]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[compat] Fix `compat.WINDOWS_VT_MODE`
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 #!/usr/bin/env python3
2 import collections
3 import contextlib
4 import datetime
5 import errno
6 import fileinput
7 import functools
8 import io
9 import itertools
10 import json
11 import locale
12 import operator
13 import os
14 import platform
15 import random
16 import re
17 import shutil
18 import subprocess
19 import sys
20 import tempfile
21 import time
22 import tokenize
23 import traceback
24 import unicodedata
25 import urllib.request
26 from string import ascii_letters
27
28 from .cache import Cache
29 from .compat import (
30 HAS_LEGACY as compat_has_legacy,
31 compat_get_terminal_size,
32 compat_os_name,
33 compat_shlex_quote,
34 compat_str,
35 compat_urllib_error,
36 compat_urllib_request,
37 )
38 from .cookies import load_cookies
39 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
40 from .downloader.rtmp import rtmpdump_version
41 from .extractor import gen_extractor_classes, get_info_extractor
42 from .extractor.openload import PhantomJSwrapper
43 from .minicurses import format_text
44 from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
45 from .postprocessor import (
46 EmbedThumbnailPP,
47 FFmpegFixupDuplicateMoovPP,
48 FFmpegFixupDurationPP,
49 FFmpegFixupM3u8PP,
50 FFmpegFixupM4aPP,
51 FFmpegFixupStretchedPP,
52 FFmpegFixupTimestampPP,
53 FFmpegMergerPP,
54 FFmpegPostProcessor,
55 MoveFilesAfterDownloadPP,
56 get_postprocessor,
57 )
58 from .update import detect_variant
59 from .utils import (
60 DEFAULT_OUTTMPL,
61 IDENTITY,
62 LINK_TEMPLATES,
63 NO_DEFAULT,
64 NUMBER_RE,
65 OUTTMPL_TYPES,
66 POSTPROCESS_WHEN,
67 STR_FORMAT_RE_TMPL,
68 STR_FORMAT_TYPES,
69 ContentTooShortError,
70 DateRange,
71 DownloadCancelled,
72 DownloadError,
73 EntryNotInPlaylist,
74 ExistingVideoReached,
75 ExtractorError,
76 GeoRestrictedError,
77 HEADRequest,
78 ISO3166Utils,
79 LazyList,
80 MaxDownloadsReached,
81 Namespace,
82 PagedList,
83 PerRequestProxyHandler,
84 PlaylistEntries,
85 Popen,
86 PostProcessingError,
87 ReExtractInfo,
88 RejectedVideoReached,
89 SameFileError,
90 UnavailableVideoError,
91 YoutubeDLCookieProcessor,
92 YoutubeDLHandler,
93 YoutubeDLRedirectHandler,
94 age_restricted,
95 args_to_str,
96 date_from_str,
97 determine_ext,
98 determine_protocol,
99 encode_compat_str,
100 encodeFilename,
101 error_to_compat_str,
102 expand_path,
103 filter_dict,
104 float_or_none,
105 format_bytes,
106 format_decimal_suffix,
107 format_field,
108 formatSeconds,
109 get_domain,
110 int_or_none,
111 iri_to_uri,
112 join_nonempty,
113 locked_file,
114 make_dir,
115 make_HTTPS_handler,
116 merge_headers,
117 network_exceptions,
118 number_of_digits,
119 orderedSet,
120 parse_filesize,
121 platform_name,
122 preferredencoding,
123 prepend_extension,
124 register_socks_protocols,
125 remove_terminal_sequences,
126 render_table,
127 replace_extension,
128 sanitize_filename,
129 sanitize_path,
130 sanitize_url,
131 sanitized_Request,
132 std_headers,
133 str_or_none,
134 strftime_or_none,
135 subtitles_filename,
136 supports_terminal_sequences,
137 timetuple_from_msec,
138 to_high_limit_path,
139 traverse_obj,
140 try_get,
141 url_basename,
142 variadic,
143 version_tuple,
144 windows_enable_vt_mode,
145 write_json_file,
146 write_string,
147 )
148 from .version import RELEASE_GIT_HEAD, __version__
149
150 if compat_os_name == 'nt':
151 import ctypes
152
153
154 class YoutubeDL:
155 """YoutubeDL class.
156
157 YoutubeDL objects are the ones responsible of downloading the
158 actual video file and writing it to disk if the user has requested
159 it, among some other tasks. In most cases there should be one per
160 program. As, given a video URL, the downloader doesn't know how to
161 extract all the needed information, task that InfoExtractors do, it
162 has to pass the URL to one of them.
163
164 For this, YoutubeDL objects have a method that allows
165 InfoExtractors to be registered in a given order. When it is passed
166 a URL, the YoutubeDL object handles it to the first InfoExtractor it
167 finds that reports being able to handle it. The InfoExtractor extracts
168 all the information about the video or videos the URL refers to, and
169 YoutubeDL process the extracted information, possibly using a File
170 Downloader to download the video.
171
172 YoutubeDL objects accept a lot of parameters. In order not to saturate
173 the object constructor with arguments, it receives a dictionary of
174 options instead. These options are available through the params
175 attribute for the InfoExtractors to use. The YoutubeDL also
176 registers itself as the downloader in charge for the InfoExtractors
177 that are added to it, so this is a "mutual registration".
178
179 Available options:
180
181 username: Username for authentication purposes.
182 password: Password for authentication purposes.
183 videopassword: Password for accessing a video.
184 ap_mso: Adobe Pass multiple-system operator identifier.
185 ap_username: Multiple-system operator account username.
186 ap_password: Multiple-system operator account password.
187 usenetrc: Use netrc for authentication instead.
188 verbose: Print additional info to stdout.
189 quiet: Do not print messages to stdout.
190 no_warnings: Do not print out anything for warnings.
191 forceprint: A dict with keys WHEN mapped to a list of templates to
192 print to stdout. The allowed keys are video or any of the
193 items in utils.POSTPROCESS_WHEN.
194 For compatibility, a single list is also accepted
195 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
196 a list of tuples with (template, filename)
197 forcejson: Force printing info_dict as JSON.
198 dump_single_json: Force printing the info_dict of the whole playlist
199 (or video) as a single JSON line.
200 force_write_download_archive: Force writing download archive regardless
201 of 'skip_download' or 'simulate'.
