]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
Remove Python 3.6 support
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 import collections
2 import contextlib
3 import datetime
4 import errno
5 import fileinput
6 import functools
7 import io
8 import itertools
9 import json
10 import locale
11 import operator
12 import os
13 import random
14 import re
15 import shutil
16 import subprocess
17 import sys
18 import tempfile
19 import time
20 import tokenize
21 import traceback
22 import unicodedata
23 import urllib.request
24 from string import ascii_letters
25
26 from .cache import Cache
27 from .compat import compat_os_name, compat_shlex_quote
28 from .cookies import load_cookies
29 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
30 from .downloader.rtmp import rtmpdump_version
31 from .extractor import gen_extractor_classes, get_info_extractor
32 from .extractor.openload import PhantomJSwrapper
33 from .minicurses import format_text
34 from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
35 from .postprocessor import (
36 EmbedThumbnailPP,
37 FFmpegFixupDuplicateMoovPP,
38 FFmpegFixupDurationPP,
39 FFmpegFixupM3u8PP,
40 FFmpegFixupM4aPP,
41 FFmpegFixupStretchedPP,
42 FFmpegFixupTimestampPP,
43 FFmpegMergerPP,
44 FFmpegPostProcessor,
45 FFmpegVideoConvertorPP,
46 MoveFilesAfterDownloadPP,
47 get_postprocessor,
48 )
49 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
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 bug_reports_message,
89 date_from_str,
90 determine_ext,
91 determine_protocol,
92 encode_compat_str,
93 encodeFilename,
94 error_to_compat_str,
95 escapeHTML,
96 expand_path,
97 filter_dict,
98 float_or_none,
99 format_bytes,
100 format_decimal_suffix,
101 format_field,
102 formatSeconds,
103 get_domain,
104 int_or_none,
105 iri_to_uri,
106 join_nonempty,
107 locked_file,
108 make_dir,
109 make_HTTPS_handler,
110 merge_headers,
111 network_exceptions,
112 number_of_digits,
113 orderedSet,
114 parse_filesize,
115 preferredencoding,
116 prepend_extension,
117 register_socks_protocols,
118 remove_terminal_sequences,
119 render_table,
120 replace_extension,
121 sanitize_filename,
122 sanitize_path,
123 sanitize_url,
124 sanitized_Request,
125 std_headers,
126 str_or_none,
127 strftime_or_none,
128 subtitles_filename,
129 supports_terminal_sequences,
130 system_identifier,
131 timetuple_from_msec,
132 to_high_limit_path,
133 traverse_obj,
134 try_get,
135 url_basename,
136 variadic,
137 version_tuple,
138 windows_enable_vt_mode,
139 write_json_file,
140 write_string,
141 )
142 from .version import RELEASE_GIT_HEAD, __version__
143
144 if compat_os_name == 'nt':
145 import ctypes
146
147
148 class YoutubeDL:
149 """YoutubeDL class.
150
151 YoutubeDL objects are the ones responsible of downloading the
152 actual video file and writing it to disk if the user has requested
153 it, among some other tasks. In most cases there should be one per
154 program. As, given a video URL, the downloader doesn't know how to
155 extract all the needed information, task that InfoExtractors do, it
156 has to pass the URL to one of them.
157
158 For this, YoutubeDL objects have a method that allows
159 InfoExtractors to be registered in a given order. When it is passed
160 a URL, the YoutubeDL object handles it to the first InfoExtractor it
161 finds that reports being able to handle it. The InfoExtractor extracts
162 all the information about the video or videos the URL refers to, and
163 YoutubeDL process the extracted information, possibly using a File
164 Downloader to download the video.
165
166 YoutubeDL objects accept a lot of parameters. In order not to saturate
167 the object constructor with arguments, it receives a dictionary of
168 options instead. These options are available through the params
169 attribute for the InfoExtractors to use. The YoutubeDL also
170 registers itself as the downloader in charge for the InfoExtractors
171 that are added to it, so this is a "mutual registration".
172
173 Available options:
174
175 username: Username for authentication purposes.
176 password: Password for authentication purposes.
177 videopassword: Password for accessing a video.
178 ap_mso: Adobe Pass multiple-system operator identifier.
179 ap_username: Multiple-system operator account username.
180 ap_password: Multiple-system operator account password.
181 usenetrc: Use netrc for authentication instead.
182 verbose: Print additional info to stdout.
183 quiet: Do not print messages to stdout.
184 no_warnings: Do not print out anything for warnings.
185 forceprint: A dict with keys WHEN mapped to a list of templates to
186 print to stdout. The allowed keys are video or any of the
187 items in utils.POSTPROCESS_WHEN.
188 For compatibility, a single list is also accepted
189 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
190 a list of tuples with (template, filename)
191 forcejson: Force printing info_dict as JSON.
192 dump_single_json: Force printing the info_dict of the whole playlist
193 (or video) as a single JSON line.
194 force_write_download_archive: Force writing download archive regardless
195 of 'skip_download' or 'simulate'.
196 simulate: Do not download the video files. If unset (or None),
197 simulate only if listsubtitles, listformats or list_thumbnails is used
198 format: Video format code. see "FORMAT SELECTION" for more details.
199 You can also pass a function. The function takes 'ctx' as
200 argument and returns the formats to download.
201 See "build_format_selector" for an implementation
202 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
203 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
204 extracting metadata even if the video is not actually
205 available for download (experimental)
206 format_sort: A list of fields by which to sort the video formats.
207 See "Sorting Formats" for more details.
208 format_sort_force: Force the given format_sort. see "Sorting Formats"
209 for more details.
210 prefer_free_formats: Whether to prefer video formats with free containers
211 over non-free ones of same quality.
212 allow_multiple_video_streams: Allow multiple video streams to be merged
213 into a single file
214 allow_multiple_audio_streams: Allow multiple audio streams to be merged
215 into a single file
216 check_formats Whether to test if the formats are downloadable.
217 Can be True (check all), False (check none),
218 'selected' (check selected formats),
219 or None (check only if requested by extractor)
220 paths: Dictionary of output paths. The allowed keys are 'home'
221 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
222 outtmpl: Dictionary of templates for output names. Allowed keys
223 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
224 For compatibility with youtube-dl, a single string can also be used
225 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
226 restrictfilenames: Do not allow "&" and spaces in file names
227 trim_file_name: Limit length of filename (extension excluded)
228 windowsfilenames: Force the filenames to be windows compatible
229 ignoreerrors: Do not stop on download/postprocessing errors.
230 Can be 'only_download' to ignore only download errors.
231 Default is 'only_download' for CLI, but False for API
232 skip_playlist_after_errors: Number of allowed failures until the rest of
233 the playlist is skipped
234 force_generic_extractor: Force downloader to use the generic extractor
235 overwrites: Overwrite all video and metadata files if True,
236 overwrite only non-video files if None
237 and don't overwrite any file if False
238 For compatibility with youtube-dl,
239 "nooverwrites" may also be used instead
240 playlist_items: Specific indices of playlist to download.
241 playlistrandom: Download playlist items in random order.
242 lazy_playlist: Process playlist entries as they are received.
243 matchtitle: Download only matching titles.
244 rejecttitle: Reject downloads for matching titles.
245 logger: Log messages to a logging.Logger instance.
246 logtostderr: Log messages to stderr instead of stdout.
247 consoletitle: Display progress in console window's titlebar.
248 writedescription: Write the video description to a .description file
249 writeinfojson: Write the video description to a .info.json file
250 clean_infojson: Remove private fields from the infojson
251 getcomments: Extract video comments. This will not be written to disk
252 unless writeinfojson is also given
253 writeannotations: Write the video annotations to a .annotations.xml file
254 writethumbnail: Write the thumbnail image to a file
255 allow_playlist_files: Whether to write playlists' description, infojson etc
256 also to disk when using the 'write*' options
257 write_all_thumbnails: Write all thumbnail formats to files
258 writelink: Write an internet shortcut file, depending on the
259 current platform (.url/.webloc/.desktop)
260 writeurllink: Write a Windows internet shortcut file (.url)
261 writewebloclink: Write a macOS internet shortcut file (.webloc)
262 writedesktoplink: Write a Linux internet shortcut file (.desktop)
263 writesubtitles: Write the video subtitles to a file
264 writeautomaticsub: Write the automatically generated subtitles to a file
265 listsubtitles: Lists all available subtitles for the video
266 subtitlesformat: The format code for subtitles
267 subtitleslangs: List of languages of the subtitles to download (can be regex).
268 The list may contain "all" to refer to all the available
269 subtitles. The language can be prefixed with a "-" to
270 exclude it from the requested languages. Eg: ['all', '-live_chat']
271 keepvideo: Keep the video file after post-processing
272 daterange: A DateRange object, download only if the upload_date is in the range.
273 skip_download: Skip the actual download of the video file
274 cachedir: Location of the cache files in the filesystem.
275 False to disable filesystem cache.
276 noplaylist: Download single video instead of a playlist if in doubt.
277 age_limit: An integer representing the user's age in years.
278 Unsuitable videos for the given age are skipped.
279 min_views: An integer representing the minimum view count the video
280 must have in order to not be skipped.
281 Videos without view count information are always
282 downloaded. None for no limit.
283 max_views: An integer representing the maximum view count.
284 Videos that are more popular than that are not
285 downloaded.
286 Videos without view count information are always
287 downloaded. None for no limit.
288 download_archive: File name of a file where all downloads are recorded.
289 Videos already present in the file are not downloaded
290 again.
291 break_on_existing: Stop the download process after attempting to download a
292 file that is in the archive.
293 break_on_reject: Stop the download process when encountering a video that
294 has been filtered out.
295 break_per_url: Whether break_on_reject and break_on_existing
296 should act on each input URL as opposed to for the entire queue
297 cookiefile: File name or text stream from where cookies should be read and dumped to
298 cookiesfrombrowser: A tuple containing the name of the browser, the profile
299 name/pathfrom where cookies are loaded, and the name of the
300 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
301 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
302 support RFC 5746 secure renegotiation
303 nocheckcertificate: Do not verify SSL certificates
304 client_certificate: Path to client certificate file in PEM format. May include the private key
305 client_certificate_key: Path to private key file for client certificate
306 client_certificate_password: Password for client certificate private key, if encrypted.
307 If not provided and the key is encrypted, yt-dlp will ask interactively
308 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
309 (Only supported by some extractors)
310 http_headers: A dictionary of custom headers to be used for all requests
311 proxy: URL of the proxy server to use
312 geo_verification_proxy: URL of the proxy to use for IP address verification
313 on geo-restricted sites.
314 socket_timeout: Time to wait for unresponsive hosts, in seconds
315 bidi_workaround: Work around buggy terminals without bidirectional text
316 support, using fridibi
317 debug_printtraffic:Print out sent and received HTTP traffic
318 default_search: Prepend this string if an input url is not valid.
319 'auto' for elaborate guessing
320 encoding: Use this encoding instead of the system-specified.
321 extract_flat: Whether to resolve and process url_results further
322 * False: Always process (default)
323 * True: Never process
324 * 'in_playlist': Do not process inside playlist/multi_video
325 * 'discard': Always process, but don't return the result
326 from inside playlist/multi_video
327 * 'discard_in_playlist': Same as "discard", but only for
328 playlists (not multi_video)
329 wait_for_video: If given, wait for scheduled streams to become available.
330 The value should be a tuple containing the range
331 (min_secs, max_secs) to wait between retries
332 postprocessors: A list of dictionaries, each with an entry
333 * key: The name of the postprocessor. See
334 yt_dlp/postprocessor/__init__.py for a list.
335 * when: When to run the postprocessor. Allowed values are
336 the entries of utils.POSTPROCESS_WHEN
337 Assumed to be 'post_process' if not given
338 progress_hooks: A list of functions that get called on download
339 progress, with a dictionary with the entries
340 * status: One of "downloading", "error", or "finished".
341 Check this first and ignore unknown values.
342 * info_dict: The extracted info_dict
343
344 If status is one of "downloading", or "finished", the
345 following properties may also be present:
346 * filename: The final filename (always present)
347 * tmpfilename: The filename we're currently writing to
348 * downloaded_bytes: Bytes on disk
349 * total_bytes: Size of the whole file, None if unknown
350 * total_bytes_estimate: Guess of the eventual file size,
351 None if unavailable.
352 * elapsed: The number of seconds since download started.
353 * eta: The estimated time in seconds, None if unknown
354 * speed: The download speed in bytes/second, None if
355 unknown
356 * fragment_index: The counter of the currently
357 downloaded video fragment.
358 * fragment_count: The number of fragments (= individual
359 files that will be merged)
360
361 Progress hooks are guaranteed to be called at least once
362 (with status "finished") if the download is successful.
363 postprocessor_hooks: A list of functions that get called on postprocessing
364 progress, with a dictionary with the entries
365 * status: One of "started", "processing", or "finished".
366 Check this first and ignore unknown values.
367 * postprocessor: Name of the postprocessor
368 * info_dict: The extracted info_dict
369
370 Progress hooks are guaranteed to be called at least twice
371 (with status "started" and "finished") if the processing is successful.
372 merge_output_format: Extension to use when merging formats.
373 final_ext: Expected final extension; used to detect when the file was
374 already downloaded and converted
375 fixup: Automatically correct known faults of the file.
376 One of:
377 - "never": do nothing
378 - "warn": only emit a warning
379 - "detect_or_warn": check whether we can do anything
380 about it, warn otherwise (default)
381 source_address: Client-side IP address to bind to.
382 sleep_interval_requests: Number of seconds to sleep between requests
383 during extraction
384 sleep_interval: Number of seconds to sleep before each download when
385 used alone or a lower bound of a range for randomized
386 sleep before each download (minimum possible number
387 of seconds to sleep) when used along with
388 max_sleep_interval.
389 max_sleep_interval:Upper bound of a range for randomized sleep before each
390 download (maximum possible number of seconds to sleep).
391 Must only be used along with sleep_interval.
392 Actual sleep time will be a random float from range
393 [sleep_interval; max_sleep_interval].
394 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
395 listformats: Print an overview of available video formats and exit.
396 list_thumbnails: Print a table of all thumbnails and exit.
397 match_filter: A function that gets called for every video with the signature
398 (info_dict, *, incomplete: bool) -> Optional[str]
399 For backward compatibility with youtube-dl, the signature
400 (info_dict) -> Optional[str] is also allowed.
401 - If it returns a message, the video is ignored.
402 - If it returns None, the video is downloaded.
403 - If it returns utils.NO_DEFAULT, the user is interactively
404 asked whether to download the video.
405 match_filter_func in utils.py is one example for this.
406 no_color: Do not emit color codes in output.
407 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
408 HTTP header
409 geo_bypass_country:
410 Two-letter ISO 3166-2 country code that will be used for
411 explicit geographic restriction bypassing via faking
412 X-Forwarded-For HTTP header
413 geo_bypass_ip_block:
414 IP range in CIDR notation that will be used similarly to
415 geo_bypass_country
416 external_downloader: A dictionary of protocol keys and the executable of the
417 external downloader to use for it. The allowed protocols
418 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
419 Set the value to 'native' to use the native downloader
420 compat_opts: Compatibility options. See "Differences in default behavior".
421 The following options do not work when used through the API:
422 filename, abort-on-error, multistreams, no-live-chat, format-sort
423 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
424 Refer __init__.py for their implementation
425 progress_template: Dictionary of templates for progress outputs.
426 Allowed keys are 'download', 'postprocess',
427 'download-title' (console title) and 'postprocess-title'.
428 The template is mapped on a dictionary with keys 'progress' and 'info'
429 retry_sleep_functions: Dictionary of functions that takes the number of attempts
430 as argument and returns the time to sleep in seconds.
431 Allowed keys are 'http', 'fragment', 'file_access'
432 download_ranges: A callback function that gets called for every video with
433 the signature (info_dict, ydl) -> Iterable[Section].
434 Only the returned sections will be downloaded.
435 Each Section is a dict with the following keys:
436 * start_time: Start time of the section in seconds
437 * end_time: End time of the section in seconds
438 * title: Section title (Optional)
439 * index: Section number (Optional)
440 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
441 noprogress: Do not print the progress bar
442
443 The following parameters are not used by YoutubeDL itself, they are used by
444 the downloader (see yt_dlp/downloader/common.py):
445 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
446 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
447 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
448 external_downloader_args, concurrent_fragment_downloads.
