534 format_field(f, 'abr', '\t%dk'),
3535 format_field(f, 'asr', '\t%dHz'),
3536 join_nonempty(
3537 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3538 format_field(f, 'language', '[%s]'),
3539 join_nonempty(format_field(f, 'format_note'),
3540 format_field(f, 'container', ignore=(None, f.get('ext'))),
3541 delim=', '),
3542 delim=' '),
3543 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3544 header_line = self._list_format_headers(
3545 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3546 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3547
3548 return render_table(
3549 header_line, table, hide_empty=True,
3550 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3551
3552 def render_thumbnails_table(self, info_dict):
3553 thumbnails = list(info_dict.get('thumbnails') or [])
3554 if not thumbnails:
3555 return None
3556 return render_table(
3557 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3558 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3559
3560 def render_subtitles_table(self, video_id, subtitles):
3561 def _row(lang, formats):
3562 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3563 if len(set(names)) == 1:
3564 names = [] if names[0] == 'unknown' else names[:1]
3565 return [lang, ', '.join(names), ', '.join(exts)]
3566
3567 if not subtitles:
3568 return None
3569 return render_table(
3570 self._list_format_headers('Language', 'Name', 'Formats'),
3571 [_row(lang, formats) for lang, formats in subtitles.items()],
3572 hide_empty=True)
3573
3574 def __list_table(self, video_id, name, func, *args):
3575 table = func(*args)
3576 if not table:
3577 self.to_screen(f'{video_id} has no {name}')
3578 return
3579 self.to_screen(f'[info] Available {name} for {video_id}:')
3580 self.to_stdout(table)
3581
3582 def list_formats(self, info_dict):
3583 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3584
3585 def list_thumbnails(self, info_dict):
3586 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3587
3588 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3589 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3590
3591 def urlopen(self, req):
3592 """ Start an HTTP download """
3593 if isinstance(req, str):
3594 req = sanitized_Request(req)
3595 return self._opener.open(req, timeout=self._socket_timeout)
3596
3597 def print_debug_header(self):
3598 if not self.params.get('verbose'):
3599 return
3600
3601 # These imports can be slow. So import them only as needed
3602 from .extractor.extractors import _LAZY_LOADER
3603 from .extractor.extractors import _PLUGIN_CLASSES as plugin_extractors
3604
3605 def get_encoding(stream):
3606 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3607 if not supports_terminal_sequences(stream):
3608 from .utils import WINDOWS_VT_MODE # Must be imported locally
3609 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3610 return ret
3611
3612 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3613 locale.getpreferredencoding(),
3614 sys.getfilesystemencoding(),
3615 self.get_encoding(),
3616 ', '.join(
3617 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3618 if stream is not None and key != 'console')
3619 )
3620
3621 logger = self.params.get('logger')
3622 if logger:
3623 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3624 write_debug(encoding_str)
3625 else:
3626 write_string(f'[debug] {encoding_str}\n', encoding=None)
3627 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3628
3629 source = detect_variant()
3630 write_debug(join_nonempty(
3631 'yt-dlp version', __version__,
3632 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3633 '' if source == 'unknown' else f'({source})',
3634 delim=' '))
3635 if not _LAZY_LOADER:
3636 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3637 write_debug('Lazy loading extractors is forcibly disabled')
3638 else:
3639 write_debug('Lazy loading extractors is disabled')
3640 if plugin_extractors or plugin_postprocessors:
3641 write_debug('Plugins: %s' % [
3642 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3643 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3644 if self.params['compat_opts']:
3645 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3646
3647 if source == 'source':
3648 try:
3649 stdout, _, _ = Popen.run(
3650 ['git', 'rev-parse', '--short', 'HEAD'],
3651 text=True, cwd=os.path.dirname(os.path.abspath(__file__)),
3652 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
3653 if re.fullmatch('[0-9a-f]+', stdout.strip()):
3654 write_debug(f'Git HEAD: {stdout.strip()}')
3655 except Exception:
3656 with contextlib.suppress(Exception):
3657 sys.exc_clear()
3658
3659 def python_implementation():
3660 impl_name = platform.python_implementation()
3661 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3662 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3663 return impl_name
3664