202 simulate: Do not download the video files. If unset (or None),
203 simulate only if listsubtitles, listformats or list_thumbnails is used
204 format: Video format code. see "FORMAT SELECTION" for more details.
205 You can also pass a function. The function takes 'ctx' as
206 argument and returns the formats to download.
207 See "build_format_selector" for an implementation
208 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
209 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
210 extracting metadata even if the video is not actually
211 available for download (experimental)
212 format_sort: A list of fields by which to sort the video formats.
213 See "Sorting Formats" for more details.
214 format_sort_force: Force the given format_sort. see "Sorting Formats"
215 for more details.
216 prefer_free_formats: Whether to prefer video formats with free containers
217 over non-free ones of same quality.
218 allow_multiple_video_streams: Allow multiple video streams to be merged
219 into a single file
220 allow_multiple_audio_streams: Allow multiple audio streams to be merged
221 into a single file
222 check_formats Whether to test if the formats are downloadable.
223 Can be True (check all), False (check none),
224 'selected' (check selected formats),
225 or None (check only if requested by extractor)
226 paths: Dictionary of output paths. The allowed keys are 'home'
227 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
228 outtmpl: Dictionary of templates for output names. Allowed keys
229 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
230 For compatibility with youtube-dl, a single string can also be used
231 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
232 restrictfilenames: Do not allow "&" and spaces in file names
233 trim_file_name: Limit length of filename (extension excluded)
234 windowsfilenames: Force the filenames to be windows compatible
235 ignoreerrors: Do not stop on download/postprocessing errors.
236 Can be 'only_download' to ignore only download errors.
237 Default is 'only_download' for CLI, but False for API
238 skip_playlist_after_errors: Number of allowed failures until the rest of
239 the playlist is skipped
240 force_generic_extractor: Force downloader to use the generic extractor
241 overwrites: Overwrite all video and metadata files if True,
242 overwrite only non-video files if None
243 and don't overwrite any file if False
244 For compatibility with youtube-dl,
245 "nooverwrites" may also be used instead
246 playlist_items: Specific indices of playlist to download.
247 playlistrandom: Download playlist items in random order.
248 lazy_playlist: Process playlist entries as they are received.
249 matchtitle: Download only matching titles.
250 rejecttitle: Reject downloads for matching titles.
251 logger: Log messages to a logging.Logger instance.
252 logtostderr: Log messages to stderr instead of stdout.
253 consoletitle: Display progress in console window's titlebar.
254 writedescription: Write the video description to a .description file
255 writeinfojson: Write the video description to a .info.json file
256 clean_infojson: Remove private fields from the infojson
257 getcomments: Extract video comments. This will not be written to disk
258 unless writeinfojson is also given
259 writeannotations: Write the video annotations to a .annotations.xml file
260 writethumbnail: Write the thumbnail image to a file
261 allow_playlist_files: Whether to write playlists' description, infojson etc
262 also to disk when using the 'write*' options
263 write_all_thumbnails: Write all thumbnail formats to files
264 writelink: Write an internet shortcut file, depending on the
265 current platform (.url/.webloc/.desktop)
266 writeurllink: Write a Windows internet shortcut file (.url)
267 writewebloclink: Write a macOS internet shortcut file (.webloc)
268 writedesktoplink: Write a Linux internet shortcut file (.desktop)
269 writesubtitles: Write the video subtitles to a file
270 writeautomaticsub: Write the automatically generated subtitles to a file
271 listsubtitles: Lists all available subtitles for the video
272 subtitlesformat: The format code for subtitles
273 subtitleslangs: List of languages of the subtitles to download (can be regex).
274 The list may contain "all" to refer to all the available
275 subtitles. The language can be prefixed with a "-" to
276 exclude it from the requested languages. Eg: ['all', '-live_chat']
277 keepvideo: Keep the video file after post-processing
278 daterange: A DateRange object, download only if the upload_date is in the range.
279 skip_download: Skip the actual download of the video file
280 cachedir: Location of the cache files in the filesystem.
281 False to disable filesystem cache.
282 noplaylist: Download single video instead of a playlist if in doubt.
283 age_limit: An integer representing the user's age in years.
284 Unsuitable videos for the given age are skipped.
285 min_views: An integer representing the minimum view count the video
286 must have in order to not be skipped.
287 Videos without view count information are always
288 downloaded. None for no limit.
289 max_views: An integer representing the maximum view count.
290 Videos that are more popular than that are not
291 downloaded.
292 Videos without view count information are always
293 downloaded. None for no limit.
294 download_archive: File name of a file where all downloads are recorded.
295 Videos already present in the file are not downloaded
296 again.
297 break_on_existing: Stop the download process after attempting to download a
298 file that is in the archive.
299 break_on_reject: Stop the download process when encountering a video that
300 has been filtered out.
301 break_per_url: Whether break_on_reject and break_on_existing
302 should act on each input URL as opposed to for the entire queue
303 cookiefile: File name or text stream from where cookies should be read and dumped to
304 cookiesfrombrowser: A tuple containing the name of the browser, the profile
305 name/pathfrom where cookies are loaded, and the name of the
306 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
307 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
308 support RFC 5746 secure renegotiation
309 nocheckcertificate: Do not verify SSL certificates
310 client_certificate: Path to client certificate file in PEM format. May include the private key
311 client_certificate_key: Path to private key file for client certificate
312 client_certificate_password: Password for client certificate private key, if encrypted.
313 If not provided and the key is encrypted, yt-dlp will ask interactively
314 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
315 At the moment, this is only supported by YouTube.
316 http_headers: A dictionary of custom headers to be used for all requests
317 proxy: URL of the proxy server to use
318 geo_verification_proxy: URL of the proxy to use for IP address verification
319 on geo-restricted sites.
320 socket_timeout: Time to wait for unresponsive hosts, in seconds
321 bidi_workaround: Work around buggy terminals without bidirectional text
322 support, using fridibi
323 debug_printtraffic:Print out sent and received HTTP traffic
324 default_search: Prepend this string if an input url is not valid.
325 'auto' for elaborate guessing
326 encoding: Use this encoding instead of the system-specified.
327 extract_flat: Do not resolve URLs, return the immediate result.
328 Pass in 'in_playlist' to only show this behavior for
329 playlist items.
330 wait_for_video: If given, wait for scheduled streams to become available.
331 The value should be a tuple containing the range
332 (min_secs, max_secs) to wait between retries
333 postprocessors: A list of dictionaries, each with an entry
334 * key: The name of the postprocessor. See
335 yt_dlp/postprocessor/__init__.py for a list.