449
450 The following options are used by the post processors:
451 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
452 to the binary or its containing directory.
453 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
454 and a list of additional command-line arguments for the
455 postprocessor/executable. The dict can also have "PP+EXE" keys
456 which are used when the given exe is used by the given PP.
457 Use 'default' as the name for arguments to passed to all PP
458 For compatibility with youtube-dl, a single list of args
459 can also be used
460
461 The following options are used by the extractors:
462 extractor_retries: Number of times to retry for known errors
463 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
464 hls_split_discontinuity: Split HLS playlists to different formats at
465 discontinuities such as ad breaks (default: False)
466 extractor_args: A dictionary of arguments to be passed to the extractors.
467 See "EXTRACTOR ARGUMENTS" for details.
468 Eg: {'youtube': {'skip': ['dash', 'hls']}}
469 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
470
471 The following options are deprecated and may be removed in the future:
472
473 playliststart: - Use playlist_items
474 Playlist item to start at.
475 playlistend: - Use playlist_items
476 Playlist item to end at.
477 playlistreverse: - Use playlist_items
478 Download playlist items in reverse order.
479 forceurl: - Use forceprint
480 Force printing final URL.
481 forcetitle: - Use forceprint
482 Force printing title.
483 forceid: - Use forceprint
484 Force printing ID.
485 forcethumbnail: - Use forceprint
486 Force printing thumbnail URL.
487 forcedescription: - Use forceprint
488 Force printing description.
489 forcefilename: - Use forceprint
490 Force printing final filename.
491 forceduration: - Use forceprint
492 Force printing duration.
493 allsubtitles: - Use subtitleslangs = ['all']
494 Downloads all the subtitles of the video
495 (requires writesubtitles or writeautomaticsub)
496 include_ads: - Doesn't work
497 Download ads as well
498 call_home: - Not implemented
499 Boolean, true iff we are allowed to contact the
500 yt-dlp servers for debugging.
501 post_hooks: - Register a custom postprocessor
502 A list of functions that get called as the final step
503 for each video file, after all postprocessors have been
504 called. The filename will be passed as the only argument.
505 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
506 Use the native HLS downloader instead of ffmpeg/avconv
507 if True, otherwise use ffmpeg/avconv if False, otherwise
508 use downloader suggested by extractor if None.
509 prefer_ffmpeg: - avconv support is deprecated
510 If False, use avconv instead of ffmpeg if both are available,
511 otherwise prefer ffmpeg.
512 youtube_include_dash_manifest: - Use extractor_args
513 If True (default), DASH manifests and related
514 data will be downloaded and processed by extractor.
515 You can reduce network I/O by disabling it if you don't
516 care about DASH. (only for youtube)
517 youtube_include_hls_manifest: - Use extractor_args
518 If True (default), HLS manifests and related
519 data will be downloaded and processed by extractor.
520 You can reduce network I/O by disabling it if you don't
521 care about HLS. (only for youtube)
522 """
523
524 _NUMERIC_FIELDS = {
525 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
526 'timestamp', 'release_timestamp',
527 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
528 'average_rating', 'comment_count', 'age_limit',
529 'start_time', 'end_time',
530 'chapter_number', 'season_number', 'episode_number',
531 'track_number', 'disc_number', 'release_year',
532 }
533
534 _format_fields = {
535 # NB: Keep in sync with the docstring of extractor/common.py
536 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
537 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
538 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
539 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
540 'preference', 'language', 'language_preference', 'quality', 'source_preference',
541 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
542 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
543 }
544 _format_selection_exts = {
545 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
546 'video': {'mp4', 'flv', 'webm', '3gp'},
547 'storyboards': {'mhtml'},
548 }
549
550 def __init__(self, params=None, auto_init=True):
551 """Create a FileDownloader object with the given options.
552 @param auto_init Whether to load the default extractors and print header (if verbose).
553 Set to 'no_verbose_header' to not print the header
554 """
555 if params is None:
556 params = {}
557 self.params = params
558 self._ies = {}
559 self._ies_instances = {}
560 self._pps = {k: [] for k in POSTPROCESS_WHEN}
561 self._printed_messages = set()
562 self._first_webpage_request = True
563 self._post_hooks = []
564 self._progress_hooks = []
565 self._postprocessor_hooks = []
566 self._download_retcode = 0
567 self._num_downloads = 0
568 self._num_videos = 0
569 self._playlist_level = 0
570 self._playlist_urls = set()
571 self.cache = Cache(self)
572
573 windows_enable_vt_mode()
574 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
575 self._out_files = Namespace(
576 out=stdout,
577 error=sys.stderr,
578 screen=sys.stderr if self.params.get('quiet') else stdout,
579 console=None if compat_os_name == 'nt' else next(
580 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
581 )
582 self._allow_colors = Namespace(**{
583 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
584 for type_, stream in self._out_files.items_ if type_ != 'console'
585 })
586
587 # The code is left like this to be reused for future deprecations
588 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 7), (3, 7)
589 current_version = sys.version_info[:2]
590 if current_version < MIN_RECOMMENDED:
591 msg = ('Support for Python version %d.%d has been deprecated. '
592 'See https://github.com/yt-dlp/yt-dlp/issues/3764 for more details.'
593 '\n You will no longer receive updates on this version')
594 if current_version < MIN_SUPPORTED:
595 msg = 'Python version %d.%d is no longer supported'
596 self.deprecation_warning(
597 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
598
599 if self.params.get('allow_unplayable_formats'):
600 self.report_warning(
601 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
602 'This is a developer option intended for debugging. \n'
603 ' If you experience any issues while using this option, '
604 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
605
606 def check_deprecated(param, option, suggestion):
607 if self.params.get(param) is not None:
608 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
609 return True
610 return False
611
612 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
613 if self.params.get('geo_verification_proxy') is None:
614 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
615
616 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
617 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
618 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
619
620 for msg in self.params.get('_warnings', []):
621 self.report_warning(msg)
622 for msg in self.params.get('_deprecation_warnings', []):
623 self.deprecation_warning(msg)
624
625 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
626 if 'list-formats' in self.params['compat_opts']:
627 self.params['listformats_table'] = False
628
629 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
630 # nooverwrites was unnecessarily changed to overwrites
631 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
632 # This ensures compatibility with both keys
633 self.params['overwrites'] = not self.params['nooverwrites']
634 elif self.params.get('overwrites') is None:
635 self.params.pop('overwrites', None)
636 else:
637 self.params['nooverwrites'] = not self.params['overwrites']
638
639 self.params.setdefault('forceprint', {})
640 self.params.setdefault('print_to_file', {})
641
642 # Compatibility with older syntax
643 if not isinstance(params['forceprint'], dict):
644 self.params['forceprint'] = {'video': params['forceprint']}
645
646 if self.params.get('bidi_workaround', False):
647 try:
648 import pty
649 master, slave = pty.openpty()
650 width = shutil.get_terminal_size().columns
651 width_args = [] if width is None else ['-w', str(width)]
652 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
653 try:
654 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
655 except OSError:
656 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
657 self._output_channel = os.fdopen(master, 'rb')
658 except OSError as ose:
659 if ose.errno == errno.ENOENT:
660 self.report_warning(
661 'Could not find fribidi executable, ignoring --bidi-workaround. '
662 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
663 else:
664 raise
665
666 if auto_init:
667 if auto_init != 'no_verbose_header':
668 self.print_debug_header()
669 self.add_default_info_extractors()
670
671 if (sys.platform != 'win32'
672 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
673 and not self.params.get('restrictfilenames', False)):
674 # Unicode filesystem API will throw errors (#1474, #13027)
675 self.report_warning(
676 'Assuming --restrict-filenames since file system encoding '
677 'cannot encode all characters. '
678 'Set the LC_ALL environment variable to fix this.')
679 self.params['restrictfilenames'] = True
680
681 self._parse_outtmpl()
682
683 # Creating format selector here allows us to catch syntax errors before the extraction
684 self.format_selector = (
685 self.params.get('format') if self.params.get('format') in (None, '-')
686 else self.params['format'] if callable(self.params['format'])
687 else self.build_format_selector(self.params['format']))
688
689 # Set http_headers defaults according to std_headers
690 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
691
692 hooks = {
693 'post_hooks': self.add_post_hook,
694 'progress_hooks': self.add_progress_hook,
695 'postprocessor_hooks': self.add_postprocessor_hook,
696 }
697 for opt, fn in hooks.items():
698 for ph in self.params.get(opt, []):
699 fn(ph)
700
701 for pp_def_raw in self.params.get('postprocessors', []):
702 pp_def = dict(pp_def_raw)
703 when = pp_def.pop('when', 'post_process')
704 self.add_post_processor(
705 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
706 when=when)
707
708 self._setup_opener()
709 register_socks_protocols()
710
711 def preload_download_archive(fn):
712 """Preload the archive, if any is specified"""
713 if fn is None:
714 return False
715 self.write_debug(f'Loading archive file {fn!r}')
716 try:
717 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
718 for line in archive_file:
719 self.archive.add(line.strip())
720 except OSError as ioe:
721 if ioe.errno != errno.ENOENT:
722 raise
723 return False
724 return True
725
726 self.archive = set()
727 preload_download_archive(self.params.get('download_archive'))
728
729 def warn_if_short_id(self, argv):
730 # short YouTube ID starting with dash?
731 idxs = [
732 i for i, a in enumerate(argv)
733 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
734 if idxs:
735 correct_argv = (
736 ['yt-dlp']
737 + [a for i, a in enumerate(argv) if i not in idxs]
738 + ['--'] + [argv[i] for i in idxs]
739 )
740 self.report_warning(
741 'Long argument string detected. '
742 'Use -- to separate parameters and URLs, like this:\n%s' %
743 args_to_str(correct_argv))
744
745 def add_info_extractor(self, ie):
746 """Add an InfoExtractor object to the end of the list."""
747 ie_key = ie.ie_key()
748 self._ies[ie_key] = ie
749 if not isinstance(ie, type):
750 self._ies_instances[ie_key] = ie
751 ie.set_downloader(self)
752
753 def _get_info_extractor_class(self, ie_key):
754 ie = self._ies.get(ie_key)
755 if ie is None:
756 ie = get_info_extractor(ie_key)
757 self.add_info_extractor(ie)
758 return ie
759
760 def get_info_extractor(self, ie_key):
761 """
762 Get an instance of an IE with name ie_key, it will try to get one from
763 the _ies list, if there's no instance it will create a new one and add
764 it to the extractor list.
765 """
766 ie = self._ies_instances.get(ie_key)
767 if ie is None:
768 ie = get_info_extractor(ie_key)()
769 self.add_info_extractor(ie)
770 return ie
771
772 def add_default_info_extractors(self):
773 """
774 Add the InfoExtractors returned by gen_extractors to the end of the list
775 """
776 for ie in gen_extractor_classes():
777 self.add_info_extractor(ie)
778
779 def add_post_processor(self, pp, when='post_process'):
780 """Add a PostProcessor object to the end of the chain."""
781 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
782 self._pps[when].append(pp)
783 pp.set_downloader(self)
784
785 def add_post_hook(self, ph):
786 """Add the post hook"""
787 self._post_hooks.append(ph)
788
789 def add_progress_hook(self, ph):
790 """Add the download progress hook"""
791 self._progress_hooks.append(ph)
792
793 def add_postprocessor_hook(self, ph):
794 """Add the postprocessing progress hook"""
795 self._postprocessor_hooks.append(ph)
796 for pps in self._pps.values():
797 for pp in pps:
798 pp.add_progress_hook(ph)
799
800 def _bidi_workaround(self, message):
801 if not hasattr(self, '_output_channel'):
802 return message
803
804 assert hasattr(self, '_output_process')
805 assert isinstance(message, str)
806 line_count = message.count('\n') + 1
807 self._output_process.stdin.write((message + '\n').encode())
808 self._output_process.stdin.flush()
809 res = ''.join(self._output_channel.readline().decode()
810 for _ in range(line_count))
811 return res[:-len('\n')]
812
813 def _write_string(self, message, out=None, only_once=False):
814 if only_once:
815 if message in self._printed_messages:
816 return
817 self._printed_messages.add(message)
818 write_string(message, out=out, encoding=self.params.get('encoding'))
819
820 def to_stdout(self, message, skip_eol=False, quiet=None):
821 """Print message to stdout"""
822 if quiet is not None:
823 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
824 if skip_eol is not False:
825 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. Use "YoutubeDL.to_screen" instead')
826 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
827
828 def to_screen(self, message, skip_eol=False, quiet=None):
829 """Print message to screen if not in quiet mode"""
830 if self.params.get('logger'):
831 self.params['logger'].debug(message)
832 return
833 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
834 return
835 self._write_string(
836 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
837 self._out_files.screen)
838
839 def to_stderr(self, message, only_once=False):
840 """Print message to stderr"""
841 assert isinstance(message, str)
842 if self.params.get('logger'):
843 self.params['logger'].error(message)
844 else:
845 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
846
847 def _send_console_code(self, code):
848 if compat_os_name == 'nt' or not self._out_files.console:
849 return
850 self._write_string(code, self._out_files.console)
851
852 def to_console_title(self, message):
853 if not self.params.get('consoletitle', False):
854 return
855 message = remove_terminal_sequences(message)
856 if compat_os_name == 'nt':
857 if ctypes.windll.kernel32.GetConsoleWindow():
858 # c_wchar_p() might not be necessary if `message` is
859 # already of type unicode()
860 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
861 else:
862 self._send_console_code(f'\033]0;{message}\007')
863
864 def save_console_title(self):
865 if not self.params.get('consoletitle') or self.params.get('simulate'):
866 return
867 self._send_console_code('\033[22;0t') # Save the title on stack
868
869 def restore_console_title(self):
870 if not self.params.get('consoletitle') or self.params.get('simulate'):
871 return
872 self._send_console_code('\033[23;0t') # Restore the title from stack
873
874 def __enter__(self):
875 self.save_console_title()
876 return self
877
878 def __exit__(self, *args):
879 self.restore_console_title()
880
881 if self.params.get('cookiefile') is not None:
882 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
883
884 def trouble(self, message=None, tb=None, is_error=True):
885 """Determine action to take when a download problem appears.
886
887 Depending on if the downloader has been configured to ignore
888 download errors or not, this method may throw an exception or
889 not when errors are found, after printing the message.
890
891 @param tb If given, is additional traceback information
892 @param is_error Whether to raise error according to ignorerrors
893 """
894 if message is not None:
895 self.to_stderr(message)
896 if self.params.get('verbose'):
897 if tb is None:
898 if sys.exc_info()[0]: # if .trouble has been called from an except block
899 tb = ''
900 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
901 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
902 tb += encode_compat_str(traceback.format_exc())
903 else:
904 tb_data = traceback.format_list(traceback.extract_stack())
905 tb = ''.join(tb_data)
906 if tb:
907 self.to_stderr(tb)
908 if not is_error:
909 return
910 if not self.params.get('ignoreerrors'):
911 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
912 exc_info = sys.exc_info()[1].exc_info
913 else:
914 exc_info = sys.exc_info()
915 raise DownloadError(message, exc_info)
916 self._download_retcode = 1
917
918 Styles = Namespace(
919 HEADERS='yellow',
920 EMPHASIS='light blue',
921 FILENAME='green',
922 ID='green',
923 DELIM='blue',
924 ERROR='red',
925 WARNING='yellow',
926 SUPPRESS='light black',
927 )
928
929 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
930 text = str(text)
931 if test_encoding:
932 original_text = text
933 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
934 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
935 text = text.encode(encoding, 'ignore').decode(encoding)
936 if fallback is not None and text != original_text:
937 text = fallback
938 return format_text(text, f) if allow_colors else text if fallback is None else fallback
939
940 def _format_out(self, *args, **kwargs):
941 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
942
943 def _format_screen(self, *args, **kwargs):
944 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
945
946 def _format_err(self, *args, **kwargs):
947 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
948
949 def report_warning(self, message, only_once=False):
950 '''
951 Print the message to stderr, it will be prefixed with 'WARNING:'
952 If stderr is a tty file the 'WARNING:' will be colored
953 '''
954 if self.params.get('logger') is not None:
955 self.params['logger'].warning(message)
956 else:
957 if self.params.get('no_warnings'):
958 return
959 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
960
961 def deprecation_warning(self, message):
962 if self.params.get('logger') is not None:
963 self.params['logger'].warning(f'DeprecationWarning: {message}')
964 else:
965 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
966
967 def report_error(self, message, *args, **kwargs):
968 '''
969 Do the same as trouble, but prefixes the message with 'ERROR:', colored
970 in red if stderr is a tty file.