3665 write_debug('Python version %s (%s %s) - %s' % (
3666 platform.python_version(),
3667 python_implementation(),
3668 platform.architecture()[0],
3669 platform_name()))
3670
3671 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3672 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3673 if ffmpeg_features:
3674 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3675
3676 exe_versions['rtmpdump'] = rtmpdump_version()
3677 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3678 exe_str = ', '.join(
3679 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3680 ) or 'none'
3681 write_debug('exe versions: %s' % exe_str)
3682
3683 from .compat.compat_utils import get_package_info
3684 from .dependencies import available_dependencies
3685
3686 write_debug('Optional libraries: %s' % (', '.join(sorted({
3687 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3688 })) or 'none'))
3689
3690 self._setup_opener()
3691 proxy_map = {}
3692 for handler in self._opener.handlers:
3693 if hasattr(handler, 'proxies'):
3694 proxy_map.update(handler.proxies)
3695 write_debug(f'Proxy map: {proxy_map}')
3696
3697 # Not implemented
3698 if False and self.params.get('call_home'):
3699 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3700 write_debug('Public IP address: %s' % ipaddr)
3701 latest_version = self.urlopen(
3702 'https://yt-dl.org/latest/version').read().decode()
3703 if version_tuple(latest_version) > version_tuple(__version__):
3704 self.report_warning(
3705 'You are using an outdated version (newest version: %s)! '
3706 'See https://yt-dl.org/update if you need help updating.' %
3707 latest_version)
3708
3709 def _setup_opener(self):
3710 if hasattr(self, '_opener'):
3711 return
3712 timeout_val = self.params.get('socket_timeout')
3713 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3714
3715 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3716 opts_cookiefile = self.params.get('cookiefile')
3717 opts_proxy = self.params.get('proxy')
3718
3719 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3720
3721 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3722 if opts_proxy is not None:
3723 if opts_proxy == '':
3724 proxies = {}
3725 else:
3726 proxies = {'http': opts_proxy, 'https': opts_proxy}
3727 else:
3728 proxies = urllib.request.getproxies()
3729 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3730 if 'http' in proxies and 'https' not in proxies:
3731 proxies['https'] = proxies['http']
3732 proxy_handler = PerRequestProxyHandler(proxies)
3733
3734 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3735 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3736 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3737 redirect_handler = YoutubeDLRedirectHandler()
3738 data_handler = urllib.request.DataHandler()
3739
3740 # When passing our own FileHandler instance, build_opener won't add the
3741 # default FileHandler and allows us to disable the file protocol, which
3742 # can be used for malicious purposes (see
3743 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3744 file_handler = urllib.request.FileHandler()
3745
3746 def file_open(*args, **kwargs):
3747 raise urllib.error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3748 file_handler.file_open = file_open
3749
3750 opener = urllib.request.build_opener(
3751 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3752
3753 # Delete the default user-agent header, which would otherwise apply in
3754 # cases where our custom HTTP handler doesn't come into play
3755 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3756 opener.addheaders = []
3757 self._opener = opener
3758
3759 def encode(self, s):
3760 if isinstance(s, bytes):
3761 return s # Already encoded
3762
3763 try:
3764 return s.encode(self.get_encoding())
3765 except UnicodeEncodeError as err:
3766 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3767 raise
3768
3769 def get_encoding(self):
3770 encoding = self.params.get('encoding')
3771 if encoding is None:
3772 encoding = preferredencoding()
3773 return encoding
3774
3775 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3776 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3777 if overwrite is None:
3778 overwrite = self.params.get('overwrites', True)
3779 if not self.params.get('writeinfojson'):
3780 return False
3781 elif not infofn:
3782 self.write_debug(f'Skipping writing {label} infojson')
3783 return False
3784 elif not self._ensure_dir_exists(infofn):
3785 return None
3786 elif not overwrite and os.path.exists(infofn):
3787 self.to_screen(f'[info] {label.title()} metadata is already present')
3788 return 'exists'
3789
3790 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3791 try:
3792 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3793 return True
3794 except OSError:
3795 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3796 return None