336 * when: When to run the postprocessor. Allowed values are
337 the entries of utils.POSTPROCESS_WHEN
338 Assumed to be 'post_process' if not given
339 progress_hooks: A list of functions that get called on download
340 progress, with a dictionary with the entries
341 * status: One of "downloading", "error", or "finished".
342 Check this first and ignore unknown values.
343 * info_dict: The extracted info_dict
344
345 If status is one of "downloading", or "finished", the
346 following properties may also be present:
347 * filename: The final filename (always present)
348 * tmpfilename: The filename we're currently writing to
349 * downloaded_bytes: Bytes on disk
350 * total_bytes: Size of the whole file, None if unknown
351 * total_bytes_estimate: Guess of the eventual file size,
352 None if unavailable.
353 * elapsed: The number of seconds since download started.
354 * eta: The estimated time in seconds, None if unknown
355 * speed: The download speed in bytes/second, None if
356 unknown
357 * fragment_index: The counter of the currently
358 downloaded video fragment.
359 * fragment_count: The number of fragments (= individual
360 files that will be merged)
361
362 Progress hooks are guaranteed to be called at least once
363 (with status "finished") if the download is successful.
364 postprocessor_hooks: A list of functions that get called on postprocessing
365 progress, with a dictionary with the entries
366 * status: One of "started", "processing", or "finished".
367 Check this first and ignore unknown values.
368 * postprocessor: Name of the postprocessor
369 * info_dict: The extracted info_dict
370
371 Progress hooks are guaranteed to be called at least twice
372 (with status "started" and "finished") if the processing is successful.
373 merge_output_format: Extension to use when merging formats.
374 final_ext: Expected final extension; used to detect when the file was
375 already downloaded and converted
376 fixup: Automatically correct known faults of the file.
377 One of:
378 - "never": do nothing
379 - "warn": only emit a warning
380 - "detect_or_warn": check whether we can do anything
381 about it, warn otherwise (default)
382 source_address: Client-side IP address to bind to.
383 sleep_interval_requests: Number of seconds to sleep between requests
384 during extraction
385 sleep_interval: Number of seconds to sleep before each download when
386 used alone or a lower bound of a range for randomized
387 sleep before each download (minimum possible number
388 of seconds to sleep) when used along with
389 max_sleep_interval.
390 max_sleep_interval:Upper bound of a range for randomized sleep before each
391 download (maximum possible number of seconds to sleep).
392 Must only be used along with sleep_interval.
393 Actual sleep time will be a random float from range
394 [sleep_interval; max_sleep_interval].
395 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
396 listformats: Print an overview of available video formats and exit.
397 list_thumbnails: Print a table of all thumbnails and exit.
398 match_filter: A function that gets called for every video with the signature
399 (info_dict, *, incomplete: bool) -> Optional[str]
400 For backward compatibility with youtube-dl, the signature
401 (info_dict) -> Optional[str] is also allowed.
402 - If it returns a message, the video is ignored.
403 - If it returns None, the video is downloaded.
404 - If it returns utils.NO_DEFAULT, the user is interactively
405 asked whether to download the video.
406 match_filter_func in utils.py is one example for this.
407 no_color: Do not emit color codes in output.
408 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
409 HTTP header
410 geo_bypass_country:
411 Two-letter ISO 3166-2 country code that will be used for
412 explicit geographic restriction bypassing via faking
413 X-Forwarded-For HTTP header
414 geo_bypass_ip_block:
415 IP range in CIDR notation that will be used similarly to
416 geo_bypass_country
417 external_downloader: A dictionary of protocol keys and the executable of the
418 external downloader to use for it. The allowed protocols
419 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
420 Set the value to 'native' to use the native downloader
421 compat_opts: Compatibility options. See "Differences in default behavior".
422 The following options do not work when used through the API:
423 filename, abort-on-error, multistreams, no-live-chat, format-sort
424 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
425 Refer __init__.py for their implementation
426 progress_template: Dictionary of templates for progress outputs.
427 Allowed keys are 'download', 'postprocess',
428 'download-title' (console title) and 'postprocess-title'.
429 The template is mapped on a dictionary with keys 'progress' and 'info'
430 retry_sleep_functions: Dictionary of functions that takes the number of attempts
431 as argument and returns the time to sleep in seconds.
432 Allowed keys are 'http', 'fragment', 'file_access'
433 download_ranges: A function that gets called for every video with the signature
434 (info_dict, *, ydl) -> Iterable[Section].
435 Only the returned sections will be downloaded. Each Section contains:
436 * start_time: Start time of the section in seconds
437 * end_time: End time of the section in seconds
438 * title: Section title (Optional)
439 * index: Section number (Optional)
440
441 The following parameters are not used by YoutubeDL itself, they are used by
442 the downloader (see yt_dlp/downloader/common.py):
443 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
444 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
445 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
446 external_downloader_args, concurrent_fragment_downloads.
447
448 The following options are used by the post processors:
449 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
450 to the binary or its containing directory.
451 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
452 and a list of additional command-line arguments for the
453 postprocessor/executable. The dict can also have "PP+EXE" keys
454 which are used when the given exe is used by the given PP.
455 Use 'default' as the name for arguments to passed to all PP
456 For compatibility with youtube-dl, a single list of args
457 can also be used
458
459 The following options are used by the extractors:
460 extractor_retries: Number of times to retry for known errors
461 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
462 hls_split_discontinuity: Split HLS playlists to different formats at
463 discontinuities such as ad breaks (default: False)
464 extractor_args: A dictionary of arguments to be passed to the extractors.
465 See "EXTRACTOR ARGUMENTS" for details.
466 Eg: {'youtube': {'skip': ['dash', 'hls']}}
467 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
468
469 The following options are deprecated and may be removed in the future:
470
471 playliststart: - Use playlist_items
472 Playlist item to start at.
473 playlistend: - Use playlist_items
474 Playlist item to end at.
475 playlistreverse: - Use playlist_items
476 Download playlist items in reverse order.
477 forceurl: - Use forceprint
478 Force printing final URL.
479 forcetitle: - Use forceprint
480 Force printing title.
481 forceid: - Use forceprint
482 Force printing ID.
483 forcethumbnail: - Use forceprint
484 Force printing thumbnail URL.
485 forcedescription: - Use forceprint
486 Force printing description.
487 forcefilename: - Use forceprint
488 Force printing final filename.
489 forceduration: - Use forceprint
490 Force printing duration.
491 allsubtitles: - Use subtitleslangs = ['all']
492 Downloads all the subtitles of the video
493 (requires writesubtitles or writeautomaticsub)
494 include_ads: - Doesn't work
495 Download ads as well
496 call_home: - Not implemented
497 Boolean, true iff we are allowed to contact the
498 yt-dlp servers for debugging.