971 '''
972 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
973
974 def write_debug(self, message, only_once=False):
975 '''Log debug message or Print message to stderr'''
976 if not self.params.get('verbose', False):
977 return
978 message = f'[debug] {message}'
979 if self.params.get('logger'):
980 self.params['logger'].debug(message)
981 else:
982 self.to_stderr(message, only_once)
983
984 def report_file_already_downloaded(self, file_name):
985 """Report file has already been fully downloaded."""
986 try:
987 self.to_screen('[download] %s has already been downloaded' % file_name)
988 except UnicodeEncodeError:
989 self.to_screen('[download] The file has already been downloaded')
990
991 def report_file_delete(self, file_name):
992 """Report that existing file will be deleted."""
993 try:
994 self.to_screen('Deleting existing file %s' % file_name)
995 except UnicodeEncodeError:
996 self.to_screen('Deleting existing file')
997
998 def raise_no_formats(self, info, forced=False, *, msg=None):
999 has_drm = info.get('_has_drm')
1000 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1001 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1002 if forced or not ignored:
1003 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1004 expected=has_drm or ignored or expected)
1005 else:
1006 self.report_warning(msg)
1007
1008 def parse_outtmpl(self):
1009 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1010 self._parse_outtmpl()
1011 return self.params['outtmpl']
1012
1013 def _parse_outtmpl(self):
1014 sanitize = IDENTITY
1015 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1016 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1017
1018 outtmpl = self.params.setdefault('outtmpl', {})
1019 if not isinstance(outtmpl, dict):
1020 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1021 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1022
1023 def get_output_path(self, dir_type='', filename=None):
1024 paths = self.params.get('paths', {})
1025 assert isinstance(paths, dict)
1026 path = os.path.join(
1027 expand_path(paths.get('home', '').strip()),
1028 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1029 filename or '')
1030 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1031
1032 @staticmethod
1033 def _outtmpl_expandpath(outtmpl):
1034 # expand_path translates '%%' into '%' and '$$' into '$'
1035 # correspondingly that is not what we want since we need to keep
1036 # '%%' intact for template dict substitution step. Working around
1037 # with boundary-alike separator hack.
1038 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1039 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1040
1041 # outtmpl should be expand_path'ed before template dict substitution
1042 # because meta fields may contain env variables we don't want to
1043 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1044 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1045 return expand_path(outtmpl).replace(sep, '')
1046
1047 @staticmethod
1048 def escape_outtmpl(outtmpl):
1049 ''' Escape any remaining strings like %s, %abc% etc. '''
1050 return re.sub(
1051 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1052 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1053 outtmpl)
1054
1055 @classmethod
1056 def validate_outtmpl(cls, outtmpl):
1057 ''' @return None or Exception object '''
1058 outtmpl = re.sub(
1059 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1060 lambda mobj: f'{mobj.group(0)[:-1]}s',
1061 cls._outtmpl_expandpath(outtmpl))
1062 try:
1063 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1064 return None
1065 except ValueError as err:
1066 return err
1067
1068 @staticmethod
1069 def _copy_infodict(info_dict):
1070 info_dict = dict(info_dict)
1071 info_dict.pop('__postprocessors', None)
1072 info_dict.pop('__pending_error', None)
1073 return info_dict
1074
1075 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1076 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1077 @param sanitize Whether to sanitize the output as a filename.
1078 For backward compatibility, a function can also be passed
1079 """
1080
1081 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1082
1083 info_dict = self._copy_infodict(info_dict)
1084 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1085 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1086 if info_dict.get('duration', None) is not None
1087 else None)
1088 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1089 info_dict['video_autonumber'] = self._num_videos
1090 if info_dict.get('resolution') is None:
1091 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1092
1093 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1094 # of %(field)s to %(field)0Nd for backward compatibility
1095 field_size_compat_map = {
1096 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1097 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1098 'autonumber': self.params.get('autonumber_size') or 5,
1099 }
1100
1101 TMPL_DICT = {}
1102 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1103 MATH_FUNCTIONS = {
1104 '+': float.__add__,
1105 '-': float.__sub__,
1106 }
1107 # Field is of the form key1.key2...
1108 # where keys (except first) can be string, int or slice
1109 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1110 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1111 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1112 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1113 (?P<negate>-)?
1114 (?P<fields>{FIELD_RE})
1115 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1116 (?:>(?P<strf_format>.+?))?
1117 (?P<remaining>
1118 (?P<alternate>(?<!\\),[^|&)]+)?
1119 (?:&(?P<replacement>.*?))?
1120 (?:\|(?P<default>.*?))?
1121 )$''')
1122
1123 def _traverse_infodict(k):
1124 k = k.split('.')
1125 if k[0] == '':
1126 k.pop(0)
1127 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1128
1129 def get_value(mdict):
1130 # Object traversal
1131 value = _traverse_infodict(mdict['fields'])
1132 # Negative
1133 if mdict['negate']:
1134 value = float_or_none(value)
1135 if value is not None:
1136 value *= -1
1137 # Do maths
1138 offset_key = mdict['maths']
1139 if offset_key:
1140 value = float_or_none(value)
1141 operator = None
1142 while offset_key:
1143 item = re.match(
1144 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1145 offset_key).group(0)
1146 offset_key = offset_key[len(item):]
1147 if operator is None:
1148 operator = MATH_FUNCTIONS[item]
1149 continue
1150 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1151 offset = float_or_none(item)
1152 if offset is None:
1153 offset = float_or_none(_traverse_infodict(item))
1154 try:
1155 value = operator(value, multiplier * offset)
1156 except (TypeError, ZeroDivisionError):
1157 return None
1158 operator = None
1159 # Datetime formatting
1160 if mdict['strf_format']:
1161 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1162
1163 return value
1164
1165 na = self.params.get('outtmpl_na_placeholder', 'NA')
1166
1167 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1168 return sanitize_filename(str(value), restricted=restricted, is_id=(
1169 bool(re.search(r'(^|[_.])id(\.|$)', key))
1170 if 'filename-sanitization' in self.params['compat_opts']
1171 else NO_DEFAULT))
1172
1173 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1174 sanitize = bool(sanitize)
1175
1176 def _dumpjson_default(obj):
1177 if isinstance(obj, (set, LazyList)):
1178 return list(obj)
1179 return repr(obj)
1180
1181 def create_key(outer_mobj):
1182 if not outer_mobj.group('has_key'):
1183 return outer_mobj.group(0)
1184 key = outer_mobj.group('key')
1185 mobj = re.match(INTERNAL_FORMAT_RE, key)
1186 initial_field = mobj.group('fields') if mobj else ''
1187 value, replacement, default = None, None, na
1188 while mobj:
1189 mobj = mobj.groupdict()
1190 default = mobj['default'] if mobj['default'] is not None else default
1191 value = get_value(mobj)
1192 replacement = mobj['replacement']
1193 if value is None and mobj['alternate']:
1194 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1195 else:
1196 break
1197
1198 fmt = outer_mobj.group('format')
1199 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1200 fmt = f'0{field_size_compat_map[key]:d}d'
1201
1202 value = default if value is None else value if replacement is None else replacement
1203
1204 flags = outer_mobj.group('conversion') or ''
1205 str_fmt = f'{fmt[:-1]}s'
1206 if fmt[-1] == 'l': # list
1207 delim = '\n' if '#' in flags else ', '
1208 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1209 elif fmt[-1] == 'j': # json
1210 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1211 elif fmt[-1] == 'h': # html
1212 value, fmt = escapeHTML(value), str_fmt
1213 elif fmt[-1] == 'q': # quoted
1214 value = map(str, variadic(value) if '#' in flags else [value])
1215 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1216 elif fmt[-1] == 'B': # bytes
1217 value = f'%{str_fmt}'.encode() % str(value).encode()
1218 value, fmt = value.decode('utf-8', 'ignore'), 's'
1219 elif fmt[-1] == 'U': # unicode normalized
1220 value, fmt = unicodedata.normalize(
1221 # "+" = compatibility equivalence, "#" = NFD
1222 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1223 value), str_fmt
1224 elif fmt[-1] == 'D': # decimal suffix
1225 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1226 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1227 factor=1024 if '#' in flags else 1000)
1228 elif fmt[-1] == 'S': # filename sanitization
1229 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1230 elif fmt[-1] == 'c':
1231 if value:
1232 value = str(value)[0]
1233 else:
1234 fmt = str_fmt
1235 elif fmt[-1] not in 'rs': # numeric
1236 value = float_or_none(value)
1237 if value is None:
1238 value, fmt = default, 's'
1239
1240 if sanitize:
1241 if fmt[-1] == 'r':
1242 # If value is an object, sanitize might convert it to a string
1243 # So we convert it to repr first
1244 value, fmt = repr(value), str_fmt
1245 if fmt[-1] in 'csr':
1246 value = sanitizer(initial_field, value)
1247
1248 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1249 TMPL_DICT[key] = value
1250 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1251
1252 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1253
1254 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1255 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1256 return self.escape_outtmpl(outtmpl) % info_dict
1257
1258 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1259 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1260 if outtmpl is None:
1261 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1262 try:
1263 outtmpl = self._outtmpl_expandpath(outtmpl)
1264 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1265 if not filename:
1266 return None
1267
1268 if tmpl_type in ('', 'temp'):
1269 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1270 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1271 filename = replace_extension(filename, ext, final_ext)
1272 elif tmpl_type:
1273 force_ext = OUTTMPL_TYPES[tmpl_type]
1274 if force_ext:
1275 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1276
1277 # https://github.com/blackjack4494/youtube-dlc/issues/85
1278 trim_file_name = self.params.get('trim_file_name', False)
1279 if trim_file_name:
1280 no_ext, *ext = filename.rsplit('.', 2)
1281 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1282
1283 return filename
1284 except ValueError as err:
1285 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1286 return None
1287
1288 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1289 """Generate the output filename"""
1290 if outtmpl:
1291 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1292 dir_type = None
1293 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1294 if not filename and dir_type not in ('', 'temp'):
1295 return ''
1296
1297 if warn:
1298 if not self.params.get('paths'):
1299 pass
1300 elif filename == '-':
1301 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1302 elif os.path.isabs(filename):
1303 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1304 if filename == '-' or not filename:
1305 return filename
1306
1307 return self.get_output_path(dir_type, filename)
1308
1309 def _match_entry(self, info_dict, incomplete=False, silent=False):
1310 """ Returns None if the file should be downloaded """
1311
1312 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1313
1314 def check_filter():
1315 if 'title' in info_dict:
1316 # This can happen when we're just evaluating the playlist
1317 title = info_dict['title']
1318 matchtitle = self.params.get('matchtitle', False)
1319 if matchtitle:
1320 if not re.search(matchtitle, title, re.IGNORECASE):
1321 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1322 rejecttitle = self.params.get('rejecttitle', False)
1323 if rejecttitle:
1324 if re.search(rejecttitle, title, re.IGNORECASE):
1325 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1326 date = info_dict.get('upload_date')
1327 if date is not None:
1328 dateRange = self.params.get('daterange', DateRange())
1329 if date not in dateRange:
1330 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1331 view_count = info_dict.get('view_count')
1332 if view_count is not None:
1333 min_views = self.params.get('min_views')
1334 if min_views is not None and view_count < min_views:
1335 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1336 max_views = self.params.get('max_views')
1337 if max_views is not None and view_count > max_views:
1338 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1339 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1340 return 'Skipping "%s" because it is age restricted' % video_title
1341
1342 match_filter = self.params.get('match_filter')
1343 if match_filter is not None:
1344 try:
1345 ret = match_filter(info_dict, incomplete=incomplete)
1346 except TypeError:
1347 # For backward compatibility
1348 ret = None if incomplete else match_filter(info_dict)
1349 if ret is NO_DEFAULT:
1350 while True:
1351 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1352 reply = input(self._format_screen(
1353 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1354 if reply in {'y', ''}:
1355 return None
1356 elif reply == 'n':
1357 return f'Skipping {video_title}'
1358 elif ret is not None:
1359 return ret
1360 return None
1361
1362 if self.in_download_archive(info_dict):
1363 reason = '%s has already been recorded in the archive' % video_title
1364 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1365 else:
1366 reason = check_filter()
1367 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1368 if reason is not None:
1369 if not silent:
1370 self.to_screen('[download] ' + reason)
1371 if self.params.get(break_opt, False):
1372 raise break_err()
1373 return reason
1374
1375 @staticmethod
1376 def add_extra_info(info_dict, extra_info):
1377 '''Set the keys from extra_info in info dict if they are missing'''
1378 for key, value in extra_info.items():
1379 info_dict.setdefault(key, value)
1380
1381 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1382 process=True, force_generic_extractor=False):
1383 """
1384 Return a list with a dictionary for each video extracted.
1385
1386 Arguments:
1387 url -- URL to extract
1388
1389 Keyword arguments:
1390 download -- whether to download videos during extraction
1391 ie_key -- extractor key hint
1392 extra_info -- dictionary containing the extra values to add to each result
1393 process -- whether to resolve all unresolved references (URLs, playlist items),
1394 must be True for download to work.
1395 force_generic_extractor -- force using the generic extractor
1396 """
1397
1398 if extra_info is None:
1399 extra_info = {}
1400
1401 if not ie_key and force_generic_extractor:
1402 ie_key = 'Generic'
1403
1404 if ie_key:
1405 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1406 else:
1407 ies = self._ies
1408
1409 for ie_key, ie in ies.items():
1410 if not ie.suitable(url):
1411 continue
1412
1413 if not ie.working():
1414 self.report_warning('The program functionality for this site has been marked as broken, '
1415 'and will probably not work.')
