]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
2b7af4cd7e6203f41da3f650a06ba424ac13ac9d
[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 MEDIA_EXTENSIONS,
56 NO_DEFAULT,
57 NUMBER_RE,
58 OUTTMPL_TYPES,
59 POSTPROCESS_WHEN,
60 STR_FORMAT_RE_TMPL,
61 STR_FORMAT_TYPES,
62 ContentTooShortError,
63 DateRange,
64 DownloadCancelled,
65 DownloadError,
66 EntryNotInPlaylist,
67 ExistingVideoReached,
68 ExtractorError,
69 GeoRestrictedError,
70 HEADRequest,
71 ISO3166Utils,
72 LazyList,
73 MaxDownloadsReached,
74 Namespace,
75 PagedList,
76 PerRequestProxyHandler,
77 PlaylistEntries,
78 Popen,
79 PostProcessingError,
80 ReExtractInfo,
81 RejectedVideoReached,
82 SameFileError,
83 UnavailableVideoError,
84 UserNotLive,
85 YoutubeDLCookieProcessor,
86 YoutubeDLHandler,
87 YoutubeDLRedirectHandler,
88 age_restricted,
89 args_to_str,
90 bug_reports_message,
91 date_from_str,
92 determine_ext,
93 determine_protocol,
94 encode_compat_str,
95 encodeFilename,
96 error_to_compat_str,
97 escapeHTML,
98 expand_path,
99 filter_dict,
100 float_or_none,
101 format_bytes,
102 format_decimal_suffix,
103 format_field,
104 formatSeconds,
105 get_compatible_ext,
106 get_domain,
107 int_or_none,
108 iri_to_uri,
109 join_nonempty,
110 locked_file,
111 make_archive_id,
112 make_dir,
113 make_HTTPS_handler,
114 merge_headers,
115 network_exceptions,
116 number_of_digits,
117 orderedSet,
118 parse_filesize,
119 preferredencoding,
120 prepend_extension,
121 register_socks_protocols,
122 remove_terminal_sequences,
123 render_table,
124 replace_extension,
125 sanitize_filename,
126 sanitize_path,
127 sanitize_url,
128 sanitized_Request,
129 std_headers,
130 str_or_none,
131 strftime_or_none,
132 subtitles_filename,
133 supports_terminal_sequences,
134 system_identifier,
135 timetuple_from_msec,
136 to_high_limit_path,
137 traverse_obj,
138 try_call,
139 try_get,
140 url_basename,
141 variadic,
142 version_tuple,
143 windows_enable_vt_mode,
144 write_json_file,
145 write_string,
146 )
147 from .version import RELEASE_GIT_HEAD, VARIANT, __version__
148
149 if compat_os_name == 'nt':
150 import ctypes
151
152
153 class YoutubeDL:
154 """YoutubeDL class.
155
156 YoutubeDL objects are the ones responsible of downloading the
157 actual video file and writing it to disk if the user has requested
158 it, among some other tasks. In most cases there should be one per
159 program. As, given a video URL, the downloader doesn't know how to
160 extract all the needed information, task that InfoExtractors do, it
161 has to pass the URL to one of them.
162
163 For this, YoutubeDL objects have a method that allows
164 InfoExtractors to be registered in a given order. When it is passed
165 a URL, the YoutubeDL object handles it to the first InfoExtractor it
166 finds that reports being able to handle it. The InfoExtractor extracts
167 all the information about the video or videos the URL refers to, and
168 YoutubeDL process the extracted information, possibly using a File
169 Downloader to download the video.
170
171 YoutubeDL objects accept a lot of parameters. In order not to saturate
172 the object constructor with arguments, it receives a dictionary of
173 options instead. These options are available through the params
174 attribute for the InfoExtractors to use. The YoutubeDL also
175 registers itself as the downloader in charge for the InfoExtractors
176 that are added to it, so this is a "mutual registration".
177
178 Available options:
179
180 username: Username for authentication purposes.
181 password: Password for authentication purposes.
182 videopassword: Password for accessing a video.
183 ap_mso: Adobe Pass multiple-system operator identifier.
184 ap_username: Multiple-system operator account username.
185 ap_password: Multiple-system operator account password.
186 usenetrc: Use netrc for authentication instead.
187 verbose: Print additional info to stdout.
188 quiet: Do not print messages to stdout.
189 no_warnings: Do not print out anything for warnings.
190 forceprint: A dict with keys WHEN mapped to a list of templates to
191 print to stdout. The allowed keys are video or any of the
192 items in utils.POSTPROCESS_WHEN.
193 For compatibility, a single list is also accepted
194 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
195 a list of tuples with (template, filename)
196 forcejson: Force printing info_dict as JSON.
197 dump_single_json: Force printing the info_dict of the whole playlist
198 (or video) as a single JSON line.
199 force_write_download_archive: Force writing download archive regardless
200 of 'skip_download' or 'simulate'.
201 simulate: Do not download the video files. If unset (or None),
202 simulate only if listsubtitles, listformats or list_thumbnails is used
203 format: Video format code. see "FORMAT SELECTION" for more details.
204 You can also pass a function. The function takes 'ctx' as
205 argument and returns the formats to download.
206 See "build_format_selector" for an implementation
207 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
208 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
209 extracting metadata even if the video is not actually
210 available for download (experimental)
211 format_sort: A list of fields by which to sort the video formats.
212 See "Sorting Formats" for more details.
213 format_sort_force: Force the given format_sort. see "Sorting Formats"
214 for more details.
215 prefer_free_formats: Whether to prefer video formats with free containers
216 over non-free ones of same quality.
217 allow_multiple_video_streams: Allow multiple video streams to be merged
218 into a single file
219 allow_multiple_audio_streams: Allow multiple audio streams to be merged
220 into a single file
221 check_formats Whether to test if the formats are downloadable.
222 Can be True (check all), False (check none),
223 'selected' (check selected formats),
224 or None (check only if requested by extractor)
225 paths: Dictionary of output paths. The allowed keys are 'home'
226 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
227 outtmpl: Dictionary of templates for output names. Allowed keys
228 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
229 For compatibility with youtube-dl, a single string can also be used
230 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
231 restrictfilenames: Do not allow "&" and spaces in file names
232 trim_file_name: Limit length of filename (extension excluded)
233 windowsfilenames: Force the filenames to be windows compatible
234 ignoreerrors: Do not stop on download/postprocessing errors.
235 Can be 'only_download' to ignore only download errors.
236 Default is 'only_download' for CLI, but False for API
237 skip_playlist_after_errors: Number of allowed failures until the rest of
238 the playlist is skipped
239 force_generic_extractor: Force downloader to use the generic extractor
240 overwrites: Overwrite all video and metadata files if True,
241 overwrite only non-video files if None
242 and don't overwrite any file if False
243 For compatibility with youtube-dl,
244 "nooverwrites" may also be used instead
245 playlist_items: Specific indices of playlist to download.
246 playlistrandom: Download playlist items in random order.
247 lazy_playlist: Process playlist entries as they are received.
248 matchtitle: Download only matching titles.
249 rejecttitle: Reject downloads for matching titles.
250 logger: Log messages to a logging.Logger instance.
251 logtostderr: Log messages to stderr instead of stdout.
252 consoletitle: Display progress in console window's titlebar.
253 writedescription: Write the video description to a .description file
254 writeinfojson: Write the video description to a .info.json file
255 clean_infojson: Remove private fields from the infojson
256 getcomments: Extract video comments. This will not be written to disk
257 unless writeinfojson is also given
258 writeannotations: Write the video annotations to a .annotations.xml file
259 writethumbnail: Write the thumbnail image to a file
260 allow_playlist_files: Whether to write playlists' description, infojson etc
261 also to disk when using the 'write*' options
262 write_all_thumbnails: Write all thumbnail formats to files
263 writelink: Write an internet shortcut file, depending on the
264 current platform (.url/.webloc/.desktop)
265 writeurllink: Write a Windows internet shortcut file (.url)
266 writewebloclink: Write a macOS internet shortcut file (.webloc)
267 writedesktoplink: Write a Linux internet shortcut file (.desktop)
268 writesubtitles: Write the video subtitles to a file
269 writeautomaticsub: Write the automatically generated subtitles to a file
270 listsubtitles: Lists all available subtitles for the video
271 subtitlesformat: The format code for subtitles
272 subtitleslangs: List of languages of the subtitles to download (can be regex).
273 The list may contain "all" to refer to all the available
274 subtitles. The language can be prefixed with a "-" to
275 exclude it from the requested languages. Eg: ['all', '-live_chat']
276 keepvideo: Keep the video file after post-processing
277 daterange: A DateRange object, download only if the upload_date is in the range.
278 skip_download: Skip the actual download of the video file
279 cachedir: Location of the cache files in the filesystem.
280 False to disable filesystem cache.
281 noplaylist: Download single video instead of a playlist if in doubt.
282 age_limit: An integer representing the user's age in years.
283 Unsuitable videos for the given age are skipped.
284 min_views: An integer representing the minimum view count the video
285 must have in order to not be skipped.
286 Videos without view count information are always
287 downloaded. None for no limit.
288 max_views: An integer representing the maximum view count.
289 Videos that are more popular than that are not
290 downloaded.
291 Videos without view count information are always
292 downloaded. None for no limit.
293 download_archive: File name of a file where all downloads are recorded.
294 Videos already present in the file are not downloaded
295 again.
296 break_on_existing: Stop the download process after attempting to download a
297 file that is in the archive.
298 break_on_reject: Stop the download process when encountering a video that
299 has been filtered out.
300 break_per_url: Whether break_on_reject and break_on_existing
301 should act on each input URL as opposed to for the entire queue
302 cookiefile: File name or text stream from where cookies should be read and dumped to
303 cookiesfrombrowser: A tuple containing the name of the browser, the profile
304 name/pathfrom where cookies are loaded, and the name of the
305 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
306 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
307 support RFC 5746 secure renegotiation
308 nocheckcertificate: Do not verify SSL certificates
309 client_certificate: Path to client certificate file in PEM format. May include the private key
310 client_certificate_key: Path to private key file for client certificate
311 client_certificate_password: Password for client certificate private key, if encrypted.
312 If not provided and the key is encrypted, yt-dlp will ask interactively
313 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
314 (Only supported by some extractors)
315 http_headers: A dictionary of custom headers to be used for all requests
316 proxy: URL of the proxy server to use
317 geo_verification_proxy: URL of the proxy to use for IP address verification
318 on geo-restricted sites.
319 socket_timeout: Time to wait for unresponsive hosts, in seconds
320 bidi_workaround: Work around buggy terminals without bidirectional text
321 support, using fridibi
322 debug_printtraffic:Print out sent and received HTTP traffic
323 default_search: Prepend this string if an input url is not valid.
324 'auto' for elaborate guessing
325 encoding: Use this encoding instead of the system-specified.
326 extract_flat: Whether to resolve and process url_results further
327 * False: Always process (default)
328 * True: Never process
329 * 'in_playlist': Do not process inside playlist/multi_video
330 * 'discard': Always process, but don't return the result
331 from inside playlist/multi_video
332 * 'discard_in_playlist': Same as "discard", but only for
333 playlists (not multi_video)
334 wait_for_video: If given, wait for scheduled streams to become available.
335 The value should be a tuple containing the range
336 (min_secs, max_secs) to wait between retries
337 postprocessors: A list of dictionaries, each with an entry
338 * key: The name of the postprocessor. See
339 yt_dlp/postprocessor/__init__.py for a list.
340 * when: When to run the postprocessor. Allowed values are
341 the entries of utils.POSTPROCESS_WHEN
342 Assumed to be 'post_process' if not given
343 progress_hooks: A list of functions that get called on download
344 progress, with a dictionary with the entries
345 * status: One of "downloading", "error", or "finished".
346 Check this first and ignore unknown values.
347 * info_dict: The extracted info_dict
348
349 If status is one of "downloading", or "finished", the
350 following properties may also be present:
351 * filename: The final filename (always present)
352 * tmpfilename: The filename we're currently writing to
353 * downloaded_bytes: Bytes on disk
354 * total_bytes: Size of the whole file, None if unknown
355 * total_bytes_estimate: Guess of the eventual file size,
356 None if unavailable.
357 * elapsed: The number of seconds since download started.
358 * eta: The estimated time in seconds, None if unknown
359 * speed: The download speed in bytes/second, None if
360 unknown
361 * fragment_index: The counter of the currently
362 downloaded video fragment.
363 * fragment_count: The number of fragments (= individual
364 files that will be merged)
365
366 Progress hooks are guaranteed to be called at least once
367 (with status "finished") if the download is successful.
368 postprocessor_hooks: A list of functions that get called on postprocessing
369 progress, with a dictionary with the entries
370 * status: One of "started", "processing", or "finished".
371 Check this first and ignore unknown values.
372 * postprocessor: Name of the postprocessor
373 * info_dict: The extracted info_dict
374
375 Progress hooks are guaranteed to be called at least twice
376 (with status "started" and "finished") if the processing is successful.
377 merge_output_format: "/" separated list of extensions to use when merging formats.
378 final_ext: Expected final extension; used to detect when the file was
379 already downloaded and converted
380 fixup: Automatically correct known faults of the file.
381 One of:
382 - "never": do nothing
383 - "warn": only emit a warning
384 - "detect_or_warn": check whether we can do anything
385 about it, warn otherwise (default)
386 source_address: Client-side IP address to bind to.
387 sleep_interval_requests: Number of seconds to sleep between requests
388 during extraction
389 sleep_interval: Number of seconds to sleep before each download when
390 used alone or a lower bound of a range for randomized
391 sleep before each download (minimum possible number
392 of seconds to sleep) when used along with
393 max_sleep_interval.
394 max_sleep_interval:Upper bound of a range for randomized sleep before each
395 download (maximum possible number of seconds to sleep).
396 Must only be used along with sleep_interval.
397 Actual sleep time will be a random float from range
398 [sleep_interval; max_sleep_interval].
399 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
400 listformats: Print an overview of available video formats and exit.
401 list_thumbnails: Print a table of all thumbnails and exit.
402 match_filter: A function that gets called for every video with the signature
403 (info_dict, *, incomplete: bool) -> Optional[str]
404 For backward compatibility with youtube-dl, the signature
405 (info_dict) -> Optional[str] is also allowed.
406 - If it returns a message, the video is ignored.
407 - If it returns None, the video is downloaded.
408 - If it returns utils.NO_DEFAULT, the user is interactively
409 asked whether to download the video.
410 match_filter_func in utils.py is one example for this.
411 no_color: Do not emit color codes in output.
412 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
413 HTTP header
414 geo_bypass_country:
415 Two-letter ISO 3166-2 country code that will be used for
416 explicit geographic restriction bypassing via faking
417 X-Forwarded-For HTTP header
418 geo_bypass_ip_block:
419 IP range in CIDR notation that will be used similarly to
420 geo_bypass_country
421 external_downloader: A dictionary of protocol keys and the executable of the
422 external downloader to use for it. The allowed protocols
423 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
424 Set the value to 'native' to use the native downloader
425 compat_opts: Compatibility options. See "Differences in default behavior".
426 The following options do not work when used through the API:
427 filename, abort-on-error, multistreams, no-live-chat, format-sort
428 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
429 Refer __init__.py for their implementation
430 progress_template: Dictionary of templates for progress outputs.
431 Allowed keys are 'download', 'postprocess',
432 'download-title' (console title) and 'postprocess-title'.
433 The template is mapped on a dictionary with keys 'progress' and 'info'
434 retry_sleep_functions: Dictionary of functions that takes the number of attempts
435 as argument and returns the time to sleep in seconds.
436 Allowed keys are 'http', 'fragment', 'file_access'
437 download_ranges: A callback function that gets called for every video with
438 the signature (info_dict, ydl) -> Iterable[Section].
439 Only the returned sections will be downloaded.
440 Each Section is a dict with the following keys:
441 * start_time: Start time of the section in seconds
442 * end_time: End time of the section in seconds
443 * title: Section title (Optional)
444 * index: Section number (Optional)
445 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
446 noprogress: Do not print the progress bar
447
448 The following parameters are not used by YoutubeDL itself, they are used by
449 the downloader (see yt_dlp/downloader/common.py):
450 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
451 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
452 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
453 external_downloader_args, concurrent_fragment_downloads.