3797
3798 def _write_description(self, label, ie_result, descfn):
3799 ''' Write description and returns True = written, False = skip, None = error '''
3800 if not self.params.get('writedescription'):
3801 return False
3802 elif not descfn:
3803 self.write_debug(f'Skipping writing {label} description')
3804 return False
3805 elif not self._ensure_dir_exists(descfn):
3806 return None
3807 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3808 self.to_screen(f'[info] {label.title()} description is already present')
3809 elif ie_result.get('description') is None:
3810 self.report_warning(f'There\'s no {label} description to write')
3811 return False
3812 else:
3813 try:
3814 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3815 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3816 descfile.write(ie_result['description'])
3817 except OSError:
3818 self.report_error(f'Cannot write {label} description file {descfn}')
3819 return None
3820 return True
3821
3822 def _write_subtitles(self, info_dict, filename):
3823 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3824 ret = []
3825 subtitles = info_dict.get('requested_subtitles')
3826 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3827 # subtitles download errors are already managed as troubles in relevant IE
3828 # that way it will silently go on when used with unsupporting IE
3829 return ret
3830
3831 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3832 if not sub_filename_base:
3833 self.to_screen('[info] Skipping writing video subtitles')
3834 return ret
3835 for sub_lang, sub_info in subtitles.items():
3836 sub_format = sub_info['ext']
3837 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3838 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3839 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3840 if existing_sub:
3841 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3842 sub_info['filepath'] = existing_sub
3843 ret.append((existing_sub, sub_filename_final))
3844 continue
3845
3846 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3847 if sub_info.get('data') is not None:
3848 try:
3849 # Use newline='' to prevent conversion of newline characters
3850 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3851 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3852 subfile.write(sub_info['data'])
3853 sub_info['filepath'] = sub_filename
3854 ret.append((sub_filename, sub_filename_final))
3855 continue
3856 except OSError:
3857 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3858 return None
3859
3860 try:
3861 sub_copy = sub_info.copy()
3862 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3863 self.dl(sub_filename, sub_copy, subtitle=True)
3864 sub_info['filepath'] = sub_filename
3865 ret.append((sub_filename, sub_filename_final))
3866 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3867 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3868 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3869 if not self.params.get('ignoreerrors'):
3870 self.report_error(msg)
3871 raise DownloadError(msg)
3872 self.report_warning(msg)
3873 return ret
3874
3875 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3876 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3877 write_all = self.params.get('write_all_thumbnails', False)
3878 thumbnails, ret = [], []
3879 if write_all or self.params.get('writethumbnail', False):
3880 thumbnails = info_dict.get('thumbnails') or []
3881 multiple = write_all and len(thumbnails) > 1
3882
3883 if thumb_filename_base is None:
3884 thumb_filename_base = filename
3885 if thumbnails and not thumb_filename_base:
3886 self.write_debug(f'Skipping writing {label} thumbnail')
3887 return ret
3888
3889 for idx, t in list(enumerate(thumbnails))[::-1]:
3890 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3891 thumb_display_id = f'{label} thumbnail {t["id"]}'
3892 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3893 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3894
3895 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3896 if existing_thumb:
3897 self.to_screen('[info] %s is already present' % (
3898 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3899 t['filepath'] = existing_thumb
3900 ret.append((existing_thumb, thumb_filename_final))
3901 else:
3902 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3903 try:
3904 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3905 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3906 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3907 shutil.copyfileobj(uf, thumbf)
3908 ret.append((thumb_filename, thumb_filename_final))
3909 t['filepath'] = thumb_filename
3910 except network_exceptions as err:
3911 thumbnails.pop(idx)
3912 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3913 if ret and not write_all:
3914 break
3915 return ret