1416
1417 temp_id = ie.get_temp_id(url)
1418 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1419 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1420 if self.params.get('break_on_existing', False):
1421 raise ExistingVideoReached()
1422 break
1423 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1424 else:
1425 self.report_error('no suitable InfoExtractor for URL %s' % url)
1426
1427 def _handle_extraction_exceptions(func):
1428 @functools.wraps(func)
1429 def wrapper(self, *args, **kwargs):
1430 while True:
1431 try:
1432 return func(self, *args, **kwargs)
1433 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1434 raise
1435 except ReExtractInfo as e:
1436 if e.expected:
1437 self.to_screen(f'{e}; Re-extracting data')
1438 else:
1439 self.to_stderr('\r')
1440 self.report_warning(f'{e}; Re-extracting data')
1441 continue
1442 except GeoRestrictedError as e:
1443 msg = e.msg
1444 if e.countries:
1445 msg += '\nThis video is available in %s.' % ', '.join(
1446 map(ISO3166Utils.short2full, e.countries))
1447 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1448 self.report_error(msg)
1449 except ExtractorError as e: # An error we somewhat expected
1450 self.report_error(str(e), e.format_traceback())
1451 except Exception as e:
1452 if self.params.get('ignoreerrors'):
1453 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1454 else:
1455 raise
1456 break
1457 return wrapper
1458
1459 def _wait_for_video(self, ie_result):
1460 if (not self.params.get('wait_for_video')
1461 or ie_result.get('_type', 'video') != 'video'
1462 or ie_result.get('formats') or ie_result.get('url')):
1463 return
1464
1465 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1466 last_msg = ''
1467
1468 def progress(msg):
1469 nonlocal last_msg
1470 full_msg = f'{msg}\n'
1471 if not self.params.get('noprogress'):
1472 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1473 elif last_msg:
1474 return
1475 self.to_screen(full_msg, skip_eol=True)
1476 last_msg = msg
1477
1478 min_wait, max_wait = self.params.get('wait_for_video')
1479 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1480 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1481 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1482 self.report_warning('Release time of video is not known')
1483 elif (diff or 0) <= 0:
1484 self.report_warning('Video should already be available according to extracted info')
1485 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1486 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1487
1488 wait_till = time.time() + diff
1489 try:
1490 while True:
1491 diff = wait_till - time.time()
1492 if diff <= 0:
1493 progress('')
1494 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1495 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1496 time.sleep(1)
1497 except KeyboardInterrupt:
1498 progress('')
1499 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1500 except BaseException as e:
1501 if not isinstance(e, ReExtractInfo):
1502 self.to_screen('')
1503 raise
1504
1505 @_handle_extraction_exceptions
1506 def __extract_info(self, url, ie, download, extra_info, process):
1507 ie_result = ie.extract(url)
1508 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1509 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1510 return
1511 if isinstance(ie_result, list):
1512 # Backwards compatibility: old IE result format
1513 ie_result = {
1514 '_type': 'compat_list',
1515 'entries': ie_result,
1516 }
1517 if extra_info.get('original_url'):
1518 ie_result.setdefault('original_url', extra_info['original_url'])
1519 self.add_default_extra_info(ie_result, ie, url)
1520 if process:
1521 self._wait_for_video(ie_result)
1522 return self.process_ie_result(ie_result, download, extra_info)
1523 else:
1524 return ie_result
1525
1526 def add_default_extra_info(self, ie_result, ie, url):
1527 if url is not None:
1528 self.add_extra_info(ie_result, {
1529 'webpage_url': url,
1530 'original_url': url,
1531 })
1532 webpage_url = ie_result.get('webpage_url')
1533 if webpage_url:
1534 self.add_extra_info(ie_result, {
1535 'webpage_url_basename': url_basename(webpage_url),
1536 'webpage_url_domain': get_domain(webpage_url),
1537 })
1538 if ie is not None:
1539 self.add_extra_info(ie_result, {
1540 'extractor': ie.IE_NAME,
1541 'extractor_key': ie.ie_key(),
1542 })
1543
1544 def process_ie_result(self, ie_result, download=True, extra_info=None):
1545 """
1546 Take the result of the ie(may be modified) and resolve all unresolved
1547 references (URLs, playlist items).
1548
1549 It will also download the videos if 'download'.
1550 Returns the resolved ie_result.
1551 """
1552 if extra_info is None:
1553 extra_info = {}
1554 result_type = ie_result.get('_type', 'video')
1555
1556 if result_type in ('url', 'url_transparent'):
1557 ie_result['url'] = sanitize_url(ie_result['url'])
1558 if ie_result.get('original_url'):
1559 extra_info.setdefault('original_url', ie_result['original_url'])
1560
1561 extract_flat = self.params.get('extract_flat', False)
1562 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1563 or extract_flat is True):
1564 info_copy = ie_result.copy()
1565 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1566 if ie and not ie_result.get('id'):
1567 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1568 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1569 self.add_extra_info(info_copy, extra_info)
1570 info_copy, _ = self.pre_process(info_copy)
1571 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1572 self._raise_pending_errors(info_copy)
1573 if self.params.get('force_write_download_archive', False):
1574 self.record_download_archive(info_copy)
1575 return ie_result
1576
1577 if result_type == 'video':
1578 self.add_extra_info(ie_result, extra_info)
1579 ie_result = self.process_video_result(ie_result, download=download)
1580 self._raise_pending_errors(ie_result)
1581 additional_urls = (ie_result or {}).get('additional_urls')
1582 if additional_urls:
1583 # TODO: Improve MetadataParserPP to allow setting a list
1584 if isinstance(additional_urls, str):
1585 additional_urls = [additional_urls]
1586 self.to_screen(
1587 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1588 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1589 ie_result['additional_entries'] = [
1590 self.extract_info(
1591 url, download, extra_info=extra_info,
1592 force_generic_extractor=self.params.get('force_generic_extractor'))
1593 for url in additional_urls
1594 ]
1595 return ie_result
1596 elif result_type == 'url':
1597 # We have to add extra_info to the results because it may be
1598 # contained in a playlist
1599 return self.extract_info(
1600 ie_result['url'], download,
1601 ie_key=ie_result.get('ie_key'),
1602 extra_info=extra_info)
1603 elif result_type == 'url_transparent':
1604 # Use the information from the embedding page
1605 info = self.extract_info(
1606 ie_result['url'], ie_key=ie_result.get('ie_key'),
1607 extra_info=extra_info, download=False, process=False)
1608
1609 # extract_info may return None when ignoreerrors is enabled and
1610 # extraction failed with an error, don't crash and return early
1611 # in this case
1612 if not info:
1613 return info
1614
1615 exempted_fields = {'_type', 'url', 'ie_key'}
1616 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1617 # For video clips, the id etc of the clip extractor should be used
1618 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1619
1620 new_result = info.copy()
1621 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1622
1623 # Extracted info may not be a video result (i.e.
1624 # info.get('_type', 'video') != video) but rather an url or
1625 # url_transparent. In such cases outer metadata (from ie_result)
1626 # should be propagated to inner one (info). For this to happen
1627 # _type of info should be overridden with url_transparent. This
1628 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1629 if new_result.get('_type') == 'url':
1630 new_result['_type'] = 'url_transparent'
1631
1632 return self.process_ie_result(
1633 new_result, download=download, extra_info=extra_info)
1634 elif result_type in ('playlist', 'multi_video'):
1635 # Protect from infinite recursion due to recursively nested playlists
1636 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1637 webpage_url = ie_result['webpage_url']
1638 if webpage_url in self._playlist_urls:
1639 self.to_screen(
1640 '[download] Skipping already downloaded playlist: %s'
1641 % ie_result.get('title') or ie_result.get('id'))
1642 return
1643
1644 self._playlist_level += 1
1645 self._playlist_urls.add(webpage_url)
1646 self._fill_common_fields(ie_result, False)
1647 self._sanitize_thumbnails(ie_result)
1648 try:
1649 return self.__process_playlist(ie_result, download)
1650 finally:
1651 self._playlist_level -= 1
1652 if not self._playlist_level:
1653 self._playlist_urls.clear()
1654 elif result_type == 'compat_list':
1655 self.report_warning(
1656 'Extractor %s returned a compat_list result. '
1657 'It needs to be updated.' % ie_result.get('extractor'))
1658
1659 def _fixup(r):
1660 self.add_extra_info(r, {
1661 'extractor': ie_result['extractor'],
1662 'webpage_url': ie_result['webpage_url'],
1663 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1664 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1665 'extractor_key': ie_result['extractor_key'],
1666 })
1667 return r
1668 ie_result['entries'] = [
1669 self.process_ie_result(_fixup(r), download, extra_info)
1670 for r in ie_result['entries']
1671 ]
1672 return ie_result
1673 else:
1674 raise Exception('Invalid result type: %s' % result_type)
1675
1676 def _ensure_dir_exists(self, path):
1677 return make_dir(path, self.report_error)
1678
1679 @staticmethod
1680 def _playlist_infodict(ie_result, **kwargs):
1681 return {
1682 **ie_result,
1683 'playlist': ie_result.get('title') or ie_result.get('id'),
1684 'playlist_id': ie_result.get('id'),
1685 'playlist_title': ie_result.get('title'),
1686 'playlist_uploader': ie_result.get('uploader'),
1687 'playlist_uploader_id': ie_result.get('uploader_id'),
1688 'playlist_index': 0,
1689 **kwargs,
1690 }
1691
1692 def __process_playlist(self, ie_result, download):
1693 """Process each entry in the playlist"""
1694 assert ie_result['_type'] in ('playlist', 'multi_video')
1695
1696 title = ie_result.get('title') or ie_result.get('id') or '<Untitled>'
1697 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1698
1699 all_entries = PlaylistEntries(self, ie_result)
1700 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1701
1702 lazy = self.params.get('lazy_playlist')
1703 if lazy:
1704 resolved_entries, n_entries = [], 'N/A'
1705 ie_result['requested_entries'], ie_result['entries'] = None, None
1706 else:
1707 entries = resolved_entries = list(entries)
1708 n_entries = len(resolved_entries)
1709 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1710 if not ie_result.get('playlist_count'):
1711 # Better to do this after potentially exhausting entries
1712 ie_result['playlist_count'] = all_entries.get_full_count()
1713
1714 _infojson_written = False
1715 write_playlist_files = self.params.get('allow_playlist_files', True)
1716 if write_playlist_files and self.params.get('list_thumbnails'):
1717 self.list_thumbnails(ie_result)
1718 if write_playlist_files and not self.params.get('simulate'):
1719 ie_copy = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1720 _infojson_written = self._write_info_json(
1721 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1722 if _infojson_written is None:
1723 return
1724 if self._write_description('playlist', ie_result,
1725 self.prepare_filename(ie_copy, 'pl_description')) is None:
1726 return
1727 # TODO: This should be passed to ThumbnailsConvertor if necessary
1728 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1729
1730 if lazy:
1731 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1732 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1733 elif self.params.get('playlistreverse'):
1734 entries.reverse()
1735 elif self.params.get('playlistrandom'):
1736 random.shuffle(entries)
1737
1738 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} videos'
1739 f'{format_field(ie_result, "playlist_count", " of %s")}')
1740
1741 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1742 if self.params.get('extract_flat') == 'discard_in_playlist':
1743 keep_resolved_entries = ie_result['_type'] != 'playlist'
1744 if keep_resolved_entries:
1745 self.write_debug('The information of all playlist entries will be held in memory')
1746
1747 failures = 0
1748 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1749 for i, (playlist_index, entry) in enumerate(entries):
1750 if lazy:
1751 resolved_entries.append((playlist_index, entry))
1752
1753 # TODO: Add auto-generated fields
1754 if not entry or self._match_entry(entry, incomplete=True) is not None:
1755 continue
1756
1757 self.to_screen('[download] Downloading video %s of %s' % (
1758 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1759
1760 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1761 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1762 playlist_index = ie_result['requested_entries'][i]
1763
1764 entry_result = self.__process_iterable_entry(entry, download, {
1765 'n_entries': int_or_none(n_entries),
1766 '__last_playlist_index': max(ie_result['requested_entries'] or (0, 0)),
1767 'playlist_count': ie_result.get('playlist_count'),
1768 'playlist_index': playlist_index,
1769 'playlist_autonumber': i + 1,
1770 'playlist': title,
1771 'playlist_id': ie_result.get('id'),
1772 'playlist_title': ie_result.get('title'),
1773 'playlist_uploader': ie_result.get('uploader'),
1774 'playlist_uploader_id': ie_result.get('uploader_id'),
1775 'extractor': ie_result['extractor'],
1776 'webpage_url': ie_result['webpage_url'],
1777 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1778 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1779 'extractor_key': ie_result['extractor_key'],
1780 })
1781 if not entry_result:
1782 failures += 1
1783 if failures >= max_failures:
1784 self.report_error(
1785 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1786 break
1787 if keep_resolved_entries:
1788 resolved_entries[i] = (playlist_index, entry_result)
1789
1790 # Update with processed data
1791 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1792
1793 # Write the updated info to json
1794 if _infojson_written is True and self._write_info_json(
1795 'updated playlist', ie_result,
1796 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1797 return
1798
1799 ie_result = self.run_all_pps('playlist', ie_result)
1800 self.to_screen(f'[download] Finished downloading playlist: {title}')
1801 return ie_result
1802
1803 @_handle_extraction_exceptions
1804 def __process_iterable_entry(self, entry, download, extra_info):
1805 return self.process_ie_result(
1806 entry, download=download, extra_info=extra_info)
1807
1808 def _build_format_filter(self, filter_spec):
1809 " Returns a function to filter the formats according to the filter_spec "
1810
1811 OPERATORS = {
1812 '<': operator.lt,
1813 '<=': operator.le,
1814 '>': operator.gt,
1815 '>=': operator.ge,
1816 '=': operator.eq,
1817 '!=': operator.ne,
1818 }
1819 operator_rex = re.compile(r'''(?x)\s*
1820 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1821 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1822 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1823 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1824 m = operator_rex.fullmatch(filter_spec)
1825 if m:
1826 try:
1827 comparison_value = int(m.group('value'))
1828 except ValueError:
1829 comparison_value = parse_filesize(m.group('value'))
1830 if comparison_value is None:
1831 comparison_value = parse_filesize(m.group('value') + 'B')
1832 if comparison_value is None:
1833 raise ValueError(
1834 'Invalid value %r in format specification %r' % (
1835 m.group('value'), filter_spec))
1836 op = OPERATORS[m.group('op')]
1837
1838 if not m:
1839 STR_OPERATORS = {
1840 '=': operator.eq,
1841 '^=': lambda attr, value: attr.startswith(value),
1842 '$=': lambda attr, value: attr.endswith(value),
1843 '*=': lambda attr, value: value in attr,
1844 '~=': lambda attr, value: value.search(attr) is not None
1845 }
1846 str_operator_rex = re.compile(r'''(?x)\s*
1847 (?P<key>[a-zA-Z0-9._-]+)\s*
1848 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1849 (?P<quote>["'])?