454
455 The following options are used by the post processors:
456 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
457 to the binary or its containing directory.
458 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
459 and a list of additional command-line arguments for the
460 postprocessor/executable. The dict can also have "PP+EXE" keys
461 which are used when the given exe is used by the given PP.
462 Use 'default' as the name for arguments to passed to all PP
463 For compatibility with youtube-dl, a single list of args
464 can also be used
465
466 The following options are used by the extractors:
467 extractor_retries: Number of times to retry for known errors
468 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
469 hls_split_discontinuity: Split HLS playlists to different formats at
470 discontinuities such as ad breaks (default: False)
471 extractor_args: A dictionary of arguments to be passed to the extractors.
472 See "EXTRACTOR ARGUMENTS" for details.
473 Eg: {'youtube': {'skip': ['dash', 'hls']}}
474 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
475
476 The following options are deprecated and may be removed in the future:
477
478 playliststart: - Use playlist_items
479 Playlist item to start at.
480 playlistend: - Use playlist_items
481 Playlist item to end at.
482 playlistreverse: - Use playlist_items
483 Download playlist items in reverse order.
484 forceurl: - Use forceprint
485 Force printing final URL.
486 forcetitle: - Use forceprint
487 Force printing title.
488 forceid: - Use forceprint
489 Force printing ID.
490 forcethumbnail: - Use forceprint
491 Force printing thumbnail URL.
492 forcedescription: - Use forceprint
493 Force printing description.
494 forcefilename: - Use forceprint
495 Force printing final filename.
496 forceduration: - Use forceprint
497 Force printing duration.
498 allsubtitles: - Use subtitleslangs = ['all']
499 Downloads all the subtitles of the video
500 (requires writesubtitles or writeautomaticsub)
501 include_ads: - Doesn't work
502 Download ads as well
503 call_home: - Not implemented
504 Boolean, true iff we are allowed to contact the
505 yt-dlp servers for debugging.
506 post_hooks: - Register a custom postprocessor
507 A list of functions that get called as the final step
508 for each video file, after all postprocessors have been
509 called. The filename will be passed as the only argument.
510 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
511 Use the native HLS downloader instead of ffmpeg/avconv
512 if True, otherwise use ffmpeg/avconv if False, otherwise
513 use downloader suggested by extractor if None.
514 prefer_ffmpeg: - avconv support is deprecated
515 If False, use avconv instead of ffmpeg if both are available,
516 otherwise prefer ffmpeg.
517 youtube_include_dash_manifest: - Use extractor_args
518 If True (default), DASH 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 DASH. (only for youtube)
522 youtube_include_hls_manifest: - Use extractor_args
523 If True (default), HLS manifests and related
524 data will be downloaded and processed by extractor.
525 You can reduce network I/O by disabling it if you don't
526 care about HLS. (only for youtube)
527 """
528
529 _NUMERIC_FIELDS = {
530 'width', 'height', 'asr', 'audio_channels', 'fps',
531 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
532 'timestamp', 'release_timestamp',
533 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
534 'average_rating', 'comment_count', 'age_limit',
535 'start_time', 'end_time',
536 'chapter_number', 'season_number', 'episode_number',
537 'track_number', 'disc_number', 'release_year',
538 }
539
540 _format_fields = {
541 # NB: Keep in sync with the docstring of extractor/common.py
542 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
543 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
544 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
545 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
546 'preference', 'language', 'language_preference', 'quality', 'source_preference',
547 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
548 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
549 }
550 _format_selection_exts = {
551 'audio': set(MEDIA_EXTENSIONS.common_audio),
552 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
553 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
554 }
555
556 def __init__(self, params=None, auto_init=True):
557 """Create a FileDownloader object with the given options.
558 @param auto_init Whether to load the default extractors and print header (if verbose).
559 Set to 'no_verbose_header' to not print the header
560 """
561 if params is None:
562 params = {}
563 self.params = params
564 self._ies = {}
565 self._ies_instances = {}
566 self._pps = {k: [] for k in POSTPROCESS_WHEN}
567 self._printed_messages = set()
568 self._first_webpage_request = True
569 self._post_hooks = []
570 self._progress_hooks = []
571 self._postprocessor_hooks = []
572 self._download_retcode = 0
573 self._num_downloads = 0
574 self._num_videos = 0
575 self._playlist_level = 0
576 self._playlist_urls = set()
577 self.cache = Cache(self)
578
579 windows_enable_vt_mode()
580 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
581 self._out_files = Namespace(
582 out=stdout,
583 error=sys.stderr,
584 screen=sys.stderr if self.params.get('quiet') else stdout,
585 console=None if compat_os_name == 'nt' else next(
586 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
587 )
588 self._allow_colors = Namespace(**{
589 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
590 for type_, stream in self._out_files.items_ if type_ != 'console'
591 })
592
593 # The code is left like this to be reused for future deprecations
594 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 7), (3, 7)
595 current_version = sys.version_info[:2]
596 if current_version < MIN_RECOMMENDED:
597 msg = ('Support for Python version %d.%d has been deprecated. '
598 'See https://github.com/yt-dlp/yt-dlp/issues/3764 for more details.'
599 '\n You will no longer receive updates on this version')
600 if current_version < MIN_SUPPORTED:
601 msg = 'Python version %d.%d is no longer supported'
602 self.deprecation_warning(
603 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
604
605 if self.params.get('allow_unplayable_formats'):
606 self.report_warning(
607 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
608 'This is a developer option intended for debugging. \n'
609 ' If you experience any issues while using this option, '
610 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
611
612 def check_deprecated(param, option, suggestion):
613 if self.params.get(param) is not None:
614 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
615 return True
616 return False
617
618 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
619 if self.params.get('geo_verification_proxy') is None:
620 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
621
622 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
623 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
624 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
625
626 for msg in self.params.get('_warnings', []):
627 self.report_warning(msg)
628 for msg in self.params.get('_deprecation_warnings', []):
629 self.deprecation_warning(msg)
630
631 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
632 if 'list-formats' in self.params['compat_opts']:
633 self.params['listformats_table'] = False
634
635 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
636 # nooverwrites was unnecessarily changed to overwrites
637 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
638 # This ensures compatibility with both keys
639 self.params['overwrites'] = not self.params['nooverwrites']
640 elif self.params.get('overwrites') is None:
641 self.params.pop('overwrites', None)
642 else:
643 self.params['nooverwrites'] = not self.params['overwrites']
644
645 self.params.setdefault('forceprint', {})
646 self.params.setdefault('print_to_file', {})
647
648 # Compatibility with older syntax
649 if not isinstance(params['forceprint'], dict):
650 self.params['forceprint'] = {'video': params['forceprint']}
651
652 if self.params.get('bidi_workaround', False):
653 try:
654 import pty
655 master, slave = pty.openpty()
656 width = shutil.get_terminal_size().columns
657 width_args = [] if width is None else ['-w', str(width)]
658 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
659 try:
660 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
661 except OSError:
662 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
663 self._output_channel = os.fdopen(master, 'rb')
664 except OSError as ose:
665 if ose.errno == errno.ENOENT:
666 self.report_warning(
667 'Could not find fribidi executable, ignoring --bidi-workaround. '
668 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
669 else:
670 raise
671
672 if auto_init:
673 if auto_init != 'no_verbose_header':
674 self.print_debug_header()
675 self.add_default_info_extractors()
676
677 if (sys.platform != 'win32'
678 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
679 and not self.params.get('restrictfilenames', False)):
680 # Unicode filesystem API will throw errors (#1474, #13027)
681 self.report_warning(
682 'Assuming --restrict-filenames since file system encoding '
683 'cannot encode all characters. '
684 'Set the LC_ALL environment variable to fix this.')
685 self.params['restrictfilenames'] = True
686
687 self._parse_outtmpl()
688
689 # Creating format selector here allows us to catch syntax errors before the extraction
690 self.format_selector = (
691 self.params.get('format') if self.params.get('format') in (None, '-')
692 else self.params['format'] if callable(self.params['format'])
693 else self.build_format_selector(self.params['format']))
694
695 # Set http_headers defaults according to std_headers
696 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
697
698 hooks = {
699 'post_hooks': self.add_post_hook,
700 'progress_hooks': self.add_progress_hook,
701 'postprocessor_hooks': self.add_postprocessor_hook,
702 }
703 for opt, fn in hooks.items():
704 for ph in self.params.get(opt, []):
705 fn(ph)
706
707 for pp_def_raw in self.params.get('postprocessors', []):
708 pp_def = dict(pp_def_raw)
709 when = pp_def.pop('when', 'post_process')
710 self.add_post_processor(
711 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
712 when=when)
713
714 self._setup_opener()
715 register_socks_protocols()
716
717 def preload_download_archive(fn):
718 """Preload the archive, if any is specified"""
719 if fn is None:
720 return False
721 self.write_debug(f'Loading archive file {fn!r}')
722 try:
723 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
724 for line in archive_file:
725 self.archive.add(line.strip())
726 except OSError as ioe:
727 if ioe.errno != errno.ENOENT:
728 raise
729 return False
730 return True
731
732 self.archive = set()
733 preload_download_archive(self.params.get('download_archive'))
734
735 def warn_if_short_id(self, argv):
736 # short YouTube ID starting with dash?
737 idxs = [
738 i for i, a in enumerate(argv)
739 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
740 if idxs:
741 correct_argv = (
742 ['yt-dlp']
743 + [a for i, a in enumerate(argv) if i not in idxs]
744 + ['--'] + [argv[i] for i in idxs]
745 )
746 self.report_warning(
747 'Long argument string detected. '
748 'Use -- to separate parameters and URLs, like this:\n%s' %
749 args_to_str(correct_argv))
750
751 def add_info_extractor(self, ie):
752 """Add an InfoExtractor object to the end of the list."""
753 ie_key = ie.ie_key()
754 self._ies[ie_key] = ie
755 if not isinstance(ie, type):
756 self._ies_instances[ie_key] = ie
757 ie.set_downloader(self)
758
759 def _get_info_extractor_class(self, ie_key):
760 ie = self._ies.get(ie_key)
761 if ie is None:
762 ie = get_info_extractor(ie_key)
763 self.add_info_extractor(ie)
764 return ie
765
766 def get_info_extractor(self, ie_key):
767 """
768 Get an instance of an IE with name ie_key, it will try to get one from
769 the _ies list, if there's no instance it will create a new one and add
770 it to the extractor list.
771 """
772 ie = self._ies_instances.get(ie_key)
773 if ie is None:
774 ie = get_info_extractor(ie_key)()
775 self.add_info_extractor(ie)
776 return ie
777
778 def add_default_info_extractors(self):
779 """
780 Add the InfoExtractors returned by gen_extractors to the end of the list
781 """
782 for ie in gen_extractor_classes():
783 self.add_info_extractor(ie)
784
785 def add_post_processor(self, pp, when='post_process'):
786 """Add a PostProcessor object to the end of the chain."""
787 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
788 self._pps[when].append(pp)
789 pp.set_downloader(self)
790
791 def add_post_hook(self, ph):
792 """Add the post hook"""
793 self._post_hooks.append(ph)
794
795 def add_progress_hook(self, ph):
796 """Add the download progress hook"""
797 self._progress_hooks.append(ph)
798
799 def add_postprocessor_hook(self, ph):
800 """Add the postprocessing progress hook"""
801 self._postprocessor_hooks.append(ph)
802 for pps in self._pps.values():
803 for pp in pps:
804 pp.add_progress_hook(ph)
805
806 def _bidi_workaround(self, message):
807 if not hasattr(self, '_output_channel'):
808 return message
809
810 assert hasattr(self, '_output_process')
811 assert isinstance(message, str)
812 line_count = message.count('\n') + 1
813 self._output_process.stdin.write((message + '\n').encode())
814 self._output_process.stdin.flush()
815 res = ''.join(self._output_channel.readline().decode()
816 for _ in range(line_count))
817 return res[:-len('\n')]
818
819 def _write_string(self, message, out=None, only_once=False):
820 if only_once:
821 if message in self._printed_messages:
822 return
823 self._printed_messages.add(message)
824 write_string(message, out=out, encoding=self.params.get('encoding'))
825
826 def to_stdout(self, message, skip_eol=False, quiet=None):
827 """Print message to stdout"""
828 if quiet is not None:
829 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
830 if skip_eol is not False:
831 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. Use "YoutubeDL.to_screen" instead')
832 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
833
834 def to_screen(self, message, skip_eol=False, quiet=None):
835 """Print message to screen if not in quiet mode"""
836 if self.params.get('logger'):
837 self.params['logger'].debug(message)
838 return
839 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
840 return
841 self._write_string(
842 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
843 self._out_files.screen)
844
845 def to_stderr(self, message, only_once=False):
846 """Print message to stderr"""
847 assert isinstance(message, str)
848 if self.params.get('logger'):
849 self.params['logger'].error(message)
850 else:
851 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
852
853 def _send_console_code(self, code):
854 if compat_os_name == 'nt' or not self._out_files.console:
855 return
856 self._write_string(code, self._out_files.console)
857
858 def to_console_title(self, message):
859 if not self.params.get('consoletitle', False):
860 return
861 message = remove_terminal_sequences(message)
862 if compat_os_name == 'nt':
863 if ctypes.windll.kernel32.GetConsoleWindow():
864 # c_wchar_p() might not be necessary if `message` is
865 # already of type unicode()
866 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
867 else:
868 self._send_console_code(f'\033]0;{message}\007')
869
870 def save_console_title(self):
871 if not self.params.get('consoletitle') or self.params.get('simulate'):
872 return
873 self._send_console_code('\033[22;0t') # Save the title on stack
874
875 def restore_console_title(self):
876 if not self.params.get('consoletitle') or self.params.get('simulate'):
877 return
878 self._send_console_code('\033[23;0t') # Restore the title from stack
879
880 def __enter__(self):
881 self.save_console_title()
882 return self
883
884 def __exit__(self, *args):
885 self.restore_console_title()
886
887 if self.params.get('cookiefile') is not None:
888 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
889
890 def trouble(self, message=None, tb=None, is_error=True):
891 """Determine action to take when a download problem appears.
892
893 Depending on if the downloader has been configured to ignore
894 download errors or not, this method may throw an exception or
895 not when errors are found, after printing the message.
896
897 @param tb If given, is additional traceback information
898 @param is_error Whether to raise error according to ignorerrors
899 """
900 if message is not None:
901 self.to_stderr(message)
902 if self.params.get('verbose'):
903 if tb is None:
904 if sys.exc_info()[0]: # if .trouble has been called from an except block
905 tb = ''
906 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
907 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
908 tb += encode_compat_str(traceback.format_exc())
909 else:
910 tb_data = traceback.format_list(traceback.extract_stack())
911 tb = ''.join(tb_data)
912 if tb:
913 self.to_stderr(tb)
914 if not is_error:
915 return
916 if not self.params.get('ignoreerrors'):
917 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
918 exc_info = sys.exc_info()[1].exc_info
919 else:
920 exc_info = sys.exc_info()
921 raise DownloadError(message, exc_info)
922 self._download_retcode = 1
923
924 Styles = Namespace(
925 HEADERS='yellow',
926 EMPHASIS='light blue',
927 FILENAME='green',
928 ID='green',
929 DELIM='blue',
930 ERROR='red',
931 WARNING='yellow',
932 SUPPRESS='light black',
933 )
934
935 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
936 text = str(text)
937 if test_encoding:
938 original_text = text
939 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
940 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
941 text = text.encode(encoding, 'ignore').decode(encoding)
942 if fallback is not None and text != original_text:
943 text = fallback
944 return format_text(text, f) if allow_colors else text if fallback is None else fallback
945
946 def _format_out(self, *args, **kwargs):
947 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
948
949 def _format_screen(self, *args, **kwargs):
950 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
951
952 def _format_err(self, *args, **kwargs):
953 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
954
955 def report_warning(self, message, only_once=False):
956 '''
957 Print the message to stderr, it will be prefixed with 'WARNING:'
958 If stderr is a tty file the 'WARNING:' will be colored
959 '''
960 if self.params.get('logger') is not None:
961 self.params['logger'].warning(message)
962 else:
963 if self.params.get('no_warnings'):
964 return
965 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
966
967 def deprecation_warning(self, message):
968 if self.params.get('logger') is not None:
969 self.params['logger'].warning(f'DeprecationWarning: {message}')
970 else:
971 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
972
973 def report_error(self, message, *args, **kwargs):
974 '''
975 Do the same as trouble, but prefixes the message with 'ERROR:', colored
976 in red if stderr is a tty file.