499 post_hooks: - Register a custom postprocessor
500 A list of functions that get called as the final step
501 for each video file, after all postprocessors have been
502 called. The filename will be passed as the only argument.
503 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
504 Use the native HLS downloader instead of ffmpeg/avconv
505 if True, otherwise use ffmpeg/avconv if False, otherwise
506 use downloader suggested by extractor if None.
507 prefer_ffmpeg: - avconv support is deprecated
508 If False, use avconv instead of ffmpeg if both are available,
509 otherwise prefer ffmpeg.
510 youtube_include_dash_manifest: - Use extractor_args
511 If True (default), DASH manifests and related
512 data will be downloaded and processed by extractor.
513 You can reduce network I/O by disabling it if you don't
514 care about DASH. (only for youtube)
515 youtube_include_hls_manifest: - Use extractor_args
516 If True (default), HLS manifests and related
517 data will be downloaded and processed by extractor.
518 You can reduce network I/O by disabling it if you don't
519 care about HLS. (only for youtube)
520 """
521
522 _NUMERIC_FIELDS = {
523 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
524 'timestamp', 'release_timestamp',
525 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
526 'average_rating', 'comment_count', 'age_limit',
527 'start_time', 'end_time',
528 'chapter_number', 'season_number', 'episode_number',
529 'track_number', 'disc_number', 'release_year',
530 }
531
532 _format_fields = {
533 # NB: Keep in sync with the docstring of extractor/common.py
534 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
535 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
536 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
537 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
538 'preference', 'language', 'language_preference', 'quality', 'source_preference',
539 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
540 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
541 }
542 _format_selection_exts = {
543 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
544 'video': {'mp4', 'flv', 'webm', '3gp'},
545 'storyboards': {'mhtml'},
546 }
547
548 def __init__(self, params=None, auto_init=True):
549 """Create a FileDownloader object with the given options.
550 @param auto_init Whether to load the default extractors and print header (if verbose).
551 Set to 'no_verbose_header' to not print the header
552 """
553 if params is None:
554 params = {}
555 self.params = params
556 self._ies = {}
557 self._ies_instances = {}
558 self._pps = {k: [] for k in POSTPROCESS_WHEN}
559 self._printed_messages = set()
560 self._first_webpage_request = True
561 self._post_hooks = []
562 self._progress_hooks = []
563 self._postprocessor_hooks = []
564 self._download_retcode = 0
565 self._num_downloads = 0
566 self._num_videos = 0
567 self._playlist_level = 0
568 self._playlist_urls = set()
569 self.cache = Cache(self)
570
571 windows_enable_vt_mode()
572 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
573 self._out_files = Namespace(
574 out=stdout,
575 error=sys.stderr,
576 screen=sys.stderr if self.params.get('quiet') else stdout,
577 console=None if compat_os_name == 'nt' else next(
578 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
579 )
580 self._allow_colors = Namespace(**{
581 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
582 for type_, stream in self._out_files.items_ if type_ != 'console'
583 })
584
585 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 6), (3, 7)
586 current_version = sys.version_info[:2]
587 if current_version < MIN_RECOMMENDED:
588 msg = 'Support for Python version %d.%d has been deprecated and will break in future versions of yt-dlp'
589 if current_version < MIN_SUPPORTED:
590 msg = 'Python version %d.%d is no longer supported'
591 self.deprecation_warning(
592 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
593
594 if self.params.get('allow_unplayable_formats'):
595 self.report_warning(
596 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
597 'This is a developer option intended for debugging. \n'
598 ' If you experience any issues while using this option, '
599 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
600
601 def check_deprecated(param, option, suggestion):
602 if self.params.get(param) is not None:
603 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
604 return True
605 return False
606
607 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
608 if self.params.get('geo_verification_proxy') is None:
609 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
610
611 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
612 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
613 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
614
615 for msg in self.params.get('_warnings', []):
616 self.report_warning(msg)
617 for msg in self.params.get('_deprecation_warnings', []):
618 self.deprecation_warning(msg)
619
620 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
621 if not compat_has_legacy:
622 self.params['compat_opts'].add('no-compat-legacy')
623 if 'list-formats' in self.params['compat_opts']:
624 self.params['listformats_table'] = False
625
626 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
627 # nooverwrites was unnecessarily changed to overwrites
628 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
629 # This ensures compatibility with both keys
630 self.params['overwrites'] = not self.params['nooverwrites']
631 elif self.params.get('overwrites') is None:
632 self.params.pop('overwrites', None)
633 else:
634 self.params['nooverwrites'] = not self.params['overwrites']
635
636 self.params.setdefault('forceprint', {})
637 self.params.setdefault('print_to_file', {})
638
639 # Compatibility with older syntax
640 if not isinstance(params['forceprint'], dict):
641 self.params['forceprint'] = {'video': params['forceprint']}
642
643 if self.params.get('bidi_workaround', False):
644 try:
645 import pty
646 master, slave = pty.openpty()
647 width = compat_get_terminal_size().columns
648 width_args = [] if width is None else ['-w', str(width)]
649 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
650 try:
651 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
652 except OSError:
653 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
654 self._output_channel = os.fdopen(master, 'rb')
655 except OSError as ose:
656 if ose.errno == errno.ENOENT:
657 self.report_warning(
658 'Could not find fribidi executable, ignoring --bidi-workaround. '
659 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
660 else:
661 raise
662
663 if auto_init:
664 if auto_init != 'no_verbose_header':
665 self.print_debug_header()
666 self.add_default_info_extractors()
667
668 if (sys.platform != 'win32'
669 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
670 and not self.params.get('restrictfilenames', False)):
671 # Unicode filesystem API will throw errors (#1474, #13027)
672 self.report_warning(
673 'Assuming --restrict-filenames since file system encoding '
674 'cannot encode all characters. '
675 'Set the LC_ALL environment variable to fix this.')