1850 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1851 (?(quote)(?P=quote))\s*
1852 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1853 m = str_operator_rex.fullmatch(filter_spec)
1854 if m:
1855 if m.group('op') == '~=':
1856 comparison_value = re.compile(m.group('value'))
1857 else:
1858 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1859 str_op = STR_OPERATORS[m.group('op')]
1860 if m.group('negation'):
1861 op = lambda attr, value: not str_op(attr, value)
1862 else:
1863 op = str_op
1864
1865 if not m:
1866 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1867
1868 def _filter(f):
1869 actual_value = f.get(m.group('key'))
1870 if actual_value is None:
1871 return m.group('none_inclusive')
1872 return op(actual_value, comparison_value)
1873 return _filter
1874
1875 def _check_formats(self, formats):
1876 for f in formats:
1877 self.to_screen('[info] Testing format %s' % f['format_id'])
1878 path = self.get_output_path('temp')
1879 if not self._ensure_dir_exists(f'{path}/'):
1880 continue
1881 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1882 temp_file.close()
1883 try:
1884 success, _ = self.dl(temp_file.name, f, test=True)
1885 except (DownloadError, OSError, ValueError) + network_exceptions:
1886 success = False
1887 finally:
1888 if os.path.exists(temp_file.name):
1889 try:
1890 os.remove(temp_file.name)
1891 except OSError:
1892 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1893 if success:
1894 yield f
1895 else:
1896 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1897
1898 def _default_format_spec(self, info_dict, download=True):
1899
1900 def can_merge():
1901 merger = FFmpegMergerPP(self)
1902 return merger.available and merger.can_merge()
1903
1904 prefer_best = (
1905 not self.params.get('simulate')
1906 and download
1907 and (
1908 not can_merge()
1909 or info_dict.get('is_live') and not self.params.get('live_from_start')
1910 or self.params['outtmpl']['default'] == '-'))
1911 compat = (
1912 prefer_best
1913 or self.params.get('allow_multiple_audio_streams', False)
1914 or 'format-spec' in self.params['compat_opts'])
1915
1916 return (
1917 'best/bestvideo+bestaudio' if prefer_best
1918 else 'bestvideo*+bestaudio/best' if not compat
1919 else 'bestvideo+bestaudio/best')
1920
1921 def build_format_selector(self, format_spec):
1922 def syntax_error(note, start):
1923 message = (
1924 'Invalid format specification: '
1925 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1926 return SyntaxError(message)
1927
1928 PICKFIRST = 'PICKFIRST'
1929 MERGE = 'MERGE'
1930 SINGLE = 'SINGLE'
1931 GROUP = 'GROUP'
1932 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1933
1934 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1935 'video': self.params.get('allow_multiple_video_streams', False)}
1936
1937 check_formats = self.params.get('check_formats') == 'selected'
1938
1939 def _parse_filter(tokens):
1940 filter_parts = []
1941 for type, string, start, _, _ in tokens:
1942 if type == tokenize.OP and string == ']':
1943 return ''.join(filter_parts)
1944 else:
1945 filter_parts.append(string)
1946
1947 def _remove_unused_ops(tokens):
1948 # Remove operators that we don't use and join them with the surrounding strings
1949 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1950 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1951 last_string, last_start, last_end, last_line = None, None, None, None
1952 for type, string, start, end, line in tokens:
1953 if type == tokenize.OP and string == '[':
1954 if last_string:
1955 yield tokenize.NAME, last_string, last_start, last_end, last_line
1956 last_string = None
1957 yield type, string, start, end, line
1958 # everything inside brackets will be handled by _parse_filter
1959 for type, string, start, end, line in tokens:
1960 yield type, string, start, end, line
1961 if type == tokenize.OP and string == ']':
1962 break
1963 elif type == tokenize.OP and string in ALLOWED_OPS:
1964 if last_string:
1965 yield tokenize.NAME, last_string, last_start, last_end, last_line
1966 last_string = None
1967 yield type, string, start, end, line
1968 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1969 if not last_string:
1970 last_string = string
1971 last_start = start
1972 last_end = end
1973 else:
1974 last_string += string
1975 if last_string:
1976 yield tokenize.NAME, last_string, last_start, last_end, last_line
1977
1978 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1979 selectors = []
1980 current_selector = None
1981 for type, string, start, _, _ in tokens:
1982 # ENCODING is only defined in python 3.x
1983 if type == getattr(tokenize, 'ENCODING', None):
1984 continue
1985 elif type in [tokenize.NAME, tokenize.NUMBER]:
1986 current_selector = FormatSelector(SINGLE, string, [])
1987 elif type == tokenize.OP:
1988 if string == ')':
1989 if not inside_group:
1990 # ')' will be handled by the parentheses group
1991 tokens.restore_last_token()
1992 break
1993 elif inside_merge and string in ['/', ',']:
1994 tokens.restore_last_token()
1995 break
1996 elif inside_choice and string == ',':
1997 tokens.restore_last_token()
1998 break
1999 elif string == ',':
2000 if not current_selector:
2001 raise syntax_error('"," must follow a format selector', start)
2002 selectors.append(current_selector)
2003 current_selector = None
2004 elif string == '/':
2005 if not current_selector:
2006 raise syntax_error('"/" must follow a format selector', start)
2007 first_choice = current_selector
2008 second_choice = _parse_format_selection(tokens, inside_choice=True)
2009 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2010 elif string == '[':
2011 if not current_selector:
2012 current_selector = FormatSelector(SINGLE, 'best', [])
2013 format_filter = _parse_filter(tokens)
2014 current_selector.filters.append(format_filter)
2015 elif string == '(':
2016 if current_selector:
2017 raise syntax_error('Unexpected "("', start)
2018 group = _parse_format_selection(tokens, inside_group=True)
2019 current_selector = FormatSelector(GROUP, group, [])
2020 elif string == '+':
2021 if not current_selector:
2022 raise syntax_error('Unexpected "+"', start)
2023 selector_1 = current_selector
2024 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2025 if not selector_2:
2026 raise syntax_error('Expected a selector', start)
2027 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2028 else:
2029 raise syntax_error(f'Operator not recognized: "{string}"', start)
2030 elif type == tokenize.ENDMARKER:
2031 break
2032 if current_selector:
2033 selectors.append(current_selector)
2034 return selectors
2035
2036 def _merge(formats_pair):
2037 format_1, format_2 = formats_pair
2038
2039 formats_info = []
2040 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2041 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2042
2043 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2044 get_no_more = {'video': False, 'audio': False}
2045 for (i, fmt_info) in enumerate(formats_info):
2046 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2047 formats_info.pop(i)
2048 continue
2049 for aud_vid in ['audio', 'video']:
2050 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2051 if get_no_more[aud_vid]:
2052 formats_info.pop(i)
2053 break
2054 get_no_more[aud_vid] = True
2055
2056 if len(formats_info) == 1:
2057 return formats_info[0]
2058
2059 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2060 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2061
2062 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2063 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2064
2065 output_ext = self.params.get('merge_output_format')
2066 if not output_ext:
2067 if the_only_video:
2068 output_ext = the_only_video['ext']
2069 elif the_only_audio and not video_fmts:
2070 output_ext = the_only_audio['ext']
2071 else:
2072 output_ext = 'mkv'
2073
2074 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2075
2076 new_dict = {
2077 'requested_formats': formats_info,
2078 'format': '+'.join(filtered('format')),
2079 'format_id': '+'.join(filtered('format_id')),
2080 'ext': output_ext,
2081 'protocol': '+'.join(map(determine_protocol, formats_info)),
2082 'language': '+'.join(orderedSet(filtered('language'))) or None,
2083 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2084 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2085 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2086 }
2087
2088 if the_only_video:
2089 new_dict.update({
2090 'width': the_only_video.get('width'),
2091 'height': the_only_video.get('height'),
2092 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2093 'fps': the_only_video.get('fps'),
2094 'dynamic_range': the_only_video.get('dynamic_range'),
2095 'vcodec': the_only_video.get('vcodec'),
2096 'vbr': the_only_video.get('vbr'),
2097 'stretched_ratio': the_only_video.get('stretched_ratio'),
2098 })
2099
2100 if the_only_audio:
2101 new_dict.update({
2102 'acodec': the_only_audio.get('acodec'),
2103 'abr': the_only_audio.get('abr'),
2104 'asr': the_only_audio.get('asr'),
2105 })
2106
2107 return new_dict
2108
2109 def _check_formats(formats):
2110 if not check_formats:
2111 yield from formats
2112 return
2113 yield from self._check_formats(formats)
2114
2115 def _build_selector_function(selector):
2116 if isinstance(selector, list): # ,
2117 fs = [_build_selector_function(s) for s in selector]
2118
2119 def selector_function(ctx):
2120 for f in fs:
2121 yield from f(ctx)
2122 return selector_function
2123
2124 elif selector.type == GROUP: # ()
2125 selector_function = _build_selector_function(selector.selector)
2126
2127 elif selector.type == PICKFIRST: # /
2128 fs = [_build_selector_function(s) for s in selector.selector]
2129
2130 def selector_function(ctx):
2131 for f in fs:
2132 picked_formats = list(f(ctx))
2133 if picked_formats:
2134 return picked_formats
2135 return []
2136
2137 elif selector.type == MERGE: # +
2138 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2139
2140 def selector_function(ctx):
2141 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2142 yield _merge(pair)
2143
2144 elif selector.type == SINGLE: # atom
2145 format_spec = selector.selector or 'best'
2146
2147 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2148 if format_spec == 'all':
2149 def selector_function(ctx):
2150 yield from _check_formats(ctx['formats'][::-1])
2151 elif format_spec == 'mergeall':
2152 def selector_function(ctx):
2153 formats = list(_check_formats(
2154 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2155 if not formats:
2156 return
2157 merged_format = formats[-1]
2158 for f in formats[-2::-1]:
2159 merged_format = _merge((merged_format, f))
2160 yield merged_format
2161
2162 else:
2163 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2164 mobj = re.match(
2165 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2166 format_spec)
2167 if mobj is not None:
2168 format_idx = int_or_none(mobj.group('n'), default=1)
2169 format_reverse = mobj.group('bw')[0] == 'b'
2170 format_type = (mobj.group('type') or [None])[0]
2171 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2172 format_modified = mobj.group('mod') is not None
2173
2174 format_fallback = not format_type and not format_modified # for b, w
2175 _filter_f = (
2176 (lambda f: f.get('%scodec' % format_type) != 'none')
2177 if format_type and format_modified # bv*, ba*, wv*, wa*
2178 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2179 if format_type # bv, ba, wv, wa
2180 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2181 if not format_modified # b, w
2182 else lambda f: True) # b*, w*
2183 filter_f = lambda f: _filter_f(f) and (
2184 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2185 else:
2186 if format_spec in self._format_selection_exts['audio']:
2187 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2188 elif format_spec in self._format_selection_exts['video']:
2189 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2190 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2191 elif format_spec in self._format_selection_exts['storyboards']:
2192 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2193 else:
2194 filter_f = lambda f: f.get('format_id') == format_spec # id
2195
2196 def selector_function(ctx):
2197 formats = list(ctx['formats'])
2198 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2199 if not matches:
2200 if format_fallback and ctx['incomplete_formats']:
2201 # for extractors with incomplete formats (audio only (soundcloud)
2202 # or video only (imgur)) best/worst will fallback to
2203 # best/worst {video,audio}-only format
2204 matches = formats
2205 elif seperate_fallback and not ctx['has_merged_format']:
2206 # for compatibility with youtube-dl when there is no pre-merged format
2207 matches = list(filter(seperate_fallback, formats))
2208 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2209 try:
2210 yield matches[format_idx - 1]
2211 except LazyList.IndexError:
2212 return
2213
2214 filters = [self._build_format_filter(f) for f in selector.filters]
2215
2216 def final_selector(ctx):
2217 ctx_copy = dict(ctx)
2218 for _filter in filters:
2219 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2220 return selector_function(ctx_copy)
2221 return final_selector
2222
2223 stream = io.BytesIO(format_spec.encode())
2224 try:
2225 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2226 except tokenize.TokenError:
2227 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2228
2229 class TokenIterator:
2230 def __init__(self, tokens):
2231 self.tokens = tokens
2232 self.counter = 0
2233
2234 def __iter__(self):
2235 return self
2236
2237 def __next__(self):
2238 if self.counter >= len(self.tokens):
2239 raise StopIteration()
2240 value = self.tokens[self.counter]
2241 self.counter += 1
2242 return value
2243
2244 next = __next__
2245
2246 def restore_last_token(self):
2247 self.counter -= 1
2248
2249 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2250 return _build_selector_function(parsed_selector)
2251
2252 def _calc_headers(self, info_dict):
2253 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2254
2255 cookies = self._calc_cookies(info_dict['url'])
2256 if cookies:
2257 res['Cookie'] = cookies
2258
2259 if 'X-Forwarded-For' not in res:
2260 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2261 if x_forwarded_for_ip:
2262 res['X-Forwarded-For'] = x_forwarded_for_ip
2263
2264 return res
2265
2266 def _calc_cookies(self, url):
2267 pr = sanitized_Request(url)
2268 self.cookiejar.add_cookie_header(pr)
2269 return pr.get_header('Cookie')
2270
2271 def _sort_thumbnails(self, thumbnails):
2272 thumbnails.sort(key=lambda t: (
2273 t.get('preference') if t.get('preference') is not None else -1,
2274 t.get('width') if t.get('width') is not None else -1,
2275 t.get('height') if t.get('height') is not None else -1,
2276 t.get('id') if t.get('id') is not None else '',
2277 t.get('url')))
2278
2279 def _sanitize_thumbnails(self, info_dict):
2280 thumbnails = info_dict.get('thumbnails')
2281 if thumbnails is None:
2282 thumbnail = info_dict.get('thumbnail')
2283 if thumbnail:
2284 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2285 if not thumbnails:
2286 return
2287
2288 def check_thumbnails(thumbnails):
2289 for t in thumbnails:
2290 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2291 try:
2292 self.urlopen(HEADRequest(t['url']))
2293 except network_exceptions as err:
2294 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2295 continue
2296 yield t
2297
2298 self._sort_thumbnails(thumbnails)
2299 for i, t in enumerate(thumbnails):
2300 if t.get('id') is None:
2301 t['id'] = '%d' % i
2302 if t.get('width') and t.get('height'):
2303 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2304 t['url'] = sanitize_url(t['url'])
2305
2306 if self.params.get('check_formats') is True:
2307 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2308 else:
2309 info_dict['thumbnails'] = thumbnails
2310
2311 def _fill_common_fields(self, info_dict, is_video=True):
2312 # TODO: move sanitization here
2313 if is_video:
2314 # playlists are allowed to lack "title"
2315 title = info_dict.get('title', NO_DEFAULT)
2316 if title is NO_DEFAULT:
2317 raise ExtractorError('Missing "title" field in extractor result',
2318 video_id=info_dict['id'], ie=info_dict['extractor'])
2319 info_dict['fulltitle'] = title
2320 if not title:
2321 if title == '':
2322 self.write_debug('Extractor gave empty title. Creating a generic title')
2323 else:
2324 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2325 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2326
2327 if info_dict.get('duration') is not None:
2328 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2329
2330 for ts_key, date_key in (
2331 ('timestamp', 'upload_date'),
2332 ('release_timestamp', 'release_date'),
2333 ('modified_timestamp', 'modified_date'),
2334 ):
2335 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2336 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2337 # see http://bugs.python.org/issue1646728)
2338 with contextlib.suppress(ValueError, OverflowError, OSError):
2339 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2340 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2341
2342 live_keys = ('is_live', 'was_live')
2343 live_status = info_dict.get('live_status')
2344 if live_status is None:
2345 for key in live_keys:
2346 if info_dict.get(key) is False:
2347 continue
2348 if info_dict.get(key):
2349 live_status = key
2350 break
2351 if all(info_dict.get(key) is False for key in live_keys):
2352 live_status = 'not_live'
2353 if live_status:
2354 info_dict['live_status'] = live_status
2355 for key in live_keys:
2356 if info_dict.get(key) is None:
2357 info_dict[key] = (live_status == key)
2358
2359 # Auto generate title fields corresponding to the *_number fields when missing
2360 # in order to always have clean titles. This is very common for TV series.