977 '''
978 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
979
980 def write_debug(self, message, only_once=False):
981 '''Log debug message or Print message to stderr'''
982 if not self.params.get('verbose', False):
983 return
984 message = f'[debug] {message}'
985 if self.params.get('logger'):
986 self.params['logger'].debug(message)
987 else:
988 self.to_stderr(message, only_once)
989
990 def report_file_already_downloaded(self, file_name):
991 """Report file has already been fully downloaded."""
992 try:
993 self.to_screen('[download] %s has already been downloaded' % file_name)
994 except UnicodeEncodeError:
995 self.to_screen('[download] The file has already been downloaded')
996
997 def report_file_delete(self, file_name):
998 """Report that existing file will be deleted."""
999 try:
1000 self.to_screen('Deleting existing file %s' % file_name)
1001 except UnicodeEncodeError:
1002 self.to_screen('Deleting existing file')
1003
1004 def raise_no_formats(self, info, forced=False, *, msg=None):
1005 has_drm = info.get('_has_drm')
1006 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1007 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1008 if forced or not ignored:
1009 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1010 expected=has_drm or ignored or expected)
1011 else:
1012 self.report_warning(msg)
1013
1014 def parse_outtmpl(self):
1015 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1016 self._parse_outtmpl()
1017 return self.params['outtmpl']
1018
1019 def _parse_outtmpl(self):
1020 sanitize = IDENTITY
1021 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1022 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1023
1024 outtmpl = self.params.setdefault('outtmpl', {})
1025 if not isinstance(outtmpl, dict):
1026 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1027 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1028
1029 def get_output_path(self, dir_type='', filename=None):
1030 paths = self.params.get('paths', {})
1031 assert isinstance(paths, dict)
1032 path = os.path.join(
1033 expand_path(paths.get('home', '').strip()),
1034 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1035 filename or '')
1036 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1037
1038 @staticmethod
1039 def _outtmpl_expandpath(outtmpl):
1040 # expand_path translates '%%' into '%' and '$$' into '$'
1041 # correspondingly that is not what we want since we need to keep
1042 # '%%' intact for template dict substitution step. Working around
1043 # with boundary-alike separator hack.
1044 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1045 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1046
1047 # outtmpl should be expand_path'ed before template dict substitution
1048 # because meta fields may contain env variables we don't want to
1049 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1050 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1051 return expand_path(outtmpl).replace(sep, '')
1052
1053 @staticmethod
1054 def escape_outtmpl(outtmpl):
1055 ''' Escape any remaining strings like %s, %abc% etc. '''
1056 return re.sub(
1057 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1058 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1059 outtmpl)
1060
1061 @classmethod
1062 def validate_outtmpl(cls, outtmpl):
1063 ''' @return None or Exception object '''
1064 outtmpl = re.sub(
1065 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1066 lambda mobj: f'{mobj.group(0)[:-1]}s',
1067 cls._outtmpl_expandpath(outtmpl))
1068 try:
1069 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1070 return None
1071 except ValueError as err:
1072 return err
1073
1074 @staticmethod
1075 def _copy_infodict(info_dict):
1076 info_dict = dict(info_dict)
1077 info_dict.pop('__postprocessors', None)
1078 info_dict.pop('__pending_error', None)
1079 return info_dict
1080
1081 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1082 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1083 @param sanitize Whether to sanitize the output as a filename.
1084 For backward compatibility, a function can also be passed
1085 """
1086
1087 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1088
1089 info_dict = self._copy_infodict(info_dict)
1090 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1091 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1092 if info_dict.get('duration', None) is not None
1093 else None)
1094 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1095 info_dict['video_autonumber'] = self._num_videos
1096 if info_dict.get('resolution') is None:
1097 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1098
1099 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1100 # of %(field)s to %(field)0Nd for backward compatibility
1101 field_size_compat_map = {
1102 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1103 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1104 'autonumber': self.params.get('autonumber_size') or 5,
1105 }
1106
1107 TMPL_DICT = {}
1108 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1109 MATH_FUNCTIONS = {
1110 '+': float.__add__,
1111 '-': float.__sub__,
1112 }
1113 # Field is of the form key1.key2...
1114 # where keys (except first) can be string, int or slice
1115 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1116 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1117 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1118 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1119 (?P<negate>-)?
1120 (?P<fields>{FIELD_RE})
1121 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1122 (?:>(?P<strf_format>.+?))?
1123 (?P<remaining>
1124 (?P<alternate>(?<!\\),[^|&)]+)?
1125 (?:&(?P<replacement>.*?))?
1126 (?:\|(?P<default>.*?))?
1127 )$''')
1128
1129 def _traverse_infodict(k):
1130 k = k.split('.')
1131 if k[0] == '':
1132 k.pop(0)
1133 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1134
1135 def get_value(mdict):
1136 # Object traversal
1137 value = _traverse_infodict(mdict['fields'])
1138 # Negative
1139 if mdict['negate']:
1140 value = float_or_none(value)
1141 if value is not None:
1142 value *= -1
1143 # Do maths
1144 offset_key = mdict['maths']
1145 if offset_key:
1146 value = float_or_none(value)
1147 operator = None
1148 while offset_key:
1149 item = re.match(
1150 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1151 offset_key).group(0)
1152 offset_key = offset_key[len(item):]
1153 if operator is None:
1154 operator = MATH_FUNCTIONS[item]
1155 continue
1156 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1157 offset = float_or_none(item)
1158 if offset is None:
1159 offset = float_or_none(_traverse_infodict(item))
1160 try:
1161 value = operator(value, multiplier * offset)
1162 except (TypeError, ZeroDivisionError):
1163 return None
1164 operator = None
1165 # Datetime formatting
1166 if mdict['strf_format']:
1167 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1168
1169 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1170 if sanitize and value == '':
1171 value = None
1172 return value
1173
1174 na = self.params.get('outtmpl_na_placeholder', 'NA')
1175
1176 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1177 return sanitize_filename(str(value), restricted=restricted, is_id=(
1178 bool(re.search(r'(^|[_.])id(\.|$)', key))
1179 if 'filename-sanitization' in self.params['compat_opts']
1180 else NO_DEFAULT))
1181
1182 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1183 sanitize = bool(sanitize)
1184
1185 def _dumpjson_default(obj):
1186 if isinstance(obj, (set, LazyList)):
1187 return list(obj)
1188 return repr(obj)
1189
1190 def create_key(outer_mobj):
1191 if not outer_mobj.group('has_key'):
1192 return outer_mobj.group(0)
1193 key = outer_mobj.group('key')
1194 mobj = re.match(INTERNAL_FORMAT_RE, key)
1195 initial_field = mobj.group('fields') if mobj else ''
1196 value, replacement, default = None, None, na
1197 while mobj:
1198 mobj = mobj.groupdict()
1199 default = mobj['default'] if mobj['default'] is not None else default
1200 value = get_value(mobj)
1201 replacement = mobj['replacement']
1202 if value is None and mobj['alternate']:
1203 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1204 else:
1205 break
1206
1207 fmt = outer_mobj.group('format')
1208 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1209 fmt = f'0{field_size_compat_map[key]:d}d'
1210
1211 value = default if value is None else value if replacement is None else replacement
1212
1213 flags = outer_mobj.group('conversion') or ''
1214 str_fmt = f'{fmt[:-1]}s'
1215 if fmt[-1] == 'l': # list
1216 delim = '\n' if '#' in flags else ', '
1217 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1218 elif fmt[-1] == 'j': # json
1219 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1220 elif fmt[-1] == 'h': # html
1221 value, fmt = escapeHTML(value), str_fmt
1222 elif fmt[-1] == 'q': # quoted
1223 value = map(str, variadic(value) if '#' in flags else [value])
1224 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1225 elif fmt[-1] == 'B': # bytes
1226 value = f'%{str_fmt}'.encode() % str(value).encode()
1227 value, fmt = value.decode('utf-8', 'ignore'), 's'
1228 elif fmt[-1] == 'U': # unicode normalized
1229 value, fmt = unicodedata.normalize(
1230 # "+" = compatibility equivalence, "#" = NFD
1231 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1232 value), str_fmt
1233 elif fmt[-1] == 'D': # decimal suffix
1234 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1235 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1236 factor=1024 if '#' in flags else 1000)
1237 elif fmt[-1] == 'S': # filename sanitization
1238 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1239 elif fmt[-1] == 'c':
1240 if value:
1241 value = str(value)[0]
1242 else:
1243 fmt = str_fmt
1244 elif fmt[-1] not in 'rs': # numeric
1245 value = float_or_none(value)
1246 if value is None:
1247 value, fmt = default, 's'
1248
1249 if sanitize:
1250 if fmt[-1] == 'r':
1251 # If value is an object, sanitize might convert it to a string
1252 # So we convert it to repr first
1253 value, fmt = repr(value), str_fmt
1254 if fmt[-1] in 'csr':
1255 value = sanitizer(initial_field, value)
1256
1257 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1258 TMPL_DICT[key] = value
1259 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1260
1261 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1262
1263 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1264 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1265 return self.escape_outtmpl(outtmpl) % info_dict
1266
1267 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1268 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1269 if outtmpl is None:
1270 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1271 try:
1272 outtmpl = self._outtmpl_expandpath(outtmpl)
1273 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1274 if not filename:
1275 return None
1276
1277 if tmpl_type in ('', 'temp'):
1278 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1279 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1280 filename = replace_extension(filename, ext, final_ext)
1281 elif tmpl_type:
1282 force_ext = OUTTMPL_TYPES[tmpl_type]
1283 if force_ext:
1284 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1285
1286 # https://github.com/blackjack4494/youtube-dlc/issues/85
1287 trim_file_name = self.params.get('trim_file_name', False)
1288 if trim_file_name:
1289 no_ext, *ext = filename.rsplit('.', 2)
1290 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1291
1292 return filename
1293 except ValueError as err:
1294 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1295 return None
1296
1297 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1298 """Generate the output filename"""
1299 if outtmpl:
1300 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1301 dir_type = None
1302 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1303 if not filename and dir_type not in ('', 'temp'):
1304 return ''
1305
1306 if warn:
1307 if not self.params.get('paths'):
1308 pass
1309 elif filename == '-':
1310 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1311 elif os.path.isabs(filename):
1312 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1313 if filename == '-' or not filename:
1314 return filename
1315
1316 return self.get_output_path(dir_type, filename)
1317
1318 def _match_entry(self, info_dict, incomplete=False, silent=False):
1319 """ Returns None if the file should be downloaded """
1320
1321 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1322
1323 def check_filter():
1324 if 'title' in info_dict:
1325 # This can happen when we're just evaluating the playlist
1326 title = info_dict['title']
1327 matchtitle = self.params.get('matchtitle', False)
1328 if matchtitle:
1329 if not re.search(matchtitle, title, re.IGNORECASE):
1330 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1331 rejecttitle = self.params.get('rejecttitle', False)
1332 if rejecttitle:
1333 if re.search(rejecttitle, title, re.IGNORECASE):
1334 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1335 date = info_dict.get('upload_date')
1336 if date is not None:
1337 dateRange = self.params.get('daterange', DateRange())
1338 if date not in dateRange:
1339 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1340 view_count = info_dict.get('view_count')
1341 if view_count is not None:
1342 min_views = self.params.get('min_views')
1343 if min_views is not None and view_count < min_views:
1344 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1345 max_views = self.params.get('max_views')
1346 if max_views is not None and view_count > max_views:
1347 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1348 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1349 return 'Skipping "%s" because it is age restricted' % video_title
1350
1351 match_filter = self.params.get('match_filter')
1352 if match_filter is not None:
1353 try:
1354 ret = match_filter(info_dict, incomplete=incomplete)
1355 except TypeError:
1356 # For backward compatibility
1357 ret = None if incomplete else match_filter(info_dict)
1358 if ret is NO_DEFAULT:
1359 while True:
1360 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1361 reply = input(self._format_screen(
1362 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1363 if reply in {'y', ''}:
1364 return None
1365 elif reply == 'n':
1366 return f'Skipping {video_title}'
1367 elif ret is not None:
1368 return ret
1369 return None
1370
1371 if self.in_download_archive(info_dict):
1372 reason = '%s has already been recorded in the archive' % video_title
1373 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1374 else:
1375 reason = check_filter()
1376 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1377 if reason is not None:
1378 if not silent:
1379 self.to_screen('[download] ' + reason)
1380 if self.params.get(break_opt, False):
1381 raise break_err()
1382 return reason
1383
1384 @staticmethod
1385 def add_extra_info(info_dict, extra_info):
1386 '''Set the keys from extra_info in info dict if they are missing'''
1387 for key, value in extra_info.items():
1388 info_dict.setdefault(key, value)
1389
1390 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1391 process=True, force_generic_extractor=False):
1392 """
1393 Return a list with a dictionary for each video extracted.
1394
1395 Arguments:
1396 url -- URL to extract
1397
1398 Keyword arguments:
1399 download -- whether to download videos during extraction
1400 ie_key -- extractor key hint
1401 extra_info -- dictionary containing the extra values to add to each result
1402 process -- whether to resolve all unresolved references (URLs, playlist items),
1403 must be True for download to work.
1404 force_generic_extractor -- force using the generic extractor
1405 """
1406
1407 if extra_info is None:
1408 extra_info = {}
1409
1410 if not ie_key and force_generic_extractor:
1411 ie_key = 'Generic'
1412
1413 if ie_key:
1414 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1415 else:
1416 ies = self._ies
1417
1418 for ie_key, ie in ies.items():
1419 if not ie.suitable(url):
1420 continue
1421
1422 if not ie.working():
1423 self.report_warning('The program functionality for this site has been marked as broken, '
1424 'and will probably not work.')