676 self.params['restrictfilenames'] = True
677
678 self._parse_outtmpl()
679
680 # Creating format selector here allows us to catch syntax errors before the extraction
681 self.format_selector = (
682 self.params.get('format') if self.params.get('format') in (None, '-')
683 else self.params['format'] if callable(self.params['format'])
684 else self.build_format_selector(self.params['format']))
685
686 # Set http_headers defaults according to std_headers
687 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
688
689 hooks = {
690 'post_hooks': self.add_post_hook,
691 'progress_hooks': self.add_progress_hook,
692 'postprocessor_hooks': self.add_postprocessor_hook,
693 }
694 for opt, fn in hooks.items():
695 for ph in self.params.get(opt, []):
696 fn(ph)
697
698 for pp_def_raw in self.params.get('postprocessors', []):
699 pp_def = dict(pp_def_raw)
700 when = pp_def.pop('when', 'post_process')
701 self.add_post_processor(
702 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
703 when=when)
704
705 self._setup_opener()
706 register_socks_protocols()
707
708 def preload_download_archive(fn):
709 """Preload the archive, if any is specified"""
710 if fn is None:
711 return False
712 self.write_debug(f'Loading archive file {fn!r}')
713 try:
714 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
715 for line in archive_file:
716 self.archive.add(line.strip())
717 except OSError as ioe:
718 if ioe.errno != errno.ENOENT:
719 raise
720 return False
721 return True
722
723 self.archive = set()
724 preload_download_archive(self.params.get('download_archive'))
725
726 def warn_if_short_id(self, argv):
727 # short YouTube ID starting with dash?
728 idxs = [
729 i for i, a in enumerate(argv)
730 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
731 if idxs:
732 correct_argv = (
733 ['yt-dlp']
734 + [a for i, a in enumerate(argv) if i not in idxs]
735 + ['--'] + [argv[i] for i in idxs]
736 )
737 self.report_warning(
738 'Long argument string detected. '
739 'Use -- to separate parameters and URLs, like this:\n%s' %
740 args_to_str(correct_argv))
741
742 def add_info_extractor(self, ie):
743 """Add an InfoExtractor object to the end of the list."""
744 ie_key = ie.ie_key()
745 self._ies[ie_key] = ie
746 if not isinstance(ie, type):
747 self._ies_instances[ie_key] = ie
748 ie.set_downloader(self)
749
750 def _get_info_extractor_class(self, ie_key):
751 ie = self._ies.get(ie_key)
752 if ie is None:
753 ie = get_info_extractor(ie_key)
754 self.add_info_extractor(ie)
755 return ie
756
757 def get_info_extractor(self, ie_key):
758 """
759 Get an instance of an IE with name ie_key, it will try to get one from
760 the _ies list, if there's no instance it will create a new one and add
761 it to the extractor list.
762 """
763 ie = self._ies_instances.get(ie_key)
764 if ie is None:
765 ie = get_info_extractor(ie_key)()
766 self.add_info_extractor(ie)
767 return ie
768
769 def add_default_info_extractors(self):
770 """
771 Add the InfoExtractors returned by gen_extractors to the end of the list
772 """
773 for ie in gen_extractor_classes():
774 self.add_info_extractor(ie)
775
776 def add_post_processor(self, pp, when='post_process'):
777 """Add a PostProcessor object to the end of the chain."""
778 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
779 self._pps[when].append(pp)
780 pp.set_downloader(self)
781
782 def add_post_hook(self, ph):
783 """Add the post hook"""
784 self._post_hooks.append(ph)
785
786 def add_progress_hook(self, ph):
787 """Add the download progress hook"""
788 self._progress_hooks.append(ph)
789
790 def add_postprocessor_hook(self, ph):
791 """Add the postprocessing progress hook"""
792 self._postprocessor_hooks.append(ph)
793 for pps in self._pps.values():
794 for pp in pps:
795 pp.add_progress_hook(ph)
796
797 def _bidi_workaround(self, message):
798 if not hasattr(self, '_output_channel'):
799 return message
800
801 assert hasattr(self, '_output_process')
802 assert isinstance(message, compat_str)
803 line_count = message.count('\n') + 1
804 self._output_process.stdin.write((message + '\n').encode())
805 self._output_process.stdin.flush()
806 res = ''.join(self._output_channel.readline().decode()
807 for _ in range(line_count))
808 return res[:-len('\n')]
809
810 def _write_string(self, message, out=None, only_once=False):
811 if only_once:
812 if message in self._printed_messages:
813 return
814 self._printed_messages.add(message)
815 write_string(message, out=out, encoding=self.params.get('encoding'))
816
817 def to_stdout(self, message, skip_eol=False, quiet=None):
818 """Print message to stdout"""
819 if quiet is not None:
820 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
821 if skip_eol is not False:
822 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. Use "YoutubeDL.to_screen" instead')
823 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
824
825 def to_screen(self, message, skip_eol=False, quiet=None):
826 """Print message to screen if not in quiet mode"""
827 if self.params.get('logger'):
828 self.params['logger'].debug(message)
829 return
830 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
831 return
832 self._write_string(
833 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
834 self._out_files.screen)
835
836 def to_stderr(self, message, only_once=False):
837 """Print message to stderr"""
838 assert isinstance(message, compat_str)
839 if self.params.get('logger'):
840 self.params['logger'].error(message)
841 else:
842 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
843
844 def _send_console_code(self, code):
845 if compat_os_name == 'nt' or not self._out_files.console:
846 return
847 self._write_string(code, self._out_files.console)
848
849 def to_console_title(self, message):
850 if not self.params.get('consoletitle', False):
851 return
852 message = remove_terminal_sequences(message)
853 if compat_os_name == 'nt':
854 if ctypes.windll.kernel32.GetConsoleWindow():
855 # c_wchar_p() might not be necessary if `message` is
856 # already of type unicode()
857 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
858 else:
859 self._send_console_code(f'\033]0;{message}\007')
860
861 def save_console_title(self):
862 if not self.params.get('consoletitle') or self.params.get('simulate'):
863 return
864 self._send_console_code('\033[22;0t') # Save the title on stack
865
866 def restore_console_title(self):
867 if not self.params.get('consoletitle') or self.params.get('simulate'):
868 return
869 self._send_console_code('\033[23;0t') # Restore the title from stack
870
871 def __enter__(self):
872 self.save_console_title()
873 return self
874
875 def __exit__(self, *args):
876 self.restore_console_title()
877
878 if self.params.get('cookiefile') is not None:
879 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
880
881 def trouble(self, message=None, tb=None, is_error=True):
882 """Determine action to take when a download problem appears.