2361 for field in ('chapter', 'season', 'episode'):
2362 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2363 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2364
2365 def _raise_pending_errors(self, info):
2366 err = info.pop('__pending_error', None)
2367 if err:
2368 self.report_error(err, tb=False)
2369
2370 def process_video_result(self, info_dict, download=True):
2371 assert info_dict.get('_type', 'video') == 'video'
2372 self._num_videos += 1
2373
2374 if 'id' not in info_dict:
2375 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2376 elif not info_dict.get('id'):
2377 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2378
2379 def report_force_conversion(field, field_not, conversion):
2380 self.report_warning(
2381 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2382 % (field, field_not, conversion))
2383
2384 def sanitize_string_field(info, string_field):
2385 field = info.get(string_field)
2386 if field is None or isinstance(field, str):
2387 return
2388 report_force_conversion(string_field, 'a string', 'string')
2389 info[string_field] = str(field)
2390
2391 def sanitize_numeric_fields(info):
2392 for numeric_field in self._NUMERIC_FIELDS:
2393 field = info.get(numeric_field)
2394 if field is None or isinstance(field, (int, float)):
2395 continue
2396 report_force_conversion(numeric_field, 'numeric', 'int')
2397 info[numeric_field] = int_or_none(field)
2398
2399 sanitize_string_field(info_dict, 'id')
2400 sanitize_numeric_fields(info_dict)
2401 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2402 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2403 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2404 self.report_warning('"duration" field is negative, there is an error in extractor')
2405
2406 chapters = info_dict.get('chapters') or []
2407 if chapters and chapters[0].get('start_time'):
2408 chapters.insert(0, {'start_time': 0})
2409
2410 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2411 for idx, (prev, current, next_) in enumerate(zip(
2412 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2413 if current.get('start_time') is None:
2414 current['start_time'] = prev.get('end_time')
2415 if not current.get('end_time'):
2416 current['end_time'] = next_.get('start_time')
2417 if not current.get('title'):
2418 current['title'] = f'<Untitled Chapter {idx}>'
2419
2420 if 'playlist' not in info_dict:
2421 # It isn't part of a playlist
2422 info_dict['playlist'] = None
2423 info_dict['playlist_index'] = None
2424
2425 self._sanitize_thumbnails(info_dict)
2426
2427 thumbnail = info_dict.get('thumbnail')
2428 thumbnails = info_dict.get('thumbnails')
2429 if thumbnail:
2430 info_dict['thumbnail'] = sanitize_url(thumbnail)
2431 elif thumbnails:
2432 info_dict['thumbnail'] = thumbnails[-1]['url']
2433
2434 if info_dict.get('display_id') is None and 'id' in info_dict:
2435 info_dict['display_id'] = info_dict['id']
2436
2437 self._fill_common_fields(info_dict)
2438
2439 for cc_kind in ('subtitles', 'automatic_captions'):
2440 cc = info_dict.get(cc_kind)
2441 if cc:
2442 for _, subtitle in cc.items():
2443 for subtitle_format in subtitle:
2444 if subtitle_format.get('url'):
2445 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2446 if subtitle_format.get('ext') is None:
2447 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2448
2449 automatic_captions = info_dict.get('automatic_captions')
2450 subtitles = info_dict.get('subtitles')
2451
2452 info_dict['requested_subtitles'] = self.process_subtitles(
2453 info_dict['id'], subtitles, automatic_captions)
2454
2455 if info_dict.get('formats') is None:
2456 # There's only one format available
2457 formats = [info_dict]
2458 else:
2459 formats = info_dict['formats']
2460
2461 # or None ensures --clean-infojson removes it
2462 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2463 if not self.params.get('allow_unplayable_formats'):
2464 formats = [f for f in formats if not f.get('has_drm')]
2465 if info_dict['_has_drm'] and all(
2466 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2467 self.report_warning(
2468 'This video is DRM protected and only images are available for download. '
2469 'Use --list-formats to see them')
2470
2471 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2472 if not get_from_start:
2473 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2474 if info_dict.get('is_live') and formats:
2475 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2476 if get_from_start and not formats:
2477 self.raise_no_formats(info_dict, msg=(
2478 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2479 'If you want to download from the current time, use --no-live-from-start'))
2480
2481 if not formats:
2482 self.raise_no_formats(info_dict)
2483
2484 def is_wellformed(f):
2485 url = f.get('url')
2486 if not url:
2487 self.report_warning(
2488 '"url" field is missing or empty - skipping format, '
2489 'there is an error in extractor')
2490 return False
2491 if isinstance(url, bytes):
2492 sanitize_string_field(f, 'url')
2493 return True
2494
2495 # Filter out malformed formats for better extraction robustness
2496 formats = list(filter(is_wellformed, formats))
2497
2498 formats_dict = {}
2499
2500 # We check that all the formats have the format and format_id fields
2501 for i, format in enumerate(formats):
2502 sanitize_string_field(format, 'format_id')
2503 sanitize_numeric_fields(format)
2504 format['url'] = sanitize_url(format['url'])
2505 if not format.get('format_id'):
2506 format['format_id'] = str(i)
2507 else:
2508 # Sanitize format_id from characters used in format selector expression
2509 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2510 format_id = format['format_id']
2511 if format_id not in formats_dict:
2512 formats_dict[format_id] = []
2513 formats_dict[format_id].append(format)
2514
2515 # Make sure all formats have unique format_id
2516 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2517 for format_id, ambiguous_formats in formats_dict.items():
2518 ambigious_id = len(ambiguous_formats) > 1
2519 for i, format in enumerate(ambiguous_formats):
2520 if ambigious_id:
2521 format['format_id'] = '%s-%d' % (format_id, i)
2522 if format.get('ext') is None:
2523 format['ext'] = determine_ext(format['url']).lower()
2524 # Ensure there is no conflict between id and ext in format selection
2525 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2526 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2527 format['format_id'] = 'f%s' % format['format_id']
2528
2529 for i, format in enumerate(formats):
2530 if format.get('format') is None:
2531 format['format'] = '{id} - {res}{note}'.format(
2532 id=format['format_id'],
2533 res=self.format_resolution(format),
2534 note=format_field(format, 'format_note', ' (%s)'),
2535 )
2536 if format.get('protocol') is None:
2537 format['protocol'] = determine_protocol(format)
2538 if format.get('resolution') is None:
2539 format['resolution'] = self.format_resolution(format, default=None)
2540 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2541 format['dynamic_range'] = 'SDR'
2542 if (info_dict.get('duration') and format.get('tbr')
2543 and not format.get('filesize') and not format.get('filesize_approx')):
2544 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2545
2546 # Add HTTP headers, so that external programs can use them from the
2547 # json output
2548 full_format_info = info_dict.copy()
2549 full_format_info.update(format)
2550 format['http_headers'] = self._calc_headers(full_format_info)
2551 # Remove private housekeeping stuff
2552 if '__x_forwarded_for_ip' in info_dict:
2553 del info_dict['__x_forwarded_for_ip']
2554
2555 if self.params.get('check_formats') is True:
2556 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2557
2558 if not formats or formats[0] is not info_dict:
2559 # only set the 'formats' fields if the original info_dict list them
2560 # otherwise we end up with a circular reference, the first (and unique)
2561 # element in the 'formats' field in info_dict is info_dict itself,
2562 # which can't be exported to json
2563 info_dict['formats'] = formats
2564
2565 info_dict, _ = self.pre_process(info_dict)
2566
2567 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2568 return info_dict
2569
2570 self.post_extract(info_dict)
2571 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2572
2573 # The pre-processors may have modified the formats
2574 formats = info_dict.get('formats', [info_dict])
2575
2576 list_only = self.params.get('simulate') is None and (
2577 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2578 interactive_format_selection = not list_only and self.format_selector == '-'
2579 if self.params.get('list_thumbnails'):
2580 self.list_thumbnails(info_dict)
2581 if self.params.get('listsubtitles'):
2582 if 'automatic_captions' in info_dict:
2583 self.list_subtitles(
2584 info_dict['id'], automatic_captions, 'automatic captions')
2585 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2586 if self.params.get('listformats') or interactive_format_selection:
2587 self.list_formats(info_dict)
2588 if list_only:
2589 # Without this printing, -F --print-json will not work
2590 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2591 return info_dict
2592
2593 format_selector = self.format_selector
2594 if format_selector is None:
2595 req_format = self._default_format_spec(info_dict, download=download)
2596 self.write_debug('Default format spec: %s' % req_format)
2597 format_selector = self.build_format_selector(req_format)
2598
2599 while True:
2600 if interactive_format_selection:
2601 req_format = input(
2602 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2603 try:
2604 format_selector = self.build_format_selector(req_format)
2605 except SyntaxError as err:
2606 self.report_error(err, tb=False, is_error=False)
2607 continue
2608
2609 formats_to_download = list(format_selector({
2610 'formats': formats,
2611 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2612 'incomplete_formats': (
2613 # All formats are video-only or
2614 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2615 # all formats are audio-only
2616 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2617 }))
2618 if interactive_format_selection and not formats_to_download:
2619 self.report_error('Requested format is not available', tb=False, is_error=False)
2620 continue
2621 break
2622
2623 if not formats_to_download:
2624 if not self.params.get('ignore_no_formats_error'):
2625 raise ExtractorError(
2626 'Requested format is not available. Use --list-formats for a list of available formats',
2627 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2628 self.report_warning('Requested format is not available')
2629 # Process what we can, even without any available formats.
2630 formats_to_download = [{}]
2631
2632 requested_ranges = self.params.get('download_ranges')
2633 if requested_ranges:
2634 requested_ranges = tuple(requested_ranges(info_dict, self))
2635
2636 best_format, downloaded_formats = formats_to_download[-1], []
2637 if download:
2638 if best_format:
2639 def to_screen(*msg):
2640 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2641
2642 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2643 (f['format_id'] for f in formats_to_download))
2644 if requested_ranges:
2645 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2646 (f'{int(c["start_time"])}-{int(c["end_time"])}' for c in requested_ranges))
2647 max_downloads_reached = False
2648
2649 for fmt, chapter in itertools.product(formats_to_download, requested_ranges or [{}]):
2650 new_info = self._copy_infodict(info_dict)
2651 new_info.update(fmt)
2652 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2653 if chapter or offset:
2654 new_info.update({
2655 'section_start': offset + chapter.get('start_time', 0),
2656 'section_end': offset + min(chapter.get('end_time', duration), duration),
2657 'section_title': chapter.get('title'),
2658 'section_number': chapter.get('index'),
2659 })
2660 downloaded_formats.append(new_info)
2661 try:
2662 self.process_info(new_info)
2663 except MaxDownloadsReached:
2664 max_downloads_reached = True
2665 self._raise_pending_errors(new_info)
2666 # Remove copied info
2667 for key, val in tuple(new_info.items()):
2668 if info_dict.get(key) == val:
2669 new_info.pop(key)
2670 if max_downloads_reached:
2671 break
2672
2673 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2674 assert write_archive.issubset({True, False, 'ignore'})
2675 if True in write_archive and False not in write_archive:
2676 self.record_download_archive(info_dict)
2677
2678 info_dict['requested_downloads'] = downloaded_formats
2679 info_dict = self.run_all_pps('after_video', info_dict)
2680 if max_downloads_reached:
2681 raise MaxDownloadsReached()
2682
2683 # We update the info dict with the selected best quality format (backwards compatibility)
2684 info_dict.update(best_format)
2685 return info_dict
2686
2687 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2688 """Select the requested subtitles and their format"""
2689 available_subs, normal_sub_langs = {}, []
2690 if normal_subtitles and self.params.get('writesubtitles'):
2691 available_subs.update(normal_subtitles)
2692 normal_sub_langs = tuple(normal_subtitles.keys())
2693 if automatic_captions and self.params.get('writeautomaticsub'):
2694 for lang, cap_info in automatic_captions.items():
2695 if lang not in available_subs:
2696 available_subs[lang] = cap_info
2697
2698 if (not self.params.get('writesubtitles') and not
2699 self.params.get('writeautomaticsub') or not
2700 available_subs):
2701 return None
2702
2703 all_sub_langs = tuple(available_subs.keys())
2704 if self.params.get('allsubtitles', False):
2705 requested_langs = all_sub_langs
2706 elif self.params.get('subtitleslangs', False):
2707 # A list is used so that the order of languages will be the same as
2708 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2709 requested_langs = []
2710 for lang_re in self.params.get('subtitleslangs'):
2711 discard = lang_re[0] == '-'
2712 if discard:
2713 lang_re = lang_re[1:]
2714 if lang_re == 'all':
2715 if discard:
2716 requested_langs = []
2717 else:
2718 requested_langs.extend(all_sub_langs)
2719 continue
2720 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2721 if discard:
2722 for lang in current_langs:
2723 while lang in requested_langs:
2724 requested_langs.remove(lang)
2725 else:
2726 requested_langs.extend(current_langs)
2727 requested_langs = orderedSet(requested_langs)
2728 elif normal_sub_langs:
2729 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2730 else:
2731 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2732 if requested_langs:
2733 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2734
2735 formats_query = self.params.get('subtitlesformat', 'best')
2736 formats_preference = formats_query.split('/') if formats_query else []
2737 subs = {}
2738 for lang in requested_langs:
2739 formats = available_subs.get(lang)
2740 if formats is None:
2741 self.report_warning(f'{lang} subtitles not available for {video_id}')
2742 continue
2743 for ext in formats_preference:
2744 if ext == 'best':
2745 f = formats[-1]
2746 break
2747 matches = list(filter(lambda f: f['ext'] == ext, formats))
2748 if matches:
2749 f = matches[-1]
2750 break
2751 else:
2752 f = formats[-1]
2753 self.report_warning(
2754 'No subtitle format found matching "%s" for language %s, '
2755 'using %s' % (formats_query, lang, f['ext']))
2756 subs[lang] = f
2757 return subs
2758
2759 def _forceprint(self, key, info_dict):
2760 if info_dict is None:
2761 return
2762 info_copy = info_dict.copy()
2763 info_copy['formats_table'] = self.render_formats_table(info_dict)
2764 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2765 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2766 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2767
2768 def format_tmpl(tmpl):
2769 mobj = re.match(r'\w+(=?)$', tmpl)
2770 if mobj and mobj.group(1):
2771 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2772 elif mobj:
2773 return f'%({tmpl})s'
2774 return tmpl
2775
2776 for tmpl in self.params['forceprint'].get(key, []):
2777 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2778
2779 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2780 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2781 tmpl = format_tmpl(tmpl)
2782 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2783 if self._ensure_dir_exists(filename):
2784 with open(filename, 'a', encoding='utf-8') as f:
2785 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2786
2787 def __forced_printings(self, info_dict, filename, incomplete):
2788 def print_mandatory(field, actual_field=None):
2789 if actual_field is None:
2790 actual_field = field
2791 if (self.params.get('force%s' % field, False)
2792 and (not incomplete or info_dict.get(actual_field) is not None)):
2793 self.to_stdout(info_dict[actual_field])
2794
2795 def print_optional(field):
2796 if (self.params.get('force%s' % field, False)
2797 and info_dict.get(field) is not None):
2798 self.to_stdout(info_dict[field])
2799
2800 info_dict = info_dict.copy()
2801 if filename is not None:
2802 info_dict['filename'] = filename
2803 if info_dict.get('requested_formats') is not None:
2804 # For RTMP URLs, also include the playpath
2805 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2806 elif info_dict.get('url'):
2807 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2808
2809 if (self.params.get('forcejson')
2810 or self.params['forceprint'].get('video')
2811 or self.params['print_to_file'].get('video')):
2812 self.post_extract(info_dict)
2813 self._forceprint('video', info_dict)
2814
2815 print_mandatory('title')
2816 print_mandatory('id')
2817 print_mandatory('url', 'urls')
2818 print_optional('thumbnail')
2819 print_optional('description')
2820 print_optional('filename')
2821 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2822 self.to_stdout(formatSeconds(info_dict['duration']))
2823 print_mandatory('format')
2824
2825 if self.params.get('forcejson'):
2826 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2827
2828 def dl(self, name, info, subtitle=False, test=False):
2829 if not info.get('url'):
2830 self.raise_no_formats(info, True)
2831
2832 if test:
2833 verbose = self.params.get('verbose')
2834 params = {
2835 'test': True,
2836 'quiet': self.params.get('quiet') or not verbose,
2837 'verbose': verbose,
2838 'noprogress': not verbose,
2839 'nopart': True,
2840 'skip_unavailable_fragments': False,
2841 'keep_fragments': False,
2842 'overwrites': True,
2843 '_no_ytdl_file': True,
2844 }
2845 else:
2846 params = self.params
2847 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2848 if not test:
2849 for ph in self._progress_hooks:
2850 fd.add_progress_hook(ph)
2851 urls = '", "'.join(
2852 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2853 for f in info.get('requested_formats', []) or [info])
2854 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2855
2856 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2857 # But it may contain objects that are not deep-copyable
2858 new_info = self._copy_infodict(info)
2859 if new_info.get('http_headers') is None:
2860 new_info['http_headers'] = self._calc_headers(new_info)
2861 return fd.download(name, new_info, subtitle)
2862
2863 def existing_file(self, filepaths, *, default_overwrite=True):
2864 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2865 if existing_files and not self.params.get('overwrites', default_overwrite):
2866 return existing_files[0]
2867
2868 for file in existing_files:
2869 self.report_file_delete(file)
2870 os.remove(file)
2871 return None
2872
2873 def process_info(self, info_dict):
2874 """Process a single resolved IE result. (Modifies it in-place)"""
2875
2876 assert info_dict.get('_type', 'video') == 'video'
2877 original_infodict = info_dict
2878
2879 if 'format' not in info_dict and 'ext' in info_dict:
2880 info_dict['format'] = info_dict['ext']
2881
2882 # This is mostly just for backward compatibility of process_info
2883 # As a side-effect, this allows for format-specific filters
2884 if self._match_entry(info_dict) is not None:
2885 info_dict['__write_download_archive'] = 'ignore'
2886 return
2887
2888 # Does nothing under normal operation - for backward compatibility of process_info
2889 self.post_extract(info_dict)
2890 self._num_downloads += 1
2891
2892 # info_dict['_filename'] needs to be set for backward compatibility
2893 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2894 temp_filename = self.prepare_filename(info_dict, 'temp')
2895 files_to_move = {}
2896
2897 # Forced printings
2898 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2899
2900 def check_max_downloads():
2901 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
2902 raise MaxDownloadsReached()
2903
2904 if self.params.get('simulate'):
2905 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2906 check_max_downloads()
2907 return
2908
2909 if full_filename is None:
2910 return
2911 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2912 return
2913 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2914 return
2915
2916 if self._write_description('video', info_dict,
2917 self.prepare_filename(info_dict, 'description')) is None:
2918 return
2919
2920 sub_files = self._write_subtitles(info_dict, temp_filename)
2921 if sub_files is None:
2922 return
2923 files_to_move.update(dict(sub_files))
2924
2925 thumb_files = self._write_thumbnails(
2926 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2927 if thumb_files is None:
2928 return
2929 files_to_move.update(dict(thumb_files))
2930
2931 infofn = self.prepare_filename(info_dict, 'infojson')
2932 _infojson_written = self._write_info_json('video', info_dict, infofn)
2933 if _infojson_written:
2934 info_dict['infojson_filename'] = infofn
2935 # For backward compatibility, even though it was a private field
2936 info_dict['__infojson_filename'] = infofn
2937 elif _infojson_written is None:
2938 return
2939
2940 # Note: Annotations are deprecated
2941 annofn = None
2942 if self.params.get('writeannotations', False):
2943 annofn = self.prepare_filename(info_dict, 'annotation')
2944 if annofn:
2945 if not self._ensure_dir_exists(encodeFilename(annofn)):
2946 return
2947 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2948 self.to_screen('[info] Video annotations are already present')
2949 elif not info_dict.get('annotations'):
2950 self.report_warning('There are no annotations to write.')