1425
1426 temp_id = ie.get_temp_id(url)
1427 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1428 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1429 if self.params.get('break_on_existing', False):
1430 raise ExistingVideoReached()
1431 break
1432 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1433 else:
1434 self.report_error('no suitable InfoExtractor for URL %s' % url)
1435
1436 def _handle_extraction_exceptions(func):
1437 @functools.wraps(func)
1438 def wrapper(self, *args, **kwargs):
1439 while True:
1440 try:
1441 return func(self, *args, **kwargs)
1442 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1443 raise
1444 except ReExtractInfo as e:
1445 if e.expected:
1446 self.to_screen(f'{e}; Re-extracting data')
1447 else:
1448 self.to_stderr('\r')
1449 self.report_warning(f'{e}; Re-extracting data')
1450 continue
1451 except GeoRestrictedError as e:
1452 msg = e.msg
1453 if e.countries:
1454 msg += '\nThis video is available in %s.' % ', '.join(
1455 map(ISO3166Utils.short2full, e.countries))
1456 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1457 self.report_error(msg)
1458 except ExtractorError as e: # An error we somewhat expected
1459 self.report_error(str(e), e.format_traceback())
1460 except Exception as e:
1461 if self.params.get('ignoreerrors'):
1462 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1463 else:
1464 raise
1465 break
1466 return wrapper
1467
1468 def _wait_for_video(self, ie_result={}):
1469 if (not self.params.get('wait_for_video')
1470 or ie_result.get('_type', 'video') != 'video'
1471 or ie_result.get('formats') or ie_result.get('url')):
1472 return
1473
1474 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1475 last_msg = ''
1476
1477 def progress(msg):
1478 nonlocal last_msg
1479 full_msg = f'{msg}\n'
1480 if not self.params.get('noprogress'):
1481 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1482 elif last_msg:
1483 return
1484 self.to_screen(full_msg, skip_eol=True)
1485 last_msg = msg
1486
1487 min_wait, max_wait = self.params.get('wait_for_video')
1488 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1489 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1490 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1491 self.report_warning('Release time of video is not known')
1492 elif ie_result and (diff or 0) <= 0:
1493 self.report_warning('Video should already be available according to extracted info')
1494 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1495 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1496
1497 wait_till = time.time() + diff
1498 try:
1499 while True:
1500 diff = wait_till - time.time()
1501 if diff <= 0:
1502 progress('')
1503 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1504 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1505 time.sleep(1)
1506 except KeyboardInterrupt:
1507 progress('')
1508 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1509 except BaseException as e:
1510 if not isinstance(e, ReExtractInfo):
1511 self.to_screen('')
1512 raise
1513
1514 @_handle_extraction_exceptions
1515 def __extract_info(self, url, ie, download, extra_info, process):
1516 try:
1517 ie_result = ie.extract(url)
1518 except UserNotLive as e:
1519 if process:
1520 if self.params.get('wait_for_video'):
1521 self.report_warning(e)
1522 self._wait_for_video()
1523 raise
1524 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1525 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1526 return
1527 if isinstance(ie_result, list):
1528 # Backwards compatibility: old IE result format
1529 ie_result = {
1530 '_type': 'compat_list',
1531 'entries': ie_result,
1532 }
1533 if extra_info.get('original_url'):
1534 ie_result.setdefault('original_url', extra_info['original_url'])
1535 self.add_default_extra_info(ie_result, ie, url)
1536 if process:
1537 self._wait_for_video(ie_result)
1538 return self.process_ie_result(ie_result, download, extra_info)
1539 else:
1540 return ie_result
1541
1542 def add_default_extra_info(self, ie_result, ie, url):
1543 if url is not None:
1544 self.add_extra_info(ie_result, {
1545 'webpage_url': url,
1546 'original_url': url,
1547 })
1548 webpage_url = ie_result.get('webpage_url')
1549 if webpage_url:
1550 self.add_extra_info(ie_result, {
1551 'webpage_url_basename': url_basename(webpage_url),
1552 'webpage_url_domain': get_domain(webpage_url),
1553 })
1554 if ie is not None:
1555 self.add_extra_info(ie_result, {
1556 'extractor': ie.IE_NAME,
1557 'extractor_key': ie.ie_key(),
1558 })
1559
1560 def process_ie_result(self, ie_result, download=True, extra_info=None):
1561 """
1562 Take the result of the ie(may be modified) and resolve all unresolved
1563 references (URLs, playlist items).
1564
1565 It will also download the videos if 'download'.
1566 Returns the resolved ie_result.
1567 """
1568 if extra_info is None:
1569 extra_info = {}
1570 result_type = ie_result.get('_type', 'video')
1571
1572 if result_type in ('url', 'url_transparent'):
1573 ie_result['url'] = sanitize_url(
1574 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1575 if ie_result.get('original_url'):
1576 extra_info.setdefault('original_url', ie_result['original_url'])
1577
1578 extract_flat = self.params.get('extract_flat', False)
1579 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1580 or extract_flat is True):
1581 info_copy = ie_result.copy()
1582 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1583 if ie and not ie_result.get('id'):
1584 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1585 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1586 self.add_extra_info(info_copy, extra_info)
1587 info_copy, _ = self.pre_process(info_copy)
1588 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1589 self._raise_pending_errors(info_copy)
1590 if self.params.get('force_write_download_archive', False):
1591 self.record_download_archive(info_copy)
1592 return ie_result
1593
1594 if result_type == 'video':
1595 self.add_extra_info(ie_result, extra_info)
1596 ie_result = self.process_video_result(ie_result, download=download)
1597 self._raise_pending_errors(ie_result)
1598 additional_urls = (ie_result or {}).get('additional_urls')
1599 if additional_urls:
1600 # TODO: Improve MetadataParserPP to allow setting a list
1601 if isinstance(additional_urls, str):
1602 additional_urls = [additional_urls]
1603 self.to_screen(
1604 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1605 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1606 ie_result['additional_entries'] = [
1607 self.extract_info(
1608 url, download, extra_info=extra_info,
1609 force_generic_extractor=self.params.get('force_generic_extractor'))
1610 for url in additional_urls
1611 ]
1612 return ie_result
1613 elif result_type == 'url':
1614 # We have to add extra_info to the results because it may be
1615 # contained in a playlist
1616 return self.extract_info(
1617 ie_result['url'], download,
1618 ie_key=ie_result.get('ie_key'),
1619 extra_info=extra_info)
1620 elif result_type == 'url_transparent':
1621 # Use the information from the embedding page
1622 info = self.extract_info(
1623 ie_result['url'], ie_key=ie_result.get('ie_key'),
1624 extra_info=extra_info, download=False, process=False)
1625
1626 # extract_info may return None when ignoreerrors is enabled and
1627 # extraction failed with an error, don't crash and return early
1628 # in this case
1629 if not info:
1630 return info
1631
1632 exempted_fields = {'_type', 'url', 'ie_key'}
1633 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1634 # For video clips, the id etc of the clip extractor should be used
1635 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1636
1637 new_result = info.copy()
1638 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1639
1640 # Extracted info may not be a video result (i.e.
1641 # info.get('_type', 'video') != video) but rather an url or
1642 # url_transparent. In such cases outer metadata (from ie_result)
1643 # should be propagated to inner one (info). For this to happen
1644 # _type of info should be overridden with url_transparent. This
1645 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1646 if new_result.get('_type') == 'url':
1647 new_result['_type'] = 'url_transparent'
1648
1649 return self.process_ie_result(
1650 new_result, download=download, extra_info=extra_info)
1651 elif result_type in ('playlist', 'multi_video'):
1652 # Protect from infinite recursion due to recursively nested playlists
1653 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1654 webpage_url = ie_result['webpage_url']
1655 if webpage_url in self._playlist_urls:
1656 self.to_screen(
1657 '[download] Skipping already downloaded playlist: %s'
1658 % ie_result.get('title') or ie_result.get('id'))
1659 return
1660
1661 self._playlist_level += 1
1662 self._playlist_urls.add(webpage_url)
1663 self._fill_common_fields(ie_result, False)
1664 self._sanitize_thumbnails(ie_result)
1665 try:
1666 return self.__process_playlist(ie_result, download)
1667 finally:
1668 self._playlist_level -= 1
1669 if not self._playlist_level:
1670 self._playlist_urls.clear()
1671 elif result_type == 'compat_list':
1672 self.report_warning(
1673 'Extractor %s returned a compat_list result. '
1674 'It needs to be updated.' % ie_result.get('extractor'))
1675
1676 def _fixup(r):
1677 self.add_extra_info(r, {
1678 'extractor': ie_result['extractor'],
1679 'webpage_url': ie_result['webpage_url'],
1680 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1681 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1682 'extractor_key': ie_result['extractor_key'],
1683 })
1684 return r
1685 ie_result['entries'] = [
1686 self.process_ie_result(_fixup(r), download, extra_info)
1687 for r in ie_result['entries']
1688 ]
1689 return ie_result
1690 else:
1691 raise Exception('Invalid result type: %s' % result_type)
1692
1693 def _ensure_dir_exists(self, path):
1694 return make_dir(path, self.report_error)
1695
1696 @staticmethod
1697 def _playlist_infodict(ie_result, strict=False, **kwargs):
1698 info = {
1699 'playlist_count': ie_result.get('playlist_count'),
1700 'playlist': ie_result.get('title') or ie_result.get('id'),
1701 'playlist_id': ie_result.get('id'),
1702 'playlist_title': ie_result.get('title'),
1703 'playlist_uploader': ie_result.get('uploader'),
1704 'playlist_uploader_id': ie_result.get('uploader_id'),
1705 **kwargs,
1706 }
1707 if strict:
1708 return info
1709 return {
1710 **info,
1711 'playlist_index': 0,
1712 '__last_playlist_index': max(ie_result['requested_entries'] or (0, 0)),
1713 'extractor': ie_result['extractor'],
1714 'webpage_url': ie_result['webpage_url'],
1715 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1716 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1717 'extractor_key': ie_result['extractor_key'],
1718 }
1719
1720 def __process_playlist(self, ie_result, download):
1721 """Process each entry in the playlist"""
1722 assert ie_result['_type'] in ('playlist', 'multi_video')
1723
1724 common_info = self._playlist_infodict(ie_result, strict=True)
1725 title = common_info.get('playlist') or '<Untitled>'
1726 if self._match_entry(common_info, incomplete=True) is not None:
1727 return
1728 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1729
1730 all_entries = PlaylistEntries(self, ie_result)
1731 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1732
1733 lazy = self.params.get('lazy_playlist')
1734 if lazy:
1735 resolved_entries, n_entries = [], 'N/A'
1736 ie_result['requested_entries'], ie_result['entries'] = None, None
1737 else:
1738 entries = resolved_entries = list(entries)
1739 n_entries = len(resolved_entries)
1740 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1741 if not ie_result.get('playlist_count'):
1742 # Better to do this after potentially exhausting entries
1743 ie_result['playlist_count'] = all_entries.get_full_count()
1744
1745 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1746 ie_copy = collections.ChainMap(ie_result, extra)
1747
1748 _infojson_written = False
1749 write_playlist_files = self.params.get('allow_playlist_files', True)
1750 if write_playlist_files and self.params.get('list_thumbnails'):
1751 self.list_thumbnails(ie_result)
1752 if write_playlist_files and not self.params.get('simulate'):
1753 _infojson_written = self._write_info_json(
1754 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1755 if _infojson_written is None:
1756 return
1757 if self._write_description('playlist', ie_result,
1758 self.prepare_filename(ie_copy, 'pl_description')) is None:
1759 return
1760 # TODO: This should be passed to ThumbnailsConvertor if necessary
1761 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1762
1763 if lazy:
1764 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1765 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1766 elif self.params.get('playlistreverse'):
1767 entries.reverse()
1768 elif self.params.get('playlistrandom'):
1769 random.shuffle(entries)
1770
1771 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} videos'
1772 f'{format_field(ie_result, "playlist_count", " of %s")}')
1773
1774 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1775 if self.params.get('extract_flat') == 'discard_in_playlist':
1776 keep_resolved_entries = ie_result['_type'] != 'playlist'
1777 if keep_resolved_entries:
1778 self.write_debug('The information of all playlist entries will be held in memory')
1779
1780 failures = 0
1781 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1782 for i, (playlist_index, entry) in enumerate(entries):
1783 if lazy:
1784 resolved_entries.append((playlist_index, entry))
1785 if not entry:
1786 continue
1787
1788 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1789 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1790 playlist_index = ie_result['requested_entries'][i]
1791
1792 entry_copy = collections.ChainMap(entry, {
1793 **common_info,
1794 'n_entries': int_or_none(n_entries),
1795 'playlist_index': playlist_index,
1796 'playlist_autonumber': i + 1,
1797 })
1798
1799 if self._match_entry(entry_copy, incomplete=True) is not None:
1800 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
1801 resolved_entries[i] = (playlist_index, NO_DEFAULT)
1802 continue
1803
1804 self.to_screen('[download] Downloading video %s of %s' % (
1805 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1806
1807 extra.update({
1808 'playlist_index': playlist_index,
1809 'playlist_autonumber': i + 1,
1810 })
1811 entry_result = self.__process_iterable_entry(entry, download, extra)
1812 if not entry_result:
1813 failures += 1
1814 if failures >= max_failures:
1815 self.report_error(
1816 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1817 break
1818 if keep_resolved_entries:
1819 resolved_entries[i] = (playlist_index, entry_result)
1820
1821 # Update with processed data
1822 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
1823 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
1824
1825 # Write the updated info to json
1826 if _infojson_written is True and self._write_info_json(
1827 'updated playlist', ie_result,
1828 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1829 return
1830
1831 ie_result = self.run_all_pps('playlist', ie_result)
1832 self.to_screen(f'[download] Finished downloading playlist: {title}')
1833 return ie_result
1834
1835 @_handle_extraction_exceptions
1836 def __process_iterable_entry(self, entry, download, extra_info):
1837 return self.process_ie_result(
1838 entry, download=download, extra_info=extra_info)
1839
1840 def _build_format_filter(self, filter_spec):
1841 " Returns a function to filter the formats according to the filter_spec "
1842
1843 OPERATORS = {
1844 '<': operator.lt,
1845 '<=': operator.le,
1846 '>': operator.gt,
1847 '>=': operator.ge,
1848 '=': operator.eq,
1849 '!=': operator.ne,
1850 }
1851 operator_rex = re.compile(r'''(?x)\s*
1852 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1853 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1854 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1855 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1856 m = operator_rex.fullmatch(filter_spec)
1857 if m:
1858 try:
1859 comparison_value = int(m.group('value'))
1860 except ValueError:
1861 comparison_value = parse_filesize(m.group('value'))
1862 if comparison_value is None:
1863 comparison_value = parse_filesize(m.group('value') + 'B')
1864 if comparison_value is None:
1865 raise ValueError(
1866 'Invalid value %r in format specification %r' % (
1867 m.group('value'), filter_spec))
1868 op = OPERATORS[m.group('op')]
1869
1870 if not m:
1871 STR_OPERATORS = {
1872 '=': operator.eq,
1873 '^=': lambda attr, value: attr.startswith(value),
1874 '$=': lambda attr, value: attr.endswith(value),
1875 '*=': lambda attr, value: value in attr,
1876 '~=': lambda attr, value: value.search(attr) is not None
1877 }
1878 str_operator_rex = re.compile(r'''(?x)\s*
1879 (?P<key>[a-zA-Z0-9._-]+)\s*
1880 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1881 (?P<quote>["'])?