883
884 Depending on if the downloader has been configured to ignore
885 download errors or not, this method may throw an exception or
886 not when errors are found, after printing the message.
887
888 @param tb If given, is additional traceback information
889 @param is_error Whether to raise error according to ignorerrors
890 """
891 if message is not None:
892 self.to_stderr(message)
893 if self.params.get('verbose'):
894 if tb is None:
895 if sys.exc_info()[0]: # if .trouble has been called from an except block
896 tb = ''
897 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
898 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
899 tb += encode_compat_str(traceback.format_exc())
900 else:
901 tb_data = traceback.format_list(traceback.extract_stack())
902 tb = ''.join(tb_data)
903 if tb:
904 self.to_stderr(tb)
905 if not is_error:
906 return
907 if not self.params.get('ignoreerrors'):
908 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
909 exc_info = sys.exc_info()[1].exc_info
910 else:
911 exc_info = sys.exc_info()
912 raise DownloadError(message, exc_info)
913 self._download_retcode = 1
914
915 Styles = Namespace(
916 HEADERS='yellow',
917 EMPHASIS='light blue',
918 FILENAME='green',
919 ID='green',
920 DELIM='blue',
921 ERROR='red',
922 WARNING='yellow',
923 SUPPRESS='light black',
924 )
925
926 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
927 text = str(text)
928 if test_encoding:
929 original_text = text
930 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
931 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
932 text = text.encode(encoding, 'ignore').decode(encoding)
933 if fallback is not None and text != original_text:
934 text = fallback
935 return format_text(text, f) if allow_colors else text if fallback is None else fallback
936
937 def _format_out(self, *args, **kwargs):
938 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
939
940 def _format_screen(self, *args, **kwargs):
941 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
942
943 def _format_err(self, *args, **kwargs):
944 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
945
946 def report_warning(self, message, only_once=False):
947 '''
948 Print the message to stderr, it will be prefixed with 'WARNING:'
949 If stderr is a tty file the 'WARNING:' will be colored
950 '''
951 if self.params.get('logger') is not None:
952 self.params['logger'].warning(message)
953 else:
954 if self.params.get('no_warnings'):
955 return
956 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
957
958 def deprecation_warning(self, message):
959 if self.params.get('logger') is not None:
960 self.params['logger'].warning(f'DeprecationWarning: {message}')
961 else:
962 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
963
964 def report_error(self, message, *args, **kwargs):
965 '''
966 Do the same as trouble, but prefixes the message with 'ERROR:', colored
967 in red if stderr is a tty file.
968 '''
969 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
970
971 def write_debug(self, message, only_once=False):
972 '''Log debug message or Print message to stderr'''
973 if not self.params.get('verbose', False):
974 return
975 message = f'[debug] {message}'
976 if self.params.get('logger'):
977 self.params['logger'].debug(message)
978 else:
979 self.to_stderr(message, only_once)
980
981 def report_file_already_downloaded(self, file_name):
982 """Report file has already been fully downloaded."""
983 try:
984 self.to_screen('[download] %s has already been downloaded' % file_name)
985 except UnicodeEncodeError:
986 self.to_screen('[download] The file has already been downloaded')
987
988 def report_file_delete(self, file_name):
989 """Report that existing file will be deleted."""
990 try:
991 self.to_screen('Deleting existing file %s' % file_name)
992 except UnicodeEncodeError:
993 self.to_screen('Deleting existing file')
994
995 def raise_no_formats(self, info, forced=False, *, msg=None):
996 has_drm = info.get('_has_drm')
997 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
998 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
999 if forced or not ignored:
1000 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1001 expected=has_drm or ignored or expected)
1002 else:
1003 self.report_warning(msg)
1004
1005 def parse_outtmpl(self):
1006 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1007 self._parse_outtmpl()
1008 return self.params['outtmpl']
1009
1010 def _parse_outtmpl(self):
1011 sanitize = IDENTITY
1012 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1013 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1014
1015 outtmpl = self.params.setdefault('outtmpl', {})
1016 if not isinstance(outtmpl, dict):
1017 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1018 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1019
1020 def get_output_path(self, dir_type='', filename=None):
1021 paths = self.params.get('paths', {})
1022 assert isinstance(paths, dict)
1023 path = os.path.join(
1024 expand_path(paths.get('home', '').strip()),
1025 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1026 filename or '')
1027 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1028
1029 @staticmethod
1030 def _outtmpl_expandpath(outtmpl):
1031 # expand_path translates '%%' into '%' and '$$' into '$'
1032 # correspondingly that is not what we want since we need to keep
1033 # '%%' intact for template dict substitution step. Working around
1034 # with boundary-alike separator hack.
1035 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1036 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1037
1038 # outtmpl should be expand_path'ed before template dict substitution
1039 # because meta fields may contain env variables we don't want to
1040 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1041 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1042 return expand_path(outtmpl).replace(sep, '')
1043
1044 @staticmethod
1045 def escape_outtmpl(outtmpl):
1046 ''' Escape any remaining strings like %s, %abc% etc. '''
1047 return re.sub(
1048 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1049 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1050 outtmpl)
1051
1052 @classmethod
1053 def validate_outtmpl(cls, outtmpl):
1054 ''' @return None or Exception object '''
1055 outtmpl = re.sub(
1056 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1057 lambda mobj: f'{mobj.group(0)[:-1]}s',
1058 cls._outtmpl_expandpath(outtmpl))
1059 try:
1060 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1061 return None
1062 except ValueError as err:
1063 return err
1064
1065 @staticmethod
1066 def _copy_infodict(info_dict):
1067 info_dict = dict(info_dict)
1068 info_dict.pop('__postprocessors', None)
1069 info_dict.pop('__pending_error', None)
1070 return info_dict
1071
1072 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1073 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1074 @param sanitize Whether to sanitize the output as a filename.
1075 For backward compatibility, a function can also be passed
1076 """
1077
1078 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1079
1080 info_dict = self._copy_infodict(info_dict)
1081 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1082 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1083 if info_dict.get('duration', None) is not None
1084 else None)
1085 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1086 info_dict['video_autonumber'] = self._num_videos
1087 if info_dict.get('resolution') is None:
1088 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1089
1090 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1091 # of %(field)s to %(field)0Nd for backward compatibility
1092 field_size_compat_map = {
1093 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1094 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1095 'autonumber': self.params.get('autonumber_size') or 5,
1096 }
1097
1098 TMPL_DICT = {}
1099 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1100 MATH_FUNCTIONS = {
1101 '+': float.__add__,
1102 '-': float.__sub__,
1103 }
1104 # Field is of the form key1.key2...