2951 else:
2952 try:
2953 self.to_screen('[info] Writing video annotations to: ' + annofn)
2954 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2955 annofile.write(info_dict['annotations'])
2956 except (KeyError, TypeError):
2957 self.report_warning('There are no annotations to write.')
2958 except OSError:
2959 self.report_error('Cannot write annotations file: ' + annofn)
2960 return
2961
2962 # Write internet shortcut files
2963 def _write_link_file(link_type):
2964 url = try_get(info_dict['webpage_url'], iri_to_uri)
2965 if not url:
2966 self.report_warning(
2967 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2968 return True
2969 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2970 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2971 return False
2972 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2973 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2974 return True
2975 try:
2976 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2977 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2978 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2979 template_vars = {'url': url}
2980 if link_type == 'desktop':
2981 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2982 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2983 except OSError:
2984 self.report_error(f'Cannot write internet shortcut {linkfn}')
2985 return False
2986 return True
2987
2988 write_links = {
2989 'url': self.params.get('writeurllink'),
2990 'webloc': self.params.get('writewebloclink'),
2991 'desktop': self.params.get('writedesktoplink'),
2992 }
2993 if self.params.get('writelink'):
2994 link_type = ('webloc' if sys.platform == 'darwin'
2995 else 'desktop' if sys.platform.startswith('linux')
2996 else 'url')
2997 write_links[link_type] = True
2998
2999 if any(should_write and not _write_link_file(link_type)
3000 for link_type, should_write in write_links.items()):
3001 return
3002
3003 def replace_info_dict(new_info):
3004 nonlocal info_dict
3005 if new_info == info_dict:
3006 return
3007 info_dict.clear()
3008 info_dict.update(new_info)
3009
3010 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3011 replace_info_dict(new_info)
3012
3013 if self.params.get('skip_download'):
3014 info_dict['filepath'] = temp_filename
3015 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3016 info_dict['__files_to_move'] = files_to_move
3017 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3018 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3019 else:
3020 # Download
3021 info_dict.setdefault('__postprocessors', [])
3022 try:
3023
3024 def existing_video_file(*filepaths):
3025 ext = info_dict.get('ext')
3026 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3027 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3028 default_overwrite=False)
3029 if file:
3030 info_dict['ext'] = os.path.splitext(file)[1][1:]
3031 return file
3032
3033 fd, success = None, True
3034 if info_dict.get('protocol') or info_dict.get('url'):
3035 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3036 if fd is not FFmpegFD and (
3037 info_dict.get('section_start') or info_dict.get('section_end')):
3038 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3039 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3040 self.report_error(f'{msg}. Aborting')
3041 return
3042
3043 if info_dict.get('requested_formats') is not None:
3044
3045 def compatible_formats(formats):
3046 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3047 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3048 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3049 if len(video_formats) > 2 or len(audio_formats) > 2:
3050 return False
3051
3052 # Check extension
3053 exts = {format.get('ext') for format in formats}
3054 COMPATIBLE_EXTS = (
3055 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3056 {'webm'},
3057 )
3058 for ext_sets in COMPATIBLE_EXTS:
3059 if ext_sets.issuperset(exts):
3060 return True
3061 # TODO: Check acodec/vcodec
3062 return False
3063
3064 requested_formats = info_dict['requested_formats']
3065 old_ext = info_dict['ext']
3066 if self.params.get('merge_output_format') is None:
3067 if not compatible_formats(requested_formats):
3068 info_dict['ext'] = 'mkv'
3069 self.report_warning(
3070 'Requested formats are incompatible for merge and will be merged into mkv')
3071 if (info_dict['ext'] == 'webm'
3072 and info_dict.get('thumbnails')
3073 # check with type instead of pp_key, __name__, or isinstance
3074 # since we dont want any custom PPs to trigger this
3075 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3076 info_dict['ext'] = 'mkv'
3077 self.report_warning(
3078 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3079 new_ext = info_dict['ext']
3080
3081 def correct_ext(filename, ext=new_ext):
3082 if filename == '-':
3083 return filename
3084 filename_real_ext = os.path.splitext(filename)[1][1:]
3085 filename_wo_ext = (
3086 os.path.splitext(filename)[0]
3087 if filename_real_ext in (old_ext, new_ext)
3088 else filename)
3089 return f'{filename_wo_ext}.{ext}'
3090
3091 # Ensure filename always has a correct extension for successful merge
3092 full_filename = correct_ext(full_filename)
3093 temp_filename = correct_ext(temp_filename)
3094 dl_filename = existing_video_file(full_filename, temp_filename)
3095 info_dict['__real_download'] = False
3096
3097 merger = FFmpegMergerPP(self)
3098 downloaded = []
3099 if dl_filename is not None:
3100 self.report_file_already_downloaded(dl_filename)
3101 elif fd:
3102 for f in requested_formats if fd != FFmpegFD else []:
3103 f['filepath'] = fname = prepend_extension(
3104 correct_ext(temp_filename, info_dict['ext']),
3105 'f%s' % f['format_id'], info_dict['ext'])
3106 downloaded.append(fname)
3107 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3108 success, real_download = self.dl(temp_filename, info_dict)
3109 info_dict['__real_download'] = real_download
3110 else:
3111 if self.params.get('allow_unplayable_formats'):
3112 self.report_warning(
3113 'You have requested merging of multiple formats '
3114 'while also allowing unplayable formats to be downloaded. '
3115 'The formats won\'t be merged to prevent data corruption.')
3116 elif not merger.available:
3117 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3118 if not self.params.get('ignoreerrors'):
3119 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3120 return
3121 self.report_warning(f'{msg}. The formats won\'t be merged')
3122
3123 if temp_filename == '-':
3124 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3125 else 'but the formats are incompatible for simultaneous download' if merger.available
3126 else 'but ffmpeg is not installed')
3127 self.report_warning(
3128 f'You have requested downloading multiple formats to stdout {reason}. '
3129 'The formats will be streamed one after the other')
3130 fname = temp_filename
3131 for f in requested_formats:
3132 new_info = dict(info_dict)
3133 del new_info['requested_formats']
3134 new_info.update(f)
3135 if temp_filename != '-':
3136 fname = prepend_extension(
3137 correct_ext(temp_filename, new_info['ext']),
3138 'f%s' % f['format_id'], new_info['ext'])
3139 if not self._ensure_dir_exists(fname):
3140 return
3141 f['filepath'] = fname
3142 downloaded.append(fname)
3143 partial_success, real_download = self.dl(fname, new_info)
3144 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3145 success = success and partial_success
3146
3147 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3148 info_dict['__postprocessors'].append(merger)
3149 info_dict['__files_to_merge'] = downloaded
3150 # Even if there were no downloads, it is being merged only now
3151 info_dict['__real_download'] = True
3152 else:
3153 for file in downloaded:
3154 files_to_move[file] = None
3155 else:
3156 # Just a single file
3157 dl_filename = existing_video_file(full_filename, temp_filename)
3158 if dl_filename is None or dl_filename == temp_filename:
3159 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3160 # So we should try to resume the download
3161 success, real_download = self.dl(temp_filename, info_dict)
3162 info_dict['__real_download'] = real_download
3163 else:
3164 self.report_file_already_downloaded(dl_filename)
3165
3166 dl_filename = dl_filename or temp_filename
3167 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3168
3169 except network_exceptions as err:
3170 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3171 return
3172 except OSError as err:
3173 raise UnavailableVideoError(err)
3174 except (ContentTooShortError, ) as err:
3175 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3176 return
3177
3178 self._raise_pending_errors(info_dict)
3179 if success and full_filename != '-':
3180
3181 def fixup():
3182 do_fixup = True
3183 fixup_policy = self.params.get('fixup')
3184 vid = info_dict['id']
3185
3186 if fixup_policy in ('ignore', 'never'):
3187 return
3188 elif fixup_policy == 'warn':
3189 do_fixup = 'warn'
3190 elif fixup_policy != 'force':
3191 assert fixup_policy in ('detect_or_warn', None)
3192 if not info_dict.get('__real_download'):
3193 do_fixup = False
3194
3195 def ffmpeg_fixup(cndn, msg, cls):
3196 if not (do_fixup and cndn):
3197 return
3198 elif do_fixup == 'warn':
3199 self.report_warning(f'{vid}: {msg}')
3200 return
3201 pp = cls(self)
3202 if pp.available:
3203 info_dict['__postprocessors'].append(pp)
3204 else:
3205 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3206
3207 stretched_ratio = info_dict.get('stretched_ratio')
3208 ffmpeg_fixup(stretched_ratio not in (1, None),
3209 f'Non-uniform pixel ratio {stretched_ratio}',
3210 FFmpegFixupStretchedPP)
3211
3212 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3213 downloader = downloader.FD_NAME if downloader else None
3214
3215 ext = info_dict.get('ext')
3216 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3217 isinstance(pp, FFmpegVideoConvertorPP)
3218 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3219 ) for pp in self._pps['post_process'])
3220
3221 if not postprocessed_by_ffmpeg:
3222 ffmpeg_fixup(ext == 'm4a' and info_dict.get('container') == 'm4a_dash',
3223 'writing DASH m4a. Only some players support this container',
3224 FFmpegFixupM4aPP)
3225 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3226 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3227 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3228 FFmpegFixupM3u8PP)
3229 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3230 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3231
3232 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3233 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3234
3235 fixup()
3236 try:
3237 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3238 except PostProcessingError as err:
3239 self.report_error('Postprocessing: %s' % str(err))
3240 return
3241 try:
3242 for ph in self._post_hooks:
3243 ph(info_dict['filepath'])
3244 except Exception as err:
3245 self.report_error('post hooks: %s' % str(err))
3246 return
3247 info_dict['__write_download_archive'] = True
3248
3249 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3250 if self.params.get('force_write_download_archive'):
3251 info_dict['__write_download_archive'] = True
3252 check_max_downloads()
3253
3254 def __download_wrapper(self, func):
3255 @functools.wraps(func)
3256 def wrapper(*args, **kwargs):
3257 try:
3258 res = func(*args, **kwargs)
3259 except UnavailableVideoError as e:
3260 self.report_error(e)
3261 except DownloadCancelled as e:
3262 self.to_screen(f'[info] {e}')
3263 if not self.params.get('break_per_url'):
3264 raise
3265 else:
3266 if self.params.get('dump_single_json', False):
3267 self.post_extract(res)
3268 self.to_stdout(json.dumps(self.sanitize_info(res)))
3269 return wrapper
3270
3271 def download(self, url_list):
3272 """Download a given list of URLs."""
3273 url_list = variadic(url_list) # Passing a single URL is a common mistake
3274 outtmpl = self.params['outtmpl']['default']
3275 if (len(url_list) > 1
3276 and outtmpl != '-'
3277 and '%' not in outtmpl
3278 and self.params.get('max_downloads') != 1):
3279 raise SameFileError(outtmpl)
3280
3281 for url in url_list:
3282 self.__download_wrapper(self.extract_info)(
3283 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3284
3285 return self._download_retcode
3286
3287 def download_with_info_file(self, info_filename):
3288 with contextlib.closing(fileinput.FileInput(
3289 [info_filename], mode='r',
3290 openhook=fileinput.hook_encoded('utf-8'))) as f:
3291 # FileInput doesn't have a read method, we can't call json.load
3292 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3293 try:
3294 self.__download_wrapper(self.process_ie_result)(info, download=True)
3295 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3296 if not isinstance(e, EntryNotInPlaylist):
3297 self.to_stderr('\r')
3298 webpage_url = info.get('webpage_url')
3299 if webpage_url is not None:
3300 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3301 return self.download([webpage_url])
3302 else:
3303 raise
3304 return self._download_retcode
3305
3306 @staticmethod
3307 def sanitize_info(info_dict, remove_private_keys=False):
3308 ''' Sanitize the infodict for converting to json '''
3309 if info_dict is None:
3310 return info_dict
3311 info_dict.setdefault('epoch', int(time.time()))
3312 info_dict.setdefault('_type', 'video')
3313
3314 if remove_private_keys:
3315 reject = lambda k, v: v is None or k.startswith('__') or k in {
3316 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3317 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3318 }
3319 else:
3320 reject = lambda k, v: False
3321
3322 def filter_fn(obj):
3323 if isinstance(obj, dict):
3324 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3325 elif isinstance(obj, (list, tuple, set, LazyList)):
3326 return list(map(filter_fn, obj))
3327 elif obj is None or isinstance(obj, (str, int, float, bool)):
3328 return obj
3329 else:
3330 return repr(obj)
3331
3332 return filter_fn(info_dict)
3333
3334 @staticmethod
3335 def filter_requested_info(info_dict, actually_filter=True):
3336 ''' Alias of sanitize_info for backward compatibility '''
3337 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3338
3339 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3340 for filename in set(filter(None, files_to_delete)):
3341 if msg:
3342 self.to_screen(msg % filename)
3343 try:
3344 os.remove(filename)
3345 except OSError:
3346 self.report_warning(f'Unable to delete file {filename}')
3347 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3348 del info['__files_to_move'][filename]
3349
3350 @staticmethod
3351 def post_extract(info_dict):
3352 def actual_post_extract(info_dict):
3353 if info_dict.get('_type') in ('playlist', 'multi_video'):
3354 for video_dict in info_dict.get('entries', {}):
3355 actual_post_extract(video_dict or {})
3356 return
3357
3358 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3359 info_dict.update(post_extractor())
3360
3361 actual_post_extract(info_dict or {})
3362
3363 def run_pp(self, pp, infodict):
3364 files_to_delete = []
3365 if '__files_to_move' not in infodict:
3366 infodict['__files_to_move'] = {}
3367 try:
3368 files_to_delete, infodict = pp.run(infodict)
3369 except PostProcessingError as e:
3370 # Must be True and not 'only_download'
3371 if self.params.get('ignoreerrors') is True:
3372 self.report_error(e)
3373 return infodict
3374 raise
3375
3376 if not files_to_delete:
3377 return infodict
3378 if self.params.get('keepvideo', False):
3379 for f in files_to_delete:
3380 infodict['__files_to_move'].setdefault(f, '')
3381 else:
3382 self._delete_downloaded_files(
3383 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3384 return infodict
3385
3386 def run_all_pps(self, key, info, *, additional_pps=None):
3387 self._forceprint(key, info)
3388 for pp in (additional_pps or []) + self._pps[key]:
3389 info = self.run_pp(pp, info)
3390 return info
3391
3392 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3393 info = dict(ie_info)
3394 info['__files_to_move'] = files_to_move or {}
3395 try:
3396 info = self.run_all_pps(key, info)
3397 except PostProcessingError as err:
3398 msg = f'Preprocessing: {err}'
3399 info.setdefault('__pending_error', msg)
3400 self.report_error(msg, is_error=False)
3401 return info, info.pop('__files_to_move', None)
3402
3403 def post_process(self, filename, info, files_to_move=None):
3404 """Run all the postprocessors on the given file."""