1882 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1883 (?(quote)(?P=quote))\s*
1884 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1885 m = str_operator_rex.fullmatch(filter_spec)
1886 if m:
1887 if m.group('op') == '~=':
1888 comparison_value = re.compile(m.group('value'))
1889 else:
1890 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1891 str_op = STR_OPERATORS[m.group('op')]
1892 if m.group('negation'):
1893 op = lambda attr, value: not str_op(attr, value)
1894 else:
1895 op = str_op
1896
1897 if not m:
1898 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1899
1900 def _filter(f):
1901 actual_value = f.get(m.group('key'))
1902 if actual_value is None:
1903 return m.group('none_inclusive')
1904 return op(actual_value, comparison_value)
1905 return _filter
1906
1907 def _check_formats(self, formats):
1908 for f in formats:
1909 self.to_screen('[info] Testing format %s' % f['format_id'])
1910 path = self.get_output_path('temp')
1911 if not self._ensure_dir_exists(f'{path}/'):
1912 continue
1913 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1914 temp_file.close()
1915 try:
1916 success, _ = self.dl(temp_file.name, f, test=True)
1917 except (DownloadError, OSError, ValueError) + network_exceptions:
1918 success = False
1919 finally:
1920 if os.path.exists(temp_file.name):
1921 try:
1922 os.remove(temp_file.name)
1923 except OSError:
1924 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1925 if success:
1926 yield f
1927 else:
1928 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1929
1930 def _default_format_spec(self, info_dict, download=True):
1931
1932 def can_merge():
1933 merger = FFmpegMergerPP(self)
1934 return merger.available and merger.can_merge()
1935
1936 prefer_best = (
1937 not self.params.get('simulate')
1938 and download
1939 and (
1940 not can_merge()
1941 or info_dict.get('is_live') and not self.params.get('live_from_start')
1942 or self.params['outtmpl']['default'] == '-'))
1943 compat = (
1944 prefer_best
1945 or self.params.get('allow_multiple_audio_streams', False)
1946 or 'format-spec' in self.params['compat_opts'])
1947
1948 return (
1949 'best/bestvideo+bestaudio' if prefer_best
1950 else 'bestvideo*+bestaudio/best' if not compat
1951 else 'bestvideo+bestaudio/best')
1952
1953 def build_format_selector(self, format_spec):
1954 def syntax_error(note, start):
1955 message = (
1956 'Invalid format specification: '
1957 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1958 return SyntaxError(message)
1959
1960 PICKFIRST = 'PICKFIRST'
1961 MERGE = 'MERGE'
1962 SINGLE = 'SINGLE'
1963 GROUP = 'GROUP'
1964 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1965
1966 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1967 'video': self.params.get('allow_multiple_video_streams', False)}
1968
1969 check_formats = self.params.get('check_formats') == 'selected'
1970
1971 def _parse_filter(tokens):
1972 filter_parts = []
1973 for type, string, start, _, _ in tokens:
1974 if type == tokenize.OP and string == ']':
1975 return ''.join(filter_parts)
1976 else:
1977 filter_parts.append(string)
1978
1979 def _remove_unused_ops(tokens):
1980 # Remove operators that we don't use and join them with the surrounding strings
1981 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1982 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1983 last_string, last_start, last_end, last_line = None, None, None, None
1984 for type, string, start, end, line in tokens:
1985 if type == tokenize.OP and string == '[':
1986 if last_string:
1987 yield tokenize.NAME, last_string, last_start, last_end, last_line
1988 last_string = None
1989 yield type, string, start, end, line
1990 # everything inside brackets will be handled by _parse_filter
1991 for type, string, start, end, line in tokens:
1992 yield type, string, start, end, line
1993 if type == tokenize.OP and string == ']':
1994 break
1995 elif type == tokenize.OP and string in ALLOWED_OPS:
1996 if last_string:
1997 yield tokenize.NAME, last_string, last_start, last_end, last_line
1998 last_string = None
1999 yield type, string, start, end, line
2000 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2001 if not last_string:
2002 last_string = string
2003 last_start = start
2004 last_end = end
2005 else:
2006 last_string += string
2007 if last_string:
2008 yield tokenize.NAME, last_string, last_start, last_end, last_line
2009
2010 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2011 selectors = []
2012 current_selector = None
2013 for type, string, start, _, _ in tokens:
2014 # ENCODING is only defined in python 3.x
2015 if type == getattr(tokenize, 'ENCODING', None):
2016 continue
2017 elif type in [tokenize.NAME, tokenize.NUMBER]:
2018 current_selector = FormatSelector(SINGLE, string, [])
2019 elif type == tokenize.OP:
2020 if string == ')':
2021 if not inside_group:
2022 # ')' will be handled by the parentheses group
2023 tokens.restore_last_token()
2024 break
2025 elif inside_merge and string in ['/', ',']:
2026 tokens.restore_last_token()
2027 break
2028 elif inside_choice and string == ',':
2029 tokens.restore_last_token()
2030 break
2031 elif string == ',':
2032 if not current_selector:
2033 raise syntax_error('"," must follow a format selector', start)
2034 selectors.append(current_selector)
2035 current_selector = None
2036 elif string == '/':
2037 if not current_selector:
2038 raise syntax_error('"/" must follow a format selector', start)
2039 first_choice = current_selector
2040 second_choice = _parse_format_selection(tokens, inside_choice=True)
2041 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2042 elif string == '[':
2043 if not current_selector:
2044 current_selector = FormatSelector(SINGLE, 'best', [])
2045 format_filter = _parse_filter(tokens)
2046 current_selector.filters.append(format_filter)
2047 elif string == '(':
2048 if current_selector:
2049 raise syntax_error('Unexpected "("', start)
2050 group = _parse_format_selection(tokens, inside_group=True)
2051 current_selector = FormatSelector(GROUP, group, [])
2052 elif string == '+':
2053 if not current_selector:
2054 raise syntax_error('Unexpected "+"', start)
2055 selector_1 = current_selector
2056 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2057 if not selector_2:
2058 raise syntax_error('Expected a selector', start)
2059 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2060 else:
2061 raise syntax_error(f'Operator not recognized: "{string}"', start)
2062 elif type == tokenize.ENDMARKER:
2063 break
2064 if current_selector:
2065 selectors.append(current_selector)
2066 return selectors
2067
2068 def _merge(formats_pair):
2069 format_1, format_2 = formats_pair
2070
2071 formats_info = []
2072 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2073 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2074
2075 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2076 get_no_more = {'video': False, 'audio': False}
2077 for (i, fmt_info) in enumerate(formats_info):
2078 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2079 formats_info.pop(i)
2080 continue
2081 for aud_vid in ['audio', 'video']:
2082 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2083 if get_no_more[aud_vid]:
2084 formats_info.pop(i)
2085 break
2086 get_no_more[aud_vid] = True
2087
2088 if len(formats_info) == 1:
2089 return formats_info[0]
2090
2091 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2092 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2093
2094 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2095 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2096
2097 output_ext = get_compatible_ext(
2098 vcodecs=[f.get('vcodec') for f in video_fmts],
2099 acodecs=[f.get('acodec') for f in audio_fmts],
2100 vexts=[f['ext'] for f in video_fmts],
2101 aexts=[f['ext'] for f in audio_fmts],
2102 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2103 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2104
2105 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2106
2107 new_dict = {
2108 'requested_formats': formats_info,
2109 'format': '+'.join(filtered('format')),
2110 'format_id': '+'.join(filtered('format_id')),
2111 'ext': output_ext,
2112 'protocol': '+'.join(map(determine_protocol, formats_info)),
2113 'language': '+'.join(orderedSet(filtered('language'))) or None,
2114 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2115 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2116 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2117 }
2118
2119 if the_only_video:
2120 new_dict.update({
2121 'width': the_only_video.get('width'),
2122 'height': the_only_video.get('height'),
2123 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2124 'fps': the_only_video.get('fps'),
2125 'dynamic_range': the_only_video.get('dynamic_range'),
2126 'vcodec': the_only_video.get('vcodec'),
2127 'vbr': the_only_video.get('vbr'),
2128 'stretched_ratio': the_only_video.get('stretched_ratio'),
2129 })
2130
2131 if the_only_audio:
2132 new_dict.update({
2133 'acodec': the_only_audio.get('acodec'),
2134 'abr': the_only_audio.get('abr'),
2135 'asr': the_only_audio.get('asr'),
2136 'audio_channels': the_only_audio.get('audio_channels')
2137 })
2138
2139 return new_dict
2140
2141 def _check_formats(formats):
2142 if not check_formats:
2143 yield from formats
2144 return
2145 yield from self._check_formats(formats)
2146
2147 def _build_selector_function(selector):
2148 if isinstance(selector, list): # ,
2149 fs = [_build_selector_function(s) for s in selector]
2150
2151 def selector_function(ctx):
2152 for f in fs:
2153 yield from f(ctx)
2154 return selector_function
2155
2156 elif selector.type == GROUP: # ()
2157 selector_function = _build_selector_function(selector.selector)
2158
2159 elif selector.type == PICKFIRST: # /
2160 fs = [_build_selector_function(s) for s in selector.selector]
2161
2162 def selector_function(ctx):
2163 for f in fs:
2164 picked_formats = list(f(ctx))
2165 if picked_formats:
2166 return picked_formats
2167 return []
2168
2169 elif selector.type == MERGE: # +
2170 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2171
2172 def selector_function(ctx):
2173 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2174 yield _merge(pair)
2175
2176 elif selector.type == SINGLE: # atom
2177 format_spec = selector.selector or 'best'
2178
2179 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2180 if format_spec == 'all':
2181 def selector_function(ctx):
2182 yield from _check_formats(ctx['formats'][::-1])
2183 elif format_spec == 'mergeall':
2184 def selector_function(ctx):
2185 formats = list(_check_formats(
2186 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2187 if not formats:
2188 return
2189 merged_format = formats[-1]
2190 for f in formats[-2::-1]:
2191 merged_format = _merge((merged_format, f))
2192 yield merged_format
2193
2194 else:
2195 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2196 mobj = re.match(
2197 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2198 format_spec)
2199 if mobj is not None:
2200 format_idx = int_or_none(mobj.group('n'), default=1)
2201 format_reverse = mobj.group('bw')[0] == 'b'
2202 format_type = (mobj.group('type') or [None])[0]
2203 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2204 format_modified = mobj.group('mod') is not None
2205
2206 format_fallback = not format_type and not format_modified # for b, w
2207 _filter_f = (
2208 (lambda f: f.get('%scodec' % format_type) != 'none')
2209 if format_type and format_modified # bv*, ba*, wv*, wa*
2210 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2211 if format_type # bv, ba, wv, wa
2212 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2213 if not format_modified # b, w
2214 else lambda f: True) # b*, w*
2215 filter_f = lambda f: _filter_f(f) and (
2216 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2217 else:
2218 if format_spec in self._format_selection_exts['audio']:
2219 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2220 elif format_spec in self._format_selection_exts['video']:
2221 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2222 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2223 elif format_spec in self._format_selection_exts['storyboards']:
2224 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2225 else:
2226 filter_f = lambda f: f.get('format_id') == format_spec # id
2227
2228 def selector_function(ctx):
2229 formats = list(ctx['formats'])
2230 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2231 if not matches:
2232 if format_fallback and ctx['incomplete_formats']:
2233 # for extractors with incomplete formats (audio only (soundcloud)
2234 # or video only (imgur)) best/worst will fallback to
2235 # best/worst {video,audio}-only format
2236 matches = formats
2237 elif seperate_fallback and not ctx['has_merged_format']:
2238 # for compatibility with youtube-dl when there is no pre-merged format
2239 matches = list(filter(seperate_fallback, formats))
2240 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2241 try:
2242 yield matches[format_idx - 1]
2243 except LazyList.IndexError:
2244 return
2245
2246 filters = [self._build_format_filter(f) for f in selector.filters]
2247
2248 def final_selector(ctx):
2249 ctx_copy = dict(ctx)
2250 for _filter in filters:
2251 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2252 return selector_function(ctx_copy)
2253 return final_selector
2254
2255 stream = io.BytesIO(format_spec.encode())
2256 try:
2257 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2258 except tokenize.TokenError:
2259 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2260
2261 class TokenIterator:
2262 def __init__(self, tokens):
2263 self.tokens = tokens
2264 self.counter = 0
2265
2266 def __iter__(self):
2267 return self
2268
2269 def __next__(self):
2270 if self.counter >= len(self.tokens):
2271 raise StopIteration()
2272 value = self.tokens[self.counter]
2273 self.counter += 1
2274 return value
2275
2276 next = __next__
2277
2278 def restore_last_token(self):
2279 self.counter -= 1
2280
2281 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2282 return _build_selector_function(parsed_selector)
2283
2284 def _calc_headers(self, info_dict):
2285 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2286
2287 cookies = self._calc_cookies(info_dict['url'])
2288 if cookies:
2289 res['Cookie'] = cookies
2290
2291 if 'X-Forwarded-For' not in res:
2292 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2293 if x_forwarded_for_ip:
2294 res['X-Forwarded-For'] = x_forwarded_for_ip
2295
2296 return res
2297
2298 def _calc_cookies(self, url):
2299 pr = sanitized_Request(url)
2300 self.cookiejar.add_cookie_header(pr)
2301 return pr.get_header('Cookie')
2302
2303 def _sort_thumbnails(self, thumbnails):
2304 thumbnails.sort(key=lambda t: (
2305 t.get('preference') if t.get('preference') is not None else -1,
2306 t.get('width') if t.get('width') is not None else -1,
2307 t.get('height') if t.get('height') is not None else -1,
2308 t.get('id') if t.get('id') is not None else '',
2309 t.get('url')))
2310
2311 def _sanitize_thumbnails(self, info_dict):
2312 thumbnails = info_dict.get('thumbnails')
2313 if thumbnails is None:
2314 thumbnail = info_dict.get('thumbnail')
2315 if thumbnail:
2316 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2317 if not thumbnails:
2318 return
2319
2320 def check_thumbnails(thumbnails):
2321 for t in thumbnails:
2322 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2323 try:
2324 self.urlopen(HEADRequest(t['url']))
2325 except network_exceptions as err:
2326 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2327 continue
2328 yield t
2329
2330 self._sort_thumbnails(thumbnails)
2331 for i, t in enumerate(thumbnails):
2332 if t.get('id') is None:
2333 t['id'] = '%d' % i
2334 if t.get('width') and t.get('height'):
2335 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2336 t['url'] = sanitize_url(t['url'])
2337
2338 if self.params.get('check_formats') is True:
2339 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2340 else:
2341 info_dict['thumbnails'] = thumbnails
2342
2343 def _fill_common_fields(self, info_dict, is_video=True):
2344 # TODO: move sanitization here
2345 if is_video:
2346 # playlists are allowed to lack "title"
2347 title = info_dict.get('title', NO_DEFAULT)
2348 if title is NO_DEFAULT:
2349 raise ExtractorError('Missing "title" field in extractor result',
2350 video_id=info_dict['id'], ie=info_dict['extractor'])
2351 info_dict['fulltitle'] = title
2352 if not title:
2353 if title == '':
2354 self.write_debug('Extractor gave empty title. Creating a generic title')
2355 else:
2356 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2357 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2358
2359 if info_dict.get('duration') is not None:
2360 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2361
2362 for ts_key, date_key in (
2363 ('timestamp', 'upload_date'),
2364 ('release_timestamp', 'release_date'),
2365 ('modified_timestamp', 'modified_date'),
2366 ):
2367 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2368 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2369 # see http://bugs.python.org/issue1646728)
2370 with contextlib.suppress(ValueError, OverflowError, OSError):
2371 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2372 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2373
2374 live_keys = ('is_live', 'was_live')
2375 live_status = info_dict.get('live_status')
2376 if live_status is None:
2377 for key in live_keys:
2378 if info_dict.get(key) is False:
2379 continue
2380 if info_dict.get(key):
2381 live_status = key
2382 break
2383 if all(info_dict.get(key) is False for key in live_keys):
2384 live_status = 'not_live'
2385 if live_status:
2386 info_dict['live_status'] = live_status
2387 for key in live_keys:
2388 if info_dict.get(key) is None:
2389 info_dict[key] = (live_status == key)
2390
2391 # Auto generate title fields corresponding to the *_number fields when missing
2392 # in order to always have clean titles. This is very common for TV series.