1105 # where keys (except first) can be string, int or slice
1106 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1107 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1108 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1109 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1110 (?P<negate>-)?
1111 (?P<fields>{FIELD_RE})
1112 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1113 (?:>(?P<strf_format>.+?))?
1114 (?P<remaining>
1115 (?P<alternate>(?<!\\),[^|&)]+)?
1116 (?:&(?P<replacement>.*?))?
1117 (?:\|(?P<default>.*?))?
1118 )$''')
1119
1120 def _traverse_infodict(k):
1121 k = k.split('.')
1122 if k[0] == '':
1123 k.pop(0)
1124 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1125
1126 def get_value(mdict):
1127 # Object traversal
1128 value = _traverse_infodict(mdict['fields'])
1129 # Negative
1130 if mdict['negate']:
1131 value = float_or_none(value)
1132 if value is not None:
1133 value *= -1
1134 # Do maths
1135 offset_key = mdict['maths']
1136 if offset_key:
1137 value = float_or_none(value)
1138 operator = None
1139 while offset_key:
1140 item = re.match(
1141 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1142 offset_key).group(0)
1143 offset_key = offset_key[len(item):]
1144 if operator is None:
1145 operator = MATH_FUNCTIONS[item]
1146 continue
1147 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1148 offset = float_or_none(item)
1149 if offset is None:
1150 offset = float_or_none(_traverse_infodict(item))
1151 try:
1152 value = operator(value, multiplier * offset)
1153 except (TypeError, ZeroDivisionError):
1154 return None
1155 operator = None
1156 # Datetime formatting
1157 if mdict['strf_format']:
1158 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1159
1160 return value
1161
1162 na = self.params.get('outtmpl_na_placeholder', 'NA')
1163
1164 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1165 return sanitize_filename(str(value), restricted=restricted, is_id=(
1166 bool(re.search(r'(^|[_.])id(\.|$)', key))
1167 if 'filename-sanitization' in self.params['compat_opts']
1168 else NO_DEFAULT))
1169
1170 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1171 sanitize = bool(sanitize)
1172
1173 def _dumpjson_default(obj):
1174 if isinstance(obj, (set, LazyList)):
1175 return list(obj)
1176 return repr(obj)
1177
1178 def create_key(outer_mobj):
1179 if not outer_mobj.group('has_key'):
1180 return outer_mobj.group(0)
1181 key = outer_mobj.group('key')
1182 mobj = re.match(INTERNAL_FORMAT_RE, key)
1183 initial_field = mobj.group('fields') if mobj else ''
1184 value, replacement, default = None, None, na
1185 while mobj:
1186 mobj = mobj.groupdict()
1187 default = mobj['default'] if mobj['default'] is not None else default
1188 value = get_value(mobj)
1189 replacement = mobj['replacement']
1190 if value is None and mobj['alternate']:
1191 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1192 else:
1193 break
1194
1195 fmt = outer_mobj.group('format')
1196 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1197 fmt = f'0{field_size_compat_map[key]:d}d'
1198
1199 value = default if value is None else value if replacement is None else replacement
1200
1201 flags = outer_mobj.group('conversion') or ''
1202 str_fmt = f'{fmt[:-1]}s'
1203 if fmt[-1] == 'l': # list
1204 delim = '\n' if '#' in flags else ', '
1205 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1206 elif fmt[-1] == 'j': # json
1207 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1208 elif fmt[-1] == 'q': # quoted
1209 value = map(str, variadic(value) if '#' in flags else [value])
1210 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1211 elif fmt[-1] == 'B': # bytes
1212 value = f'%{str_fmt}'.encode() % str(value).encode()
1213 value, fmt = value.decode('utf-8', 'ignore'), 's'
1214 elif fmt[-1] == 'U': # unicode normalized
1215 value, fmt = unicodedata.normalize(
1216 # "+" = compatibility equivalence, "#" = NFD
1217 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1218 value), str_fmt
1219 elif fmt[-1] == 'D': # decimal suffix
1220 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1221 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1222 factor=1024 if '#' in flags else 1000)
1223 elif fmt[-1] == 'S': # filename sanitization
1224 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1225 elif fmt[-1] == 'c':
1226 if value:
1227 value = str(value)[0]
1228 else:
1229 fmt = str_fmt
1230 elif fmt[-1] not in 'rs': # numeric
1231 value = float_or_none(value)
1232 if value is None:
1233 value, fmt = default, 's'
1234
1235 if sanitize:
1236 if fmt[-1] == 'r':
1237 # If value is an object, sanitize might convert it to a string
1238 # So we convert it to repr first
1239 value, fmt = repr(value), str_fmt
1240 if fmt[-1] in 'csr':
1241 value = sanitizer(initial_field, value)
1242
1243 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1244 TMPL_DICT[key] = value
1245 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1246
1247 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1248
1249 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1250 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1251 return self.escape_outtmpl(outtmpl) % info_dict
1252
1253 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1254 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1255 if outtmpl is None:
1256 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1257 try:
1258 outtmpl = self._outtmpl_expandpath(outtmpl)
1259 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1260 if not filename:
1261 return None
1262
1263 if tmpl_type in ('', 'temp'):
1264 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1265 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1266 filename = replace_extension(filename, ext, final_ext)
1267 elif tmpl_type:
1268 force_ext = OUTTMPL_TYPES[tmpl_type]
1269 if force_ext:
1270 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1271
1272 # https://github.com/blackjack4494/youtube-dlc/issues/85
1273 trim_file_name = self.params.get('trim_file_name', False)
1274 if trim_file_name:
1275 no_ext, *ext = filename.rsplit('.', 2)
1276 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1277
1278 return filename
1279 except ValueError as err:
1280 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1281 return None
1282
1283 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1284 """Generate the output filename"""
1285 if outtmpl:
1286 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1287 dir_type = None
1288 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1289 if not filename and dir_type not in ('', 'temp'):
1290 return ''
1291
1292 if warn:
1293 if not self.params.get('paths'):
1294 pass
1295 elif filename == '-':
1296 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1297 elif os.path.isabs(filename):
1298 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1299 if filename == '-' or not filename:
1300 return filename
1301
1302 return self.get_output_path(dir_type, filename)
1303
1304 def _match_entry(self, info_dict, incomplete=False, silent=False):
1305 """ Returns None if the file should be downloaded """
1306
1307 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1308