3405 info['filepath'] = filename
3406 info['__files_to_move'] = files_to_move or {}
3407 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3408 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3409 del info['__files_to_move']
3410 return self.run_all_pps('after_move', info)
3411
3412 def _make_archive_id(self, info_dict):
3413 video_id = info_dict.get('id')
3414 if not video_id:
3415 return
3416 # Future-proof against any change in case
3417 # and backwards compatibility with prior versions
3418 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3419 if extractor is None:
3420 url = str_or_none(info_dict.get('url'))
3421 if not url:
3422 return
3423 # Try to find matching extractor for the URL and take its ie_key
3424 for ie_key, ie in self._ies.items():
3425 if ie.suitable(url):
3426 extractor = ie_key
3427 break
3428 else:
3429 return
3430 return f'{extractor.lower()} {video_id}'
3431
3432 def in_download_archive(self, info_dict):
3433 fn = self.params.get('download_archive')
3434 if fn is None:
3435 return False
3436
3437 vid_id = self._make_archive_id(info_dict)
3438 if not vid_id:
3439 return False # Incomplete video information
3440
3441 return vid_id in self.archive
3442
3443 def record_download_archive(self, info_dict):
3444 fn = self.params.get('download_archive')
3445 if fn is None:
3446 return
3447 vid_id = self._make_archive_id(info_dict)
3448 assert vid_id
3449 self.write_debug(f'Adding to archive: {vid_id}')
3450 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3451 archive_file.write(vid_id + '\n')
3452 self.archive.add(vid_id)
3453
3454 @staticmethod
3455 def format_resolution(format, default='unknown'):
3456 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3457 return 'audio only'
3458 if format.get('resolution') is not None:
3459 return format['resolution']
3460 if format.get('width') and format.get('height'):
3461 return '%dx%d' % (format['width'], format['height'])
3462 elif format.get('height'):
3463 return '%sp' % format['height']
3464 elif format.get('width'):
3465 return '%dx?' % format['width']
3466 return default
3467
3468 def _list_format_headers(self, *headers):
3469 if self.params.get('listformats_table', True) is not False:
3470 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3471 return headers
3472
3473 def _format_note(self, fdict):
3474 res = ''
3475 if fdict.get('ext') in ['f4f', 'f4m']:
3476 res += '(unsupported)'
3477 if fdict.get('language'):
3478 if res:
3479 res += ' '
3480 res += '[%s]' % fdict['language']
3481 if fdict.get('format_note') is not None:
3482 if res:
3483 res += ' '
3484 res += fdict['format_note']
3485 if fdict.get('tbr') is not None:
3486 if res:
3487 res += ', '
3488 res += '%4dk' % fdict['tbr']
3489 if fdict.get('container') is not None:
3490 if res:
3491 res += ', '
3492 res += '%s container' % fdict['container']
3493 if (fdict.get('vcodec') is not None
3494 and fdict.get('vcodec') != 'none'):
3495 if res:
3496 res += ', '
3497 res += fdict['vcodec']
3498 if fdict.get('vbr') is not None:
3499 res += '@'
3500 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3501 res += 'video@'
3502 if fdict.get('vbr') is not None:
3503 res += '%4dk' % fdict['vbr']
3504 if fdict.get('fps') is not None:
3505 if res:
3506 res += ', '
3507 res += '%sfps' % fdict['fps']
3508 if fdict.get('acodec') is not None:
3509 if res:
3510 res += ', '
3511 if fdict['acodec'] == 'none':
3512 res += 'video only'
3513 else:
3514 res += '%-5s' % fdict['acodec']
3515 elif fdict.get('abr') is not None:
3516 if res:
3517 res += ', '
3518 res += 'audio'
3519 if fdict.get('abr') is not None:
3520 res += '@%3dk' % fdict['abr']
3521 if fdict.get('asr') is not None:
3522 res += ' (%5dHz)' % fdict['asr']
3523 if fdict.get('filesize') is not None:
3524 if res:
3525 res += ', '
3526 res += format_bytes(fdict['filesize'])
3527 elif fdict.get('filesize_approx') is not None:
3528 if res:
3529 res += ', '
3530 res += '~' + format_bytes(fdict['filesize_approx'])
3531 return res
3532
3533 def render_formats_table(self, info_dict):
3534 if not info_dict.get('formats') and not info_dict.get('url'):
3535 return None
3536
3537 formats = info_dict.get('formats', [info_dict])
3538 if not self.params.get('listformats_table', True) is not False:
3539 table = [
3540 [
3541 format_field(f, 'format_id'),
3542 format_field(f, 'ext'),
3543 self.format_resolution(f),
3544 self._format_note(f)
3545 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3546 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3547
3548 def simplified_codec(f, field):
3549 assert field in ('acodec', 'vcodec')
3550 codec = f.get(field, 'unknown')
3551 if not codec:
3552 return 'unknown'
3553 elif codec != 'none':
3554 return '.'.join(codec.split('.')[:4])
3555
3556 if field == 'vcodec' and f.get('acodec') == 'none':
3557 return 'images'
3558 elif field == 'acodec' and f.get('vcodec') == 'none':
3559 return ''
3560 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3561 self.Styles.SUPPRESS)
3562
3563 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3564 table = [
3565 [
3566 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3567 format_field(f, 'ext'),
3568 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3569 format_field(f, 'fps', '\t%d', func=round),
3570 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3571 delim,
3572 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3573 format_field(f, 'tbr', '\t%dk', func=round),
3574 shorten_protocol_name(f.get('protocol', '')),
3575 delim,
3576 simplified_codec(f, 'vcodec'),
3577 format_field(f, 'vbr', '\t%dk', func=round),
3578 simplified_codec(f, 'acodec'),
3579 format_field(f, 'abr', '\t%dk', func=round),
3580 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3581 join_nonempty(
3582 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3583 format_field(f, 'language', '[%s]'),
3584 join_nonempty(format_field(f, 'format_note'),
3585 format_field(f, 'container', ignore=(None, f.get('ext'))),
3586 delim=', '),
3587 delim=' '),
3588 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3589 header_line = self._list_format_headers(
3590 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3591 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3592
3593 return render_table(
3594 header_line, table, hide_empty=True,
3595 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3596
3597 def render_thumbnails_table(self, info_dict):
3598 thumbnails = list(info_dict.get('thumbnails') or [])
3599 if not thumbnails:
3600 return None
3601 return render_table(
3602 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3603 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3604
3605 def render_subtitles_table(self, video_id, subtitles):
3606 def _row(lang, formats):
3607 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3608 if len(set(names)) == 1:
3609 names = [] if names[0] == 'unknown' else names[:1]
3610 return [lang, ', '.join(names), ', '.join(exts)]
3611
3612 if not subtitles:
3613 return None
3614 return render_table(
3615 self._list_format_headers('Language', 'Name', 'Formats'),
3616 [_row(lang, formats) for lang, formats in subtitles.items()],
3617 hide_empty=True)
3618
3619 def __list_table(self, video_id, name, func, *args):
3620 table = func(*args)
3621 if not table:
3622 self.to_screen(f'{video_id} has no {name}')
3623 return
3624 self.to_screen(f'[info] Available {name} for {video_id}:')
3625 self.to_stdout(table)
3626
3627 def list_formats(self, info_dict):
3628 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3629
3630 def list_thumbnails(self, info_dict):
3631 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3632
3633 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3634 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3635
3636 def urlopen(self, req):
3637 """ Start an HTTP download """
3638 if isinstance(req, str):
3639 req = sanitized_Request(req)
3640 return self._opener.open(req, timeout=self._socket_timeout)
3641
3642 def print_debug_header(self):
3643 if not self.params.get('verbose'):
3644 return
3645
3646 # These imports can be slow. So import them only as needed
3647 from .extractor.extractors import _LAZY_LOADER
3648 from .extractor.extractors import _PLUGIN_CLASSES as plugin_extractors
3649
3650 def get_encoding(stream):
3651 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3652 if not supports_terminal_sequences(stream):
3653 from .utils import WINDOWS_VT_MODE # Must be imported locally
3654 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3655 return ret
3656
3657 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3658 locale.getpreferredencoding(),
3659 sys.getfilesystemencoding(),
3660 self.get_encoding(),
3661 ', '.join(
3662 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3663 if stream is not None and key != 'console')
3664 )
3665
3666 logger = self.params.get('logger')
3667 if logger:
3668 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3669 write_debug(encoding_str)
3670 else:
3671 write_string(f'[debug] {encoding_str}\n', encoding=None)
3672 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3673
3674 source = detect_variant()
3675 write_debug(join_nonempty(
3676 'yt-dlp version', __version__,
3677 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3678 '' if source == 'unknown' else f'({source})',
3679 delim=' '))
3680 if not _LAZY_LOADER:
3681 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3682 write_debug('Lazy loading extractors is forcibly disabled')
3683 else:
3684 write_debug('Lazy loading extractors is disabled')
3685 if plugin_extractors or plugin_postprocessors:
3686 write_debug('Plugins: %s' % [
3687 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3688 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3689 if self.params['compat_opts']:
3690 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3691
3692 if source == 'source':
3693 try:
3694 stdout, _, _ = Popen.run(
3695 ['git', 'rev-parse', '--short', 'HEAD'],
3696 text=True, cwd=os.path.dirname(os.path.abspath(__file__)),
3697 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
3698 if re.fullmatch('[0-9a-f]+', stdout.strip()):
3699 write_debug(f'Git HEAD: {stdout.strip()}')
3700 except Exception:
3701 with contextlib.suppress(Exception):
3702 sys.exc_clear()
3703
3704 write_debug(system_identifier())
3705
3706 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3707 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3708 if ffmpeg_features:
3709 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3710
3711 exe_versions['rtmpdump'] = rtmpdump_version()
3712 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3713 exe_str = ', '.join(
3714 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3715 ) or 'none'
3716 write_debug('exe versions: %s' % exe_str)
3717
3718 from .compat.compat_utils import get_package_info
3719 from .dependencies import available_dependencies
3720
3721 write_debug('Optional libraries: %s' % (', '.join(sorted({
3722 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3723 })) or 'none'))
3724
3725 self._setup_opener()
3726 proxy_map = {}
3727 for handler in self._opener.handlers:
3728 if hasattr(handler, 'proxies'):
3729 proxy_map.update(handler.proxies)
3730 write_debug(f'Proxy map: {proxy_map}')
3731
3732 # Not implemented
3733 if False and self.params.get('call_home'):
3734 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3735 write_debug('Public IP address: %s' % ipaddr)
3736 latest_version = self.urlopen(
3737 'https://yt-dl.org/latest/version').read().decode()
3738 if version_tuple(latest_version) > version_tuple(__version__):
3739 self.report_warning(
3740 'You are using an outdated version (newest version: %s)! '
3741 'See https://yt-dl.org/update if you need help updating.' %
3742 latest_version)
3743
3744 def _setup_opener(self):
3745 if hasattr(self, '_opener'):
3746 return
3747 timeout_val = self.params.get('socket_timeout')
3748 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3749
3750 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3751 opts_cookiefile = self.params.get('cookiefile')
3752 opts_proxy = self.params.get('proxy')
3753
3754 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3755
3756 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3757 if opts_proxy is not None:
3758 if opts_proxy == '':
3759 proxies = {}
3760 else:
3761 proxies = {'http': opts_proxy, 'https': opts_proxy}
3762 else:
3763 proxies = urllib.request.getproxies()
3764 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3765 if 'http' in proxies and 'https' not in proxies:
3766 proxies['https'] = proxies['http']
3767 proxy_handler = PerRequestProxyHandler(proxies)
3768
3769 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3770 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3771 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3772 redirect_handler = YoutubeDLRedirectHandler()
3773 data_handler = urllib.request.DataHandler()
3774
3775 # When passing our own FileHandler instance, build_opener won't add the
3776 # default FileHandler and allows us to disable the file protocol, which
3777 # can be used for malicious purposes (see
3778 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3779 file_handler = urllib.request.FileHandler()
3780
3781 def file_open(*args, **kwargs):
3782 raise urllib.error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3783 file_handler.file_open = file_open
3784
3785 opener = urllib.request.build_opener(
3786 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3787
3788 # Delete the default user-agent header, which would otherwise apply in
3789 # cases where our custom HTTP handler doesn't come into play
3790 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3791 opener.addheaders = []
3792 self._opener = opener
3793
3794 def encode(self, s):
3795 if isinstance(s, bytes):
3796 return s # Already encoded
3797
3798 try:
3799 return s.encode(self.get_encoding())
3800 except UnicodeEncodeError as err:
3801 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3802 raise
3803
3804 def get_encoding(self):
3805 encoding = self.params.get('encoding')
3806 if encoding is None:
3807 encoding = preferredencoding()
3808 return encoding
3809
3810 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3811 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3812 if overwrite is None:
3813 overwrite = self.params.get('overwrites', True)
3814 if not self.params.get('writeinfojson'):
3815 return False
3816 elif not infofn:
3817 self.write_debug(f'Skipping writing {label} infojson')
3818 return False
3819 elif not self._ensure_dir_exists(infofn):
3820 return None
3821 elif not overwrite and os.path.exists(infofn):
3822 self.to_screen(f'[info] {label.title()} metadata is already present')
3823 return 'exists'
3824
3825 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3826 try:
3827 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3828 return True
3829 except OSError:
3830 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3831 return None
3832
3833 def _write_description(self, label, ie_result, descfn):
3834 ''' Write description and returns True = written, False = skip, None = error '''
3835 if not self.params.get('writedescription'):
3836 return False
3837 elif not descfn:
3838 self.write_debug(f'Skipping writing {label} description')
3839 return False
3840 elif not self._ensure_dir_exists(descfn):
3841 return None
3842 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3843 self.to_screen(f'[info] {label.title()} description is already present')
3844 elif ie_result.get('description') is None:
3845 self.report_warning(f'There\'s no {label} description to write')
3846 return False
3847 else:
3848 try:
3849 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3850 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3851 descfile.write(ie_result['description'])
3852 except OSError:
3853 self.report_error(f'Cannot write {label} description file {descfn}')
3854 return None
3855 return True
3856
3857 def _write_subtitles(self, info_dict, filename):
3858 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3859 ret = []
3860 subtitles = info_dict.get('requested_subtitles')
3861 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3862 # subtitles download errors are already managed as troubles in relevant IE
3863 # that way it will silently go on when used with unsupporting IE
3864 return ret
3865
3866 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3867 if not sub_filename_base:
3868 self.to_screen('[info] Skipping writing video subtitles')
3869 return ret
3870 for sub_lang, sub_info in subtitles.items():
3871 sub_format = sub_info['ext']
3872 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3873 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3874 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3875 if existing_sub:
3876 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3877 sub_info['filepath'] = existing_sub
3878 ret.append((existing_sub, sub_filename_final))
3879 continue
3880
3881 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3882 if sub_info.get('data') is not None:
3883 try:
3884 # Use newline='' to prevent conversion of newline characters
3885 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3886 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3887 subfile.write(sub_info['data'])
3888 sub_info['filepath'] = sub_filename
3889 ret.append((sub_filename, sub_filename_final))
3890 continue
3891 except OSError:
3892 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3893 return None
3894
3895 try:
3896 sub_copy = sub_info.copy()
3897 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3898 self.dl(sub_filename, sub_copy, subtitle=True)
3899 sub_info['filepath'] = sub_filename
3900 ret.append((sub_filename, sub_filename_final))
3901 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3902 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3903 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3904 if not self.params.get('ignoreerrors'):
3905 self.report_error(msg)
3906 raise DownloadError(msg)
3907 self.report_warning(msg)
3908 return ret
3909
3910 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3911 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3912 write_all = self.params.get('write_all_thumbnails', False)
3913 thumbnails, ret = [], []
3914 if write_all or self.params.get('writethumbnail', False):
3915 thumbnails = info_dict.get('thumbnails') or []
3916 multiple = write_all and len(thumbnails) > 1
3917
3918 if thumb_filename_base is None:
3919 thumb_filename_base = filename
3920 if thumbnails and not thumb_filename_base:
3921 self.write_debug(f'Skipping writing {label} thumbnail')
3922 return ret
3923
3924 for idx, t in list(enumerate(thumbnails))[::-1]:
3925 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3926 thumb_display_id = f'{label} thumbnail {t["id"]}'
3927 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3928 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3929
3930 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3931 if existing_thumb:
3932 self.to_screen('[info] %s is already present' % (
3933 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3934 t['filepath'] = existing_thumb
3935 ret.append((existing_thumb, thumb_filename_final))
3936 else:
3937 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3938 try:
3939 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3940 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3941 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3942 shutil.copyfileobj(uf, thumbf)
3943 ret.append((thumb_filename, thumb_filename_final))
3944 t['filepath'] = thumb_filename
3945 except network_exceptions as err:
3946 thumbnails.pop(idx)
3947 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3948 if ret and not write_all:
3949 break
3950 return ret