2393 for field in ('chapter', 'season', 'episode'):
2394 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2395 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2396
2397 def _raise_pending_errors(self, info):
2398 err = info.pop('__pending_error', None)
2399 if err:
2400 self.report_error(err, tb=False)
2401
2402 def process_video_result(self, info_dict, download=True):
2403 assert info_dict.get('_type', 'video') == 'video'
2404 self._num_videos += 1
2405
2406 if 'id' not in info_dict:
2407 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2408 elif not info_dict.get('id'):
2409 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2410
2411 def report_force_conversion(field, field_not, conversion):
2412 self.report_warning(
2413 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2414 % (field, field_not, conversion))
2415
2416 def sanitize_string_field(info, string_field):
2417 field = info.get(string_field)
2418 if field is None or isinstance(field, str):
2419 return
2420 report_force_conversion(string_field, 'a string', 'string')
2421 info[string_field] = str(field)
2422
2423 def sanitize_numeric_fields(info):
2424 for numeric_field in self._NUMERIC_FIELDS:
2425 field = info.get(numeric_field)
2426 if field is None or isinstance(field, (int, float)):
2427 continue
2428 report_force_conversion(numeric_field, 'numeric', 'int')
2429 info[numeric_field] = int_or_none(field)
2430
2431 sanitize_string_field(info_dict, 'id')
2432 sanitize_numeric_fields(info_dict)
2433 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2434 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2435 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2436 self.report_warning('"duration" field is negative, there is an error in extractor')
2437
2438 chapters = info_dict.get('chapters') or []
2439 if chapters and chapters[0].get('start_time'):
2440 chapters.insert(0, {'start_time': 0})
2441
2442 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2443 for idx, (prev, current, next_) in enumerate(zip(
2444 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2445 if current.get('start_time') is None:
2446 current['start_time'] = prev.get('end_time')
2447 if not current.get('end_time'):
2448 current['end_time'] = next_.get('start_time')
2449 if not current.get('title'):
2450 current['title'] = f'<Untitled Chapter {idx}>'
2451
2452 if 'playlist' not in info_dict:
2453 # It isn't part of a playlist
2454 info_dict['playlist'] = None
2455 info_dict['playlist_index'] = None
2456
2457 self._sanitize_thumbnails(info_dict)
2458
2459 thumbnail = info_dict.get('thumbnail')
2460 thumbnails = info_dict.get('thumbnails')
2461 if thumbnail:
2462 info_dict['thumbnail'] = sanitize_url(thumbnail)
2463 elif thumbnails:
2464 info_dict['thumbnail'] = thumbnails[-1]['url']
2465
2466 if info_dict.get('display_id') is None and 'id' in info_dict:
2467 info_dict['display_id'] = info_dict['id']
2468
2469 self._fill_common_fields(info_dict)
2470
2471 for cc_kind in ('subtitles', 'automatic_captions'):
2472 cc = info_dict.get(cc_kind)
2473 if cc:
2474 for _, subtitle in cc.items():
2475 for subtitle_format in subtitle:
2476 if subtitle_format.get('url'):
2477 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2478 if subtitle_format.get('ext') is None:
2479 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2480
2481 automatic_captions = info_dict.get('automatic_captions')
2482 subtitles = info_dict.get('subtitles')
2483
2484 info_dict['requested_subtitles'] = self.process_subtitles(
2485 info_dict['id'], subtitles, automatic_captions)
2486
2487 if info_dict.get('formats') is None:
2488 # There's only one format available
2489 formats = [info_dict]
2490 else:
2491 formats = info_dict['formats']
2492
2493 # or None ensures --clean-infojson removes it
2494 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2495 if not self.params.get('allow_unplayable_formats'):
2496 formats = [f for f in formats if not f.get('has_drm')]
2497 if info_dict['_has_drm'] and formats and all(
2498 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2499 self.report_warning(
2500 'This video is DRM protected and only images are available for download. '
2501 'Use --list-formats to see them')
2502
2503 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2504 if not get_from_start:
2505 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2506 if info_dict.get('is_live') and formats:
2507 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2508 if get_from_start and not formats:
2509 self.raise_no_formats(info_dict, msg=(
2510 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2511 'If you want to download from the current time, use --no-live-from-start'))
2512
2513 if not formats:
2514 self.raise_no_formats(info_dict)
2515
2516 def is_wellformed(f):
2517 url = f.get('url')
2518 if not url:
2519 self.report_warning(
2520 '"url" field is missing or empty - skipping format, '
2521 'there is an error in extractor')
2522 return False
2523 if isinstance(url, bytes):
2524 sanitize_string_field(f, 'url')
2525 return True
2526
2527 # Filter out malformed formats for better extraction robustness
2528 formats = list(filter(is_wellformed, formats))
2529
2530 formats_dict = {}
2531
2532 # We check that all the formats have the format and format_id fields
2533 for i, format in enumerate(formats):
2534 sanitize_string_field(format, 'format_id')
2535 sanitize_numeric_fields(format)
2536 format['url'] = sanitize_url(format['url'])
2537 if not format.get('format_id'):
2538 format['format_id'] = str(i)
2539 else:
2540 # Sanitize format_id from characters used in format selector expression
2541 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2542 format_id = format['format_id']
2543 if format_id not in formats_dict:
2544 formats_dict[format_id] = []
2545 formats_dict[format_id].append(format)
2546
2547 # Make sure all formats have unique format_id
2548 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2549 for format_id, ambiguous_formats in formats_dict.items():
2550 ambigious_id = len(ambiguous_formats) > 1
2551 for i, format in enumerate(ambiguous_formats):
2552 if ambigious_id:
2553 format['format_id'] = '%s-%d' % (format_id, i)
2554 if format.get('ext') is None:
2555 format['ext'] = determine_ext(format['url']).lower()
2556 # Ensure there is no conflict between id and ext in format selection
2557 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2558 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2559 format['format_id'] = 'f%s' % format['format_id']
2560
2561 for i, format in enumerate(formats):
2562 if format.get('format') is None:
2563 format['format'] = '{id} - {res}{note}'.format(
2564 id=format['format_id'],
2565 res=self.format_resolution(format),
2566 note=format_field(format, 'format_note', ' (%s)'),
2567 )
2568 if format.get('protocol') is None:
2569 format['protocol'] = determine_protocol(format)
2570 if format.get('resolution') is None:
2571 format['resolution'] = self.format_resolution(format, default=None)
2572 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2573 format['dynamic_range'] = 'SDR'
2574 if (info_dict.get('duration') and format.get('tbr')
2575 and not format.get('filesize') and not format.get('filesize_approx')):
2576 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2577
2578 # Add HTTP headers, so that external programs can use them from the
2579 # json output
2580 full_format_info = info_dict.copy()
2581 full_format_info.update(format)
2582 format['http_headers'] = self._calc_headers(full_format_info)
2583 # Remove private housekeeping stuff
2584 if '__x_forwarded_for_ip' in info_dict:
2585 del info_dict['__x_forwarded_for_ip']
2586
2587 if self.params.get('check_formats') is True:
2588 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2589
2590 if not formats or formats[0] is not info_dict:
2591 # only set the 'formats' fields if the original info_dict list them
2592 # otherwise we end up with a circular reference, the first (and unique)
2593 # element in the 'formats' field in info_dict is info_dict itself,
2594 # which can't be exported to json
2595 info_dict['formats'] = formats
2596
2597 info_dict, _ = self.pre_process(info_dict)
2598
2599 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2600 return info_dict
2601
2602 self.post_extract(info_dict)
2603 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2604
2605 # The pre-processors may have modified the formats
2606 formats = info_dict.get('formats', [info_dict])
2607
2608 list_only = self.params.get('simulate') is None and (
2609 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2610 interactive_format_selection = not list_only and self.format_selector == '-'
2611 if self.params.get('list_thumbnails'):
2612 self.list_thumbnails(info_dict)
2613 if self.params.get('listsubtitles'):
2614 if 'automatic_captions' in info_dict:
2615 self.list_subtitles(
2616 info_dict['id'], automatic_captions, 'automatic captions')
2617 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2618 if self.params.get('listformats') or interactive_format_selection:
2619 self.list_formats(info_dict)
2620 if list_only:
2621 # Without this printing, -F --print-json will not work
2622 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2623 return info_dict
2624
2625 format_selector = self.format_selector
2626 if format_selector is None:
2627 req_format = self._default_format_spec(info_dict, download=download)
2628 self.write_debug('Default format spec: %s' % req_format)
2629 format_selector = self.build_format_selector(req_format)
2630
2631 while True:
2632 if interactive_format_selection:
2633 req_format = input(
2634 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2635 try:
2636 format_selector = self.build_format_selector(req_format)
2637 except SyntaxError as err:
2638 self.report_error(err, tb=False, is_error=False)
2639 continue
2640
2641 formats_to_download = list(format_selector({
2642 'formats': formats,
2643 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2644 'incomplete_formats': (
2645 # All formats are video-only or
2646 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2647 # all formats are audio-only
2648 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2649 }))
2650 if interactive_format_selection and not formats_to_download:
2651 self.report_error('Requested format is not available', tb=False, is_error=False)
2652 continue
2653 break
2654
2655 if not formats_to_download:
2656 if not self.params.get('ignore_no_formats_error'):
2657 raise ExtractorError(
2658 'Requested format is not available. Use --list-formats for a list of available formats',
2659 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2660 self.report_warning('Requested format is not available')
2661 # Process what we can, even without any available formats.
2662 formats_to_download = [{}]
2663
2664 requested_ranges = self.params.get('download_ranges')
2665 if requested_ranges:
2666 requested_ranges = tuple(requested_ranges(info_dict, self))
2667
2668 best_format, downloaded_formats = formats_to_download[-1], []
2669 if download:
2670 if best_format:
2671 def to_screen(*msg):
2672 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2673
2674 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2675 (f['format_id'] for f in formats_to_download))
2676 if requested_ranges:
2677 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2678 (f'{int(c["start_time"])}-{int(c["end_time"])}' for c in requested_ranges))
2679 max_downloads_reached = False
2680
2681 for fmt, chapter in itertools.product(formats_to_download, requested_ranges or [{}]):
2682 new_info = self._copy_infodict(info_dict)
2683 new_info.update(fmt)
2684 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2685 if chapter or offset:
2686 new_info.update({
2687 'section_start': offset + chapter.get('start_time', 0),
2688 'section_end': offset + min(chapter.get('end_time', duration), duration),
2689 'section_title': chapter.get('title'),
2690 'section_number': chapter.get('index'),
2691 })
2692 downloaded_formats.append(new_info)
2693 try:
2694 self.process_info(new_info)
2695 except MaxDownloadsReached:
2696 max_downloads_reached = True
2697 self._raise_pending_errors(new_info)
2698 # Remove copied info
2699 for key, val in tuple(new_info.items()):
2700 if info_dict.get(key) == val:
2701 new_info.pop(key)
2702 if max_downloads_reached:
2703 break
2704
2705 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2706 assert write_archive.issubset({True, False, 'ignore'})
2707 if True in write_archive and False not in write_archive:
2708 self.record_download_archive(info_dict)
2709
2710 info_dict['requested_downloads'] = downloaded_formats
2711 info_dict = self.run_all_pps('after_video', info_dict)
2712 if max_downloads_reached:
2713 raise MaxDownloadsReached()
2714
2715 # We update the info dict with the selected best quality format (backwards compatibility)
2716 info_dict.update(best_format)
2717 return info_dict
2718
2719 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2720 """Select the requested subtitles and their format"""
2721 available_subs, normal_sub_langs = {}, []
2722 if normal_subtitles and self.params.get('writesubtitles'):
2723 available_subs.update(normal_subtitles)
2724 normal_sub_langs = tuple(normal_subtitles.keys())
2725 if automatic_captions and self.params.get('writeautomaticsub'):
2726 for lang, cap_info in automatic_captions.items():
2727 if lang not in available_subs:
2728 available_subs[lang] = cap_info
2729
2730 if (not self.params.get('writesubtitles') and not
2731 self.params.get('writeautomaticsub') or not
2732 available_subs):
2733 return None
2734
2735 all_sub_langs = tuple(available_subs.keys())
2736 if self.params.get('allsubtitles', False):
2737 requested_langs = all_sub_langs
2738 elif self.params.get('subtitleslangs', False):
2739 # A list is used so that the order of languages will be the same as
2740 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2741 requested_langs = []
2742 for lang_re in self.params.get('subtitleslangs'):
2743 discard = lang_re[0] == '-'
2744 if discard:
2745 lang_re = lang_re[1:]
2746 if lang_re == 'all':
2747 if discard:
2748 requested_langs = []
2749 else:
2750 requested_langs.extend(all_sub_langs)
2751 continue
2752 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2753 if discard:
2754 for lang in current_langs:
2755 while lang in requested_langs:
2756 requested_langs.remove(lang)
2757 else:
2758 requested_langs.extend(current_langs)
2759 requested_langs = orderedSet(requested_langs)
2760 elif normal_sub_langs:
2761 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2762 else:
2763 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2764 if requested_langs:
2765 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2766
2767 formats_query = self.params.get('subtitlesformat', 'best')
2768 formats_preference = formats_query.split('/') if formats_query else []
2769 subs = {}
2770 for lang in requested_langs:
2771 formats = available_subs.get(lang)
2772 if formats is None:
2773 self.report_warning(f'{lang} subtitles not available for {video_id}')
2774 continue
2775 for ext in formats_preference:
2776 if ext == 'best':
2777 f = formats[-1]
2778 break
2779 matches = list(filter(lambda f: f['ext'] == ext, formats))
2780 if matches:
2781 f = matches[-1]
2782 break
2783 else:
2784 f = formats[-1]
2785 self.report_warning(
2786 'No subtitle format found matching "%s" for language %s, '
2787 'using %s' % (formats_query, lang, f['ext']))
2788 subs[lang] = f
2789 return subs
2790
2791 def _forceprint(self, key, info_dict):
2792 if info_dict is None:
2793 return
2794 info_copy = info_dict.copy()
2795 info_copy['formats_table'] = self.render_formats_table(info_dict)
2796 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2797 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2798 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2799
2800 def format_tmpl(tmpl):
2801 mobj = re.match(r'\w+(=?)$', tmpl)
2802 if mobj and mobj.group(1):
2803 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2804 elif mobj:
2805 return f'%({tmpl})s'
2806 return tmpl
2807
2808 for tmpl in self.params['forceprint'].get(key, []):
2809 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2810
2811 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2812 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2813 tmpl = format_tmpl(tmpl)
2814 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2815 if self._ensure_dir_exists(filename):
2816 with open(filename, 'a', encoding='utf-8') as f:
2817 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2818
2819 def __forced_printings(self, info_dict, filename, incomplete):
2820 def print_mandatory(field, actual_field=None):
2821 if actual_field is None:
2822 actual_field = field
2823 if (self.params.get('force%s' % field, False)
2824 and (not incomplete or info_dict.get(actual_field) is not None)):
2825 self.to_stdout(info_dict[actual_field])
2826
2827 def print_optional(field):
2828 if (self.params.get('force%s' % field, False)
2829 and info_dict.get(field) is not None):
2830 self.to_stdout(info_dict[field])
2831
2832 info_dict = info_dict.copy()
2833 if filename is not None:
2834 info_dict['filename'] = filename
2835 if info_dict.get('requested_formats') is not None:
2836 # For RTMP URLs, also include the playpath
2837 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2838 elif info_dict.get('url'):
2839 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2840
2841 if (self.params.get('forcejson')
2842 or self.params['forceprint'].get('video')
2843 or self.params['print_to_file'].get('video')):
2844 self.post_extract(info_dict)
2845 self._forceprint('video', info_dict)
2846
2847 print_mandatory('title')
2848 print_mandatory('id')
2849 print_mandatory('url', 'urls')
2850 print_optional('thumbnail')
2851 print_optional('description')
2852 print_optional('filename')
2853 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2854 self.to_stdout(formatSeconds(info_dict['duration']))
2855 print_mandatory('format')
2856
2857 if self.params.get('forcejson'):
2858 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2859
2860 def dl(self, name, info, subtitle=False, test=False):
2861 if not info.get('url'):
2862 self.raise_no_formats(info, True)
2863
2864 if test:
2865 verbose = self.params.get('verbose')
2866 params = {
2867 'test': True,
2868 'quiet': self.params.get('quiet') or not verbose,
2869 'verbose': verbose,
2870 'noprogress': not verbose,
2871 'nopart': True,
2872 'skip_unavailable_fragments': False,
2873 'keep_fragments': False,
2874 'overwrites': True,
2875 '_no_ytdl_file': True,
2876 }
2877 else:
2878 params = self.params
2879 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2880 if not test:
2881 for ph in self._progress_hooks:
2882 fd.add_progress_hook(ph)
2883 urls = '", "'.join(
2884 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2885 for f in info.get('requested_formats', []) or [info])
2886 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2887
2888 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2889 # But it may contain objects that are not deep-copyable
2890 new_info = self._copy_infodict(info)
2891 if new_info.get('http_headers') is None:
2892 new_info['http_headers'] = self._calc_headers(new_info)
2893 return fd.download(name, new_info, subtitle)
2894
2895 def existing_file(self, filepaths, *, default_overwrite=True):
2896 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2897 if existing_files and not self.params.get('overwrites', default_overwrite):
2898 return existing_files[0]
2899
2900 for file in existing_files:
2901 self.report_file_delete(file)
2902 os.remove(file)
2903 return None
2904
2905 def process_info(self, info_dict):
2906 """Process a single resolved IE result. (Modifies it in-place)"""
2907
2908 assert info_dict.get('_type', 'video') == 'video'
2909 original_infodict = info_dict
2910
2911 if 'format' not in info_dict and 'ext' in info_dict:
2912 info_dict['format'] = info_dict['ext']
2913
2914 # This is mostly just for backward compatibility of process_info
2915 # As a side-effect, this allows for format-specific filters
2916 if self._match_entry(info_dict) is not None:
2917 info_dict['__write_download_archive'] = 'ignore'
2918 return
2919
2920 # Does nothing under normal operation - for backward compatibility of process_info
2921 self.post_extract(info_dict)
2922 self._num_downloads += 1
2923
2924 # info_dict['_filename'] needs to be set for backward compatibility
2925 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2926 temp_filename = self.prepare_filename(info_dict, 'temp')
2927 files_to_move = {}
2928
2929 # Forced printings
2930 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2931
2932 def check_max_downloads():
2933 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
2934 raise MaxDownloadsReached()
2935
2936 if self.params.get('simulate'):
2937 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2938 check_max_downloads()
2939 return
2940
2941 if full_filename is None:
2942 return
2943 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2944 return
2945 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2946 return
2947
2948 if self._write_description('video', info_dict,
2949 self.prepare_filename(info_dict, 'description')) is None:
2950 return
2951
2952 sub_files = self._write_subtitles(info_dict, temp_filename)
2953 if sub_files is None:
2954 return
2955 files_to_move.update(dict(sub_files))
2956
2957 thumb_files = self._write_thumbnails(
2958 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2959 if thumb_files is None:
2960 return
2961 files_to_move.update(dict(thumb_files))
2962
2963 infofn = self.prepare_filename(info_dict, 'infojson')
2964 _infojson_written = self._write_info_json('video', info_dict, infofn)
2965 if _infojson_written:
2966 info_dict['infojson_filename'] = infofn
2967 # For backward compatibility, even though it was a private field
2968 info_dict['__infojson_filename'] = infofn
2969 elif _infojson_written is None:
2970 return
2971
2972 # Note: Annotations are deprecated
2973 annofn = None
2974 if self.params.get('writeannotations', False):
2975 annofn = self.prepare_filename(info_dict, 'annotation')
2976 if annofn:
2977 if not self._ensure_dir_exists(encodeFilename(annofn)):
2978 return
2979 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2980 self.to_screen('[info] Video annotations are already present')
2981 elif not info_dict.get('annotations'):
2982 self.report_warning('There are no annotations to write.')
2983 else:
2984 try:
2985 self.to_screen('[info] Writing video annotations to: ' + annofn)
2986 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2987 annofile.write(info_dict['annotations'])
2988 except (KeyError, TypeError):
2989 self.report_warning('There are no annotations to write.')
2990 except OSError:
2991 self.report_error('Cannot write annotations file: ' + annofn)
2992 return
2993
2994 # Write internet shortcut files
2995 def _write_link_file(link_type):
2996 url = try_get(info_dict['webpage_url'], iri_to_uri)
2997 if not url:
2998 self.report_warning(
2999 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3000 return True
3001 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3002 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3003 return False
3004 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3005 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3006 return True
3007 try:
3008 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3009 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3010 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3011 template_vars = {'url': url}
3012 if link_type == 'desktop':
3013 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3014 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3015 except OSError:
3016 self.report_error(f'Cannot write internet shortcut {linkfn}')
3017 return False
3018 return True
3019
3020 write_links = {
3021 'url': self.params.get('writeurllink'),
3022 'webloc': self.params.get('writewebloclink'),
3023 'desktop': self.params.get('writedesktoplink'),
3024 }
3025 if self.params.get('writelink'):
3026 link_type = ('webloc' if sys.platform == 'darwin'
3027 else 'desktop' if sys.platform.startswith('linux')
3028 else 'url')
3029 write_links[link_type] = True
3030
3031 if any(should_write and not _write_link_file(link_type)
3032 for link_type, should_write in write_links.items()):
3033 return
3034
3035 def replace_info_dict(new_info):
3036 nonlocal info_dict
3037 if new_info == info_dict:
3038 return
3039 info_dict.clear()
3040 info_dict.update(new_info)
3041
3042 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3043 replace_info_dict(new_info)
3044
3045 if self.params.get('skip_download'):
3046 info_dict['filepath'] = temp_filename
3047 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3048 info_dict['__files_to_move'] = files_to_move
3049 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3050 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3051 else:
3052 # Download
3053 info_dict.setdefault('__postprocessors', [])
3054 try:
3055
3056 def existing_video_file(*filepaths):
3057 ext = info_dict.get('ext')
3058 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3059 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3060 default_overwrite=False)
3061 if file:
3062 info_dict['ext'] = os.path.splitext(file)[1][1:]
3063 return file
3064
3065 fd, success = None, True
3066 if info_dict.get('protocol') or info_dict.get('url'):
3067 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3068 if fd is not FFmpegFD and (
3069 info_dict.get('section_start') or info_dict.get('section_end')):
3070 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3071 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3072 self.report_error(f'{msg}. Aborting')
3073 return
3074
3075 if info_dict.get('requested_formats') is not None:
3076 requested_formats = info_dict['requested_formats']
3077 old_ext = info_dict['ext']
3078 if self.params.get('merge_output_format') is None:
3079 if (info_dict['ext'] == 'webm'
3080 and info_dict.get('thumbnails')
3081 # check with type instead of pp_key, __name__, or isinstance
3082 # since we dont want any custom PPs to trigger this
3083 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3084 info_dict['ext'] = 'mkv'
3085 self.report_warning(
3086 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3087 new_ext = info_dict['ext']
3088
3089 def correct_ext(filename, ext=new_ext):
3090 if filename == '-':
3091 return filename
3092 filename_real_ext = os.path.splitext(filename)[1][1:]
3093 filename_wo_ext = (
3094 os.path.splitext(filename)[0]
3095 if filename_real_ext in (old_ext, new_ext)
3096 else filename)
3097 return f'{filename_wo_ext}.{ext}'
3098
3099 # Ensure filename always has a correct extension for successful merge
3100 full_filename = correct_ext(full_filename)
3101 temp_filename = correct_ext(temp_filename)
3102 dl_filename = existing_video_file(full_filename, temp_filename)
3103 info_dict['__real_download'] = False
3104
3105 merger = FFmpegMergerPP(self)
3106 downloaded = []
3107 if dl_filename is not None:
3108 self.report_file_already_downloaded(dl_filename)
3109 elif fd:
3110 for f in requested_formats if fd != FFmpegFD else []:
3111 f['filepath'] = fname = prepend_extension(
3112 correct_ext(temp_filename, info_dict['ext']),
3113 'f%s' % f['format_id'], info_dict['ext'])
3114 downloaded.append(fname)
3115 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3116 success, real_download = self.dl(temp_filename, info_dict)
3117 info_dict['__real_download'] = real_download
3118 else:
3119 if self.params.get('allow_unplayable_formats'):
3120 self.report_warning(
3121 'You have requested merging of multiple formats '
3122 'while also allowing unplayable formats to be downloaded. '
3123 'The formats won\'t be merged to prevent data corruption.')
3124 elif not merger.available:
3125 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3126 if not self.params.get('ignoreerrors'):
3127 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3128 return
3129 self.report_warning(f'{msg}. The formats won\'t be merged')
3130
3131 if temp_filename == '-':
3132 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3133 else 'but the formats are incompatible for simultaneous download' if merger.available
3134 else 'but ffmpeg is not installed')
3135 self.report_warning(
3136 f'You have requested downloading multiple formats to stdout {reason}. '
3137 'The formats will be streamed one after the other')
3138 fname = temp_filename
3139 for f in requested_formats:
3140 new_info = dict(info_dict)
3141 del new_info['requested_formats']
3142 new_info.update(f)
3143 if temp_filename != '-':
3144 fname = prepend_extension(
3145 correct_ext(temp_filename, new_info['ext']),
3146 'f%s' % f['format_id'], new_info['ext'])
3147 if not self._ensure_dir_exists(fname):
3148 return
3149 f['filepath'] = fname
3150 downloaded.append(fname)
3151 partial_success, real_download = self.dl(fname, new_info)
3152 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3153 success = success and partial_success
3154
3155 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3156 info_dict['__postprocessors'].append(merger)
3157 info_dict['__files_to_merge'] = downloaded
3158 # Even if there were no downloads, it is being merged only now
3159 info_dict['__real_download'] = True
3160 else:
3161 for file in downloaded:
3162 files_to_move[file] = None
3163 else:
3164 # Just a single file
3165 dl_filename = existing_video_file(full_filename, temp_filename)
3166 if dl_filename is None or dl_filename == temp_filename:
3167 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3168 # So we should try to resume the download
3169 success, real_download = self.dl(temp_filename, info_dict)
3170 info_dict['__real_download'] = real_download
3171 else:
3172 self.report_file_already_downloaded(dl_filename)
3173
3174 dl_filename = dl_filename or temp_filename
3175 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3176
3177 except network_exceptions as err:
3178 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3179 return
3180 except OSError as err:
3181 raise UnavailableVideoError(err)
3182 except (ContentTooShortError, ) as err:
3183 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3184 return
3185
3186 self._raise_pending_errors(info_dict)
3187 if success and full_filename != '-':
3188
3189 def fixup():
3190 do_fixup = True
3191 fixup_policy = self.params.get('fixup')
3192 vid = info_dict['id']
3193
3194 if fixup_policy in ('ignore', 'never'):
3195 return
3196 elif fixup_policy == 'warn':
3197 do_fixup = 'warn'
3198 elif fixup_policy != 'force':
3199 assert fixup_policy in ('detect_or_warn', None)
3200 if not info_dict.get('__real_download'):
3201 do_fixup = False
3202
3203 def ffmpeg_fixup(cndn, msg, cls):
3204 if not (do_fixup and cndn):
3205 return
3206 elif do_fixup == 'warn':
3207 self.report_warning(f'{vid}: {msg}')
3208 return
3209 pp = cls(self)
3210 if pp.available:
3211 info_dict['__postprocessors'].append(pp)
3212 else:
3213 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3214
3215 stretched_ratio = info_dict.get('stretched_ratio')
3216 ffmpeg_fixup(stretched_ratio not in (1, None),
3217 f'Non-uniform pixel ratio {stretched_ratio}',
3218 FFmpegFixupStretchedPP)
3219
3220 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3221 downloader = downloader.FD_NAME if downloader else None
3222
3223 ext = info_dict.get('ext')
3224 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3225 isinstance(pp, FFmpegVideoConvertorPP)
3226 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3227 ) for pp in self._pps['post_process'])
3228
3229 if not postprocessed_by_ffmpeg:
3230 ffmpeg_fixup(ext == 'm4a' and info_dict.get('container') == 'm4a_dash',
3231 'writing DASH m4a. Only some players support this container',
3232 FFmpegFixupM4aPP)
3233 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3234 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3235 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3236 FFmpegFixupM3u8PP)
3237 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3238 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3239
3240 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3241 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3242
3243 fixup()
3244 try:
3245 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3246 except PostProcessingError as err:
3247 self.report_error('Postprocessing: %s' % str(err))
3248 return
3249 try:
3250 for ph in self._post_hooks:
3251 ph(info_dict['filepath'])
3252 except Exception as err:
3253 self.report_error('post hooks: %s' % str(err))
3254 return
3255 info_dict['__write_download_archive'] = True
3256
3257 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3258 if self.params.get('force_write_download_archive'):
3259 info_dict['__write_download_archive'] = True
3260 check_max_downloads()
3261
3262 def __download_wrapper(self, func):
3263 @functools.wraps(func)
3264 def wrapper(*args, **kwargs):
3265 try:
3266 res = func(*args, **kwargs)
3267 except UnavailableVideoError as e:
3268 self.report_error(e)
3269 except DownloadCancelled as e:
3270 self.to_screen(f'[info] {e}')
3271 if not self.params.get('break_per_url'):
3272 raise
3273 else:
3274 if self.params.get('dump_single_json', False):
3275 self.post_extract(res)
3276 self.to_stdout(json.dumps(self.sanitize_info(res)))
3277 return wrapper
3278
3279 def download(self, url_list):
3280 """Download a given list of URLs."""
3281 url_list = variadic(url_list) # Passing a single URL is a common mistake
3282 outtmpl = self.params['outtmpl']['default']
3283 if (len(url_list) > 1
3284 and outtmpl != '-'
3285 and '%' not in outtmpl
3286 and self.params.get('max_downloads') != 1):
3287 raise SameFileError(outtmpl)
3288
3289 for url in url_list:
3290 self.__download_wrapper(self.extract_info)(
3291 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3292
3293 return self._download_retcode
3294
3295 def download_with_info_file(self, info_filename):
3296 with contextlib.closing(fileinput.FileInput(
3297 [info_filename], mode='r',
3298 openhook=fileinput.hook_encoded('utf-8'))) as f:
3299 # FileInput doesn't have a read method, we can't call json.load
3300 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3301 try:
3302 self.__download_wrapper(self.process_ie_result)(info, download=True)
3303 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3304 if not isinstance(e, EntryNotInPlaylist):
3305 self.to_stderr('\r')
3306 webpage_url = info.get('webpage_url')
3307 if webpage_url is not None:
3308 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3309 return self.download([webpage_url])
3310 else:
3311 raise
3312 return self._download_retcode
3313
3314 @staticmethod
3315 def sanitize_info(info_dict, remove_private_keys=False):
3316 ''' Sanitize the infodict for converting to json '''
3317 if info_dict is None:
3318 return info_dict
3319 info_dict.setdefault('epoch', int(time.time()))
3320 info_dict.setdefault('_type', 'video')
3321
3322 if remove_private_keys:
3323 reject = lambda k, v: v is None or k.startswith('__') or k in {
3324 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3325 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3326 }
3327 else:
3328 reject = lambda k, v: False
3329
3330 def filter_fn(obj):
3331 if isinstance(obj, dict):
3332 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3333 elif isinstance(obj, (list, tuple, set, LazyList)):
3334 return list(map(filter_fn, obj))
3335 elif obj is None or isinstance(obj, (str, int, float, bool)):
3336 return obj
3337 else:
3338 return repr(obj)
3339
3340 return filter_fn(info_dict)
3341
3342 @staticmethod
3343 def filter_requested_info(info_dict, actually_filter=True):
3344 ''' Alias of sanitize_info for backward compatibility '''
3345 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3346
3347 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3348 for filename in set(filter(None, files_to_delete)):
3349 if msg:
3350 self.to_screen(msg % filename)
3351 try:
3352 os.remove(filename)
3353 except OSError:
3354 self.report_warning(f'Unable to delete file {filename}')
3355 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3356 del info['__files_to_move'][filename]
3357
3358 @staticmethod
3359 def post_extract(info_dict):
3360 def actual_post_extract(info_dict):
3361 if info_dict.get('_type') in ('playlist', 'multi_video'):
3362 for video_dict in info_dict.get('entries', {}):
3363 actual_post_extract(video_dict or {})
3364 return
3365
3366 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3367 info_dict.update(post_extractor())
3368
3369 actual_post_extract(info_dict or {})
3370
3371 def run_pp(self, pp, infodict):
3372 files_to_delete = []
3373 if '__files_to_move' not in infodict:
3374 infodict['__files_to_move'] = {}
3375 try:
3376 files_to_delete, infodict = pp.run(infodict)
3377 except PostProcessingError as e:
3378 # Must be True and not 'only_download'
3379 if self.params.get('ignoreerrors') is True:
3380 self.report_error(e)
3381 return infodict
3382 raise
3383
3384 if not files_to_delete:
3385 return infodict
3386 if self.params.get('keepvideo', False):
3387 for f in files_to_delete:
3388 infodict['__files_to_move'].setdefault(f, '')
3389 else:
3390 self._delete_downloaded_files(
3391 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3392 return infodict
3393
3394 def run_all_pps(self, key, info, *, additional_pps=None):
3395 self._forceprint(key, info)
3396 for pp in (additional_pps or []) + self._pps[key]:
3397 info = self.run_pp(pp, info)
3398 return info
3399
3400 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3401 info = dict(ie_info)
3402 info['__files_to_move'] = files_to_move or {}
3403 try:
3404 info = self.run_all_pps(key, info)
3405 except PostProcessingError as err:
3406 msg = f'Preprocessing: {err}'
3407 info.setdefault('__pending_error', msg)
3408 self.report_error(msg, is_error=False)
3409 return info, info.pop('__files_to_move', None)
3410
3411 def post_process(self, filename, info, files_to_move=None):
3412 """Run all the postprocessors on the given file."""
3413 info['filepath'] = filename
3414 info['__files_to_move'] = files_to_move or {}
3415 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3416 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3417 del info['__files_to_move']
3418 return self.run_all_pps('after_move', info)
3419
3420 def _make_archive_id(self, info_dict):
3421 video_id = info_dict.get('id')
3422 if not video_id:
3423 return
3424 # Future-proof against any change in case
3425 # and backwards compatibility with prior versions
3426 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3427 if extractor is None:
3428 url = str_or_none(info_dict.get('url'))
3429 if not url:
3430 return
3431 # Try to find matching extractor for the URL and take its ie_key
3432 for ie_key, ie in self._ies.items():
3433 if ie.suitable(url):
3434 extractor = ie_key
3435 break
3436 else:
3437 return
3438 return make_archive_id(extractor, video_id)
3439
3440 def in_download_archive(self, info_dict):
3441 fn = self.params.get('download_archive')
3442 if fn is None:
3443 return False
3444
3445 vid_ids = [self._make_archive_id(info_dict)]
3446 vid_ids.extend(info_dict.get('_old_archive_ids', []))
3447 return any(id_ in self.archive for id_ in vid_ids)
3448
3449 def record_download_archive(self, info_dict):
3450 fn = self.params.get('download_archive')
3451 if fn is None:
3452 return
3453 vid_id = self._make_archive_id(info_dict)
3454 assert vid_id
3455 self.write_debug(f'Adding to archive: {vid_id}')
3456 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3457 archive_file.write(vid_id + '\n')
3458 self.archive.add(vid_id)
3459
3460 @staticmethod
3461 def format_resolution(format, default='unknown'):
3462 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3463 return 'audio only'
3464 if format.get('resolution') is not None:
3465 return format['resolution']
3466 if format.get('width') and format.get('height'):
3467 return '%dx%d' % (format['width'], format['height'])
3468 elif format.get('height'):
3469 return '%sp' % format['height']
3470 elif format.get('width'):
3471 return '%dx?' % format['width']
3472 return default
3473
3474 def _list_format_headers(self, *headers):
3475 if self.params.get('listformats_table', True) is not False:
3476 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3477 return headers
3478
3479 def _format_note(self, fdict):
3480 res = ''
3481 if fdict.get('ext') in ['f4f', 'f4m']:
3482 res += '(unsupported)'
3483 if fdict.get('language'):
3484 if res:
3485 res += ' '
3486 res += '[%s]' % fdict['language']
3487 if fdict.get('format_note') is not None:
3488 if res:
3489 res += ' '
3490 res += fdict['format_note']
3491 if fdict.get('tbr') is not None:
3492 if res:
3493 res += ', '
3494 res += '%4dk' % fdict['tbr']
3495 if fdict.get('container') is not None:
3496 if res:
3497 res += ', '
3498 res += '%s container' % fdict['container']
3499 if (fdict.get('vcodec') is not None
3500 and fdict.get('vcodec') != 'none'):
3501 if res:
3502 res += ', '
3503 res += fdict['vcodec']
3504 if fdict.get('vbr') is not None:
3505 res += '@'
3506 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3507 res += 'video@'
3508 if fdict.get('vbr') is not None:
3509 res += '%4dk' % fdict['vbr']
3510 if fdict.get('fps') is not None:
3511 if res:
3512 res += ', '
3513 res += '%sfps' % fdict['fps']
3514 if fdict.get('acodec') is not None:
3515 if res:
3516 res += ', '
3517 if fdict['acodec'] == 'none':
3518 res += 'video only'
3519 else:
3520 res += '%-5s' % fdict['acodec']
3521 elif fdict.get('abr') is not None:
3522 if res:
3523 res += ', '
3524 res += 'audio'
3525 if fdict.get('abr') is not None:
3526 res += '@%3dk' % fdict['abr']
3527 if fdict.get('asr') is not None:
3528 res += ' (%5dHz)' % fdict['asr']
3529 if fdict.get('filesize') is not None:
3530 if res:
3531 res += ', '
3532 res += format_bytes(fdict['filesize'])
3533 elif fdict.get('filesize_approx') is not None:
3534 if res:
3535 res += ', '
3536 res += '~' + format_bytes(fdict['filesize_approx'])
3537 return res
3538
3539 def render_formats_table(self, info_dict):
3540 if not info_dict.get('formats') and not info_dict.get('url'):
3541 return None
3542
3543 formats = info_dict.get('formats', [info_dict])
3544 if not self.params.get('listformats_table', True) is not False:
3545 table = [
3546 [
3547 format_field(f, 'format_id'),
3548 format_field(f, 'ext'),
3549 self.format_resolution(f),
3550 self._format_note(f)
3551 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3552 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3553
3554 def simplified_codec(f, field):
3555 assert field in ('acodec', 'vcodec')
3556 codec = f.get(field, 'unknown')
3557 if not codec:
3558 return 'unknown'
3559 elif codec != 'none':
3560 return '.'.join(codec.split('.')[:4])
3561
3562 if field == 'vcodec' and f.get('acodec') == 'none':
3563 return 'images'
3564 elif field == 'acodec' and f.get('vcodec') == 'none':
3565 return ''
3566 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3567 self.Styles.SUPPRESS)
3568
3569 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3570 table = [
3571 [
3572 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3573 format_field(f, 'ext'),
3574 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3575 format_field(f, 'fps', '\t%d', func=round),
3576 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3577 format_field(f, 'audio_channels', '\t%s'),
3578 delim,
3579 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3580 format_field(f, 'tbr', '\t%dk', func=round),
3581 shorten_protocol_name(f.get('protocol', '')),
3582 delim,
3583 simplified_codec(f, 'vcodec'),
3584 format_field(f, 'vbr', '\t%dk', func=round),
3585 simplified_codec(f, 'acodec'),
3586 format_field(f, 'abr', '\t%dk', func=round),
3587 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3588 join_nonempty(
3589 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3590 format_field(f, 'language', '[%s]'),
3591 join_nonempty(format_field(f, 'format_note'),
3592 format_field(f, 'container', ignore=(None, f.get('ext'))),
3593 delim=', '),
3594 delim=' '),
3595 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3596 header_line = self._list_format_headers(
3597 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3598 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3599
3600 return render_table(
3601 header_line, table, hide_empty=True,
3602 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3603
3604 def render_thumbnails_table(self, info_dict):
3605 thumbnails = list(info_dict.get('thumbnails') or [])
3606 if not thumbnails:
3607 return None
3608 return render_table(
3609 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3610 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3611
3612 def render_subtitles_table(self, video_id, subtitles):
3613 def _row(lang, formats):
3614 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3615 if len(set(names)) == 1:
3616 names = [] if names[0] == 'unknown' else names[:1]
3617 return [lang, ', '.join(names), ', '.join(exts)]
3618
3619 if not subtitles:
3620 return None
3621 return render_table(
3622 self._list_format_headers('Language', 'Name', 'Formats'),
3623 [_row(lang, formats) for lang, formats in subtitles.items()],
3624 hide_empty=True)
3625
3626 def __list_table(self, video_id, name, func, *args):
3627 table = func(*args)
3628 if not table:
3629 self.to_screen(f'{video_id} has no {name}')
3630 return
3631 self.to_screen(f'[info] Available {name} for {video_id}:')
3632 self.to_stdout(table)
3633
3634 def list_formats(self, info_dict):
3635 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3636
3637 def list_thumbnails(self, info_dict):
3638 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3639
3640 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3641 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3642
3643 def urlopen(self, req):
3644 """ Start an HTTP download """
3645 if isinstance(req, str):
3646 req = sanitized_Request(req)
3647 return self._opener.open(req, timeout=self._socket_timeout)
3648
3649 def print_debug_header(self):
3650 if not self.params.get('verbose'):
3651 return
3652
3653 # These imports can be slow. So import them only as needed
3654 from .extractor.extractors import _LAZY_LOADER
3655 from .extractor.extractors import _PLUGIN_CLASSES as plugin_extractors
3656
3657 def get_encoding(stream):
3658 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3659 if not supports_terminal_sequences(stream):
3660 from .utils import WINDOWS_VT_MODE # Must be imported locally
3661 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3662 return ret
3663
3664 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3665 locale.getpreferredencoding(),
3666 sys.getfilesystemencoding(),
3667 self.get_encoding(),
3668 ', '.join(
3669 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3670 if stream is not None and key != 'console')
3671 )
3672
3673 logger = self.params.get('logger')
3674 if logger:
3675 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3676 write_debug(encoding_str)
3677 else:
3678 write_string(f'[debug] {encoding_str}\n', encoding=None)
3679 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3680
3681 source = detect_variant()
3682 if VARIANT not in (None, 'pip'):
3683 source += '*'
3684 write_debug(join_nonempty(
3685 'yt-dlp version', __version__,
3686 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3687 '' if source == 'unknown' else f'({source})',
3688 delim=' '))
3689 if not _LAZY_LOADER:
3690 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3691 write_debug('Lazy loading extractors is forcibly disabled')
3692 else:
3693 write_debug('Lazy loading extractors is disabled')
3694 if plugin_extractors or plugin_postprocessors:
3695 write_debug('Plugins: %s' % [
3696 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3697 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3698 if self.params['compat_opts']:
3699 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3700
3701 if source == 'source':
3702 try:
3703 stdout, _, _ = Popen.run(
3704 ['git', 'rev-parse', '--short', 'HEAD'],
3705 text=True, cwd=os.path.dirname(os.path.abspath(__file__)),
3706 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
3707 if re.fullmatch('[0-9a-f]+', stdout.strip()):
3708 write_debug(f'Git HEAD: {stdout.strip()}')
3709 except Exception:
3710 with contextlib.suppress(Exception):
3711 sys.exc_clear()
3712
3713 write_debug(system_identifier())
3714
3715 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3716 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3717 if ffmpeg_features:
3718 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3719
3720 exe_versions['rtmpdump'] = rtmpdump_version()
3721 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3722 exe_str = ', '.join(
3723 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3724 ) or 'none'
3725 write_debug('exe versions: %s' % exe_str)
3726
3727 from .compat.compat_utils import get_package_info
3728 from .dependencies import available_dependencies
3729
3730 write_debug('Optional libraries: %s' % (', '.join(sorted({
3731 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3732 })) or 'none'))
3733
3734 self._setup_opener()
3735 proxy_map = {}
3736 for handler in self._opener.handlers:
3737 if hasattr(handler, 'proxies'):
3738 proxy_map.update(handler.proxies)
3739 write_debug(f'Proxy map: {proxy_map}')
3740
3741 # Not implemented
3742 if False and self.params.get('call_home'):
3743 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3744 write_debug('Public IP address: %s' % ipaddr)
3745 latest_version = self.urlopen(
3746 'https://yt-dl.org/latest/version').read().decode()
3747 if version_tuple(latest_version) > version_tuple(__version__):
3748 self.report_warning(
3749 'You are using an outdated version (newest version: %s)! '
3750 'See https://yt-dl.org/update if you need help updating.' %
3751 latest_version)
3752
3753 def _setup_opener(self):
3754 if hasattr(self, '_opener'):
3755 return
3756 timeout_val = self.params.get('socket_timeout')
3757 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3758
3759 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3760 opts_cookiefile = self.params.get('cookiefile')
3761 opts_proxy = self.params.get('proxy')
3762
3763 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3764
3765 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3766 if opts_proxy is not None:
3767 if opts_proxy == '':
3768 proxies = {}
3769 else:
3770 proxies = {'http': opts_proxy, 'https': opts_proxy}
3771 else:
3772 proxies = urllib.request.getproxies()
3773 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3774 if 'http' in proxies and 'https' not in proxies:
3775 proxies['https'] = proxies['http']
3776 proxy_handler = PerRequestProxyHandler(proxies)
3777
3778 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3779 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3780 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3781 redirect_handler = YoutubeDLRedirectHandler()
3782 data_handler = urllib.request.DataHandler()
3783
3784 # When passing our own FileHandler instance, build_opener won't add the
3785 # default FileHandler and allows us to disable the file protocol, which
3786 # can be used for malicious purposes (see
3787 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3788 file_handler = urllib.request.FileHandler()
3789
3790 def file_open(*args, **kwargs):
3791 raise urllib.error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3792 file_handler.file_open = file_open
3793
3794 opener = urllib.request.build_opener(
3795 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3796
3797 # Delete the default user-agent header, which would otherwise apply in
3798 # cases where our custom HTTP handler doesn't come into play
3799 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3800 opener.addheaders = []
3801 self._opener = opener
3802
3803 def encode(self, s):
3804 if isinstance(s, bytes):
3805 return s # Already encoded
3806
3807 try:
3808 return s.encode(self.get_encoding())
3809 except UnicodeEncodeError as err:
3810 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3811 raise
3812
3813 def get_encoding(self):
3814 encoding = self.params.get('encoding')
3815 if encoding is None:
3816 encoding = preferredencoding()
3817 return encoding
3818
3819 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3820 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3821 if overwrite is None:
3822 overwrite = self.params.get('overwrites', True)
3823 if not self.params.get('writeinfojson'):
3824 return False
3825 elif not infofn:
3826 self.write_debug(f'Skipping writing {label} infojson')
3827 return False
3828 elif not self._ensure_dir_exists(infofn):
3829 return None
3830 elif not overwrite and os.path.exists(infofn):
3831 self.to_screen(f'[info] {label.title()} metadata is already present')
3832 return 'exists'
3833
3834 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3835 try:
3836 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3837 return True
3838 except OSError:
3839 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3840 return None
3841
3842 def _write_description(self, label, ie_result, descfn):
3843 ''' Write description and returns True = written, False = skip, None = error '''
3844 if not self.params.get('writedescription'):
3845 return False
3846 elif not descfn:
3847 self.write_debug(f'Skipping writing {label} description')
3848 return False
3849 elif not self._ensure_dir_exists(descfn):
3850 return None
3851 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3852 self.to_screen(f'[info] {label.title()} description is already present')
3853 elif ie_result.get('description') is None:
3854 self.report_warning(f'There\'s no {label} description to write')
3855 return False
3856 else:
3857 try:
3858 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3859 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3860 descfile.write(ie_result['description'])
3861 except OSError:
3862 self.report_error(f'Cannot write {label} description file {descfn}')
3863 return None
3864 return True
3865
3866 def _write_subtitles(self, info_dict, filename):
3867 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3868 ret = []
3869 subtitles = info_dict.get('requested_subtitles')
3870 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3871 # subtitles download errors are already managed as troubles in relevant IE
3872 # that way it will silently go on when used with unsupporting IE
3873 return ret
3874
3875 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3876 if not sub_filename_base:
3877 self.to_screen('[info] Skipping writing video subtitles')
3878 return ret
3879 for sub_lang, sub_info in subtitles.items():
3880 sub_format = sub_info['ext']
3881 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3882 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3883 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3884 if existing_sub:
3885 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3886 sub_info['filepath'] = existing_sub
3887 ret.append((existing_sub, sub_filename_final))
3888 continue
3889
3890 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3891 if sub_info.get('data') is not None:
3892 try:
3893 # Use newline='' to prevent conversion of newline characters
3894 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3895 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3896 subfile.write(sub_info['data'])
3897 sub_info['filepath'] = sub_filename
3898 ret.append((sub_filename, sub_filename_final))
3899 continue
3900 except OSError:
3901 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3902 return None
3903
3904 try:
3905 sub_copy = sub_info.copy()
3906 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3907 self.dl(sub_filename, sub_copy, subtitle=True)
3908 sub_info['filepath'] = sub_filename
3909 ret.append((sub_filename, sub_filename_final))
3910 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3911 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3912 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3913 if not self.params.get('ignoreerrors'):
3914 self.report_error(msg)
3915 raise DownloadError(msg)
3916 self.report_warning(msg)
3917 return ret
3918
3919 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3920 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3921 write_all = self.params.get('write_all_thumbnails', False)
3922 thumbnails, ret = [], []
3923 if write_all or self.params.get('writethumbnail', False):
3924 thumbnails = info_dict.get('thumbnails') or []
3925 multiple = write_all and len(thumbnails) > 1
3926
3927 if thumb_filename_base is None:
3928 thumb_filename_base = filename
3929 if thumbnails and not thumb_filename_base:
3930 self.write_debug(f'Skipping writing {label} thumbnail')
3931 return ret
3932
3933 for idx, t in list(enumerate(thumbnails))[::-1]:
3934 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3935 thumb_display_id = f'{label} thumbnail {t["id"]}'
3936 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3937 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3938
3939 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3940 if existing_thumb:
3941 self.to_screen('[info] %s is already present' % (
3942 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3943 t['filepath'] = existing_thumb
3944 ret.append((existing_thumb, thumb_filename_final))
3945 else:
3946 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3947 try:
3948 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3949 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3950 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3951 shutil.copyfileobj(uf, thumbf)
3952 ret.append((thumb_filename, thumb_filename_final))
3953 t['filepath'] = thumb_filename
3954 except network_exceptions as err:
3955 thumbnails.pop(idx)
3956 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3957 if ret and not write_all:
3958 break
3959 return ret