1309 def check_filter():
1310 if 'title' in info_dict:
1311 # This can happen when we're just evaluating the playlist
1312 title = info_dict['title']
1313 matchtitle = self.params.get('matchtitle', False)
1314 if matchtitle:
1315 if not re.search(matchtitle, title, re.IGNORECASE):
1316 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1317 rejecttitle = self.params.get('rejecttitle', False)
1318 if rejecttitle:
1319 if re.search(rejecttitle, title, re.IGNORECASE):
1320 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1321 date = info_dict.get('upload_date')
1322 if date is not None:
1323 dateRange = self.params.get('daterange', DateRange())
1324 if date not in dateRange:
1325 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1326 view_count = info_dict.get('view_count')
1327 if view_count is not None:
1328 min_views = self.params.get('min_views')
1329 if min_views is not None and view_count < min_views:
1330 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1331 max_views = self.params.get('max_views')
1332 if max_views is not None and view_count > max_views:
1333 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1334 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1335 return 'Skipping "%s" because it is age restricted' % video_title
1336
1337 match_filter = self.params.get('match_filter')
1338 if match_filter is not None:
1339 try:
1340 ret = match_filter(info_dict, incomplete=incomplete)
1341 except TypeError:
1342 # For backward compatibility
1343 ret = None if incomplete else match_filter(info_dict)
1344 if ret is NO_DEFAULT:
1345 while True:
1346 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1347 reply = input(self._format_screen(
1348 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1349 if reply in {'y', ''}:
1350 return None
1351 elif reply == 'n':
1352 return f'Skipping {video_title}'
1353 elif ret is not None:
1354 return ret
1355 return None
1356
1357 if self.in_download_archive(info_dict):
1358 reason = '%s has already been recorded in the archive' % video_title
1359 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1360 else:
1361 reason = check_filter()
1362 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1363 if reason is not None:
1364 if not silent:
1365 self.to_screen('[download] ' + reason)
1366 if self.params.get(break_opt, False):
1367 raise break_err()
1368 return reason
1369
1370 @staticmethod
1371 def add_extra_info(info_dict, extra_info):
1372 '''Set the keys from extra_info in info dict if they are missing'''
1373 for key, value in extra_info.items():
1374 info_dict.setdefault(key, value)
1375
1376 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1377 process=True, force_generic_extractor=False):
1378 """
1379 Return a list with a dictionary for each video extracted.
1380
1381 Arguments:
1382 url -- URL to extract
1383
1384 Keyword arguments:
1385 download -- whether to download videos during extraction
1386 ie_key -- extractor key hint
1387 extra_info -- dictionary containing the extra values to add to each result
1388 process -- whether to resolve all unresolved references (URLs, playlist items),
1389 must be True for download to work.
1390 force_generic_extractor -- force using the generic extractor
1391 """
1392
1393 if extra_info is None:
1394 extra_info = {}
1395
1396 if not ie_key and force_generic_extractor:
1397 ie_key = 'Generic'
1398
1399 if ie_key:
1400 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1401 else:
1402 ies = self._ies
1403
1404 for ie_key, ie in ies.items():
1405 if not ie.suitable(url):
1406 continue
1407
1408 if not ie.working():
1409 self.report_warning('The program functionality for this site has been marked as broken, '
1410 'and will probably not work.')
1411
1412 temp_id = ie.get_temp_id(url)
1413 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1414 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1415 if self.params.get('break_on_existing', False):
1416 raise ExistingVideoReached()
1417 break
1418 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1419 else:
1420 self.report_error('no suitable InfoExtractor for URL %s' % url)
1421
1422 def _handle_extraction_exceptions(func):
1423 @functools.wraps(func)
1424 def wrapper(self, *args, **kwargs):
1425 while True:
1426 try:
1427 return func(self, *args, **kwargs)
1428 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1429 raise
1430 except ReExtractInfo as e:
1431 if e.expected:
1432 self.to_screen(f'{e}; Re-extracting data')
1433 else:
1434 self.to_stderr('\r')
1435 self.report_warning(f'{e}; Re-extracting data')
1436 continue
1437 except GeoRestrictedError as e:
1438 msg = e.msg
1439 if e.countries:
1440 msg += '\nThis video is available in %s.' % ', '.join(
1441 map(ISO3166Utils.short2full, e.countries))
1442 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1443 self.report_error(msg)
1444 except ExtractorError as e: # An error we somewhat expected
1445 self.report_error(str(e), e.format_traceback())
1446 except Exception as e:
1447 if self.params.get('ignoreerrors'):
1448 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1449 else:
1450 raise
1451 break
1452 return wrapper
1453
1454 def _wait_for_video(self, ie_result):
1455 if (not self.params.get('wait_for_video')
1456 or ie_result.get('_type', 'video') != 'video'
1457 or ie_result.get('formats') or ie_result.get('url')):
1458 return
1459
1460 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1461 last_msg = ''
1462
1463 def progress(msg):
1464 nonlocal last_msg
1465 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1466 last_msg = msg
1467
1468 min_wait, max_wait = self.params.get('wait_for_video')
1469 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1470 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1471 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1472 self.report_warning('Release time of video is not known')
1473 elif (diff or 0) <= 0:
1474 self.report_warning('Video should already be available according to extracted info')
1475 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1476 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1477
1478 wait_till = time.time() + diff
1479 try:
1480 while True:
1481 diff = wait_till - time.time()
1482 if diff <= 0:
1483 progress('')
1484 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1485 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1486 time.sleep(1)
1487 except KeyboardInterrupt:
1488 progress('')
1489 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1490 except BaseException as e:
1491 if not isinstance(e, ReExtractInfo):
1492 self.to_screen('')
1493 raise
1494
1495 @_handle_extraction_exceptions
1496 def __extract_info(self, url, ie, download, extra_info, process):
1497 ie_result = ie.extract(url)
1498 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1499 return
1500 if isinstance(ie_result, list):
1501 # Backwards compatibility: old IE result format
1502 ie_result = {
1503 '_type': 'compat_list',
1504 'entries': ie_result,
1505 }
1506 if extra_info.get('original_url'):
1507 ie_result.setdefault('original_url', extra_info['original_url'])
1508 self.add_default_extra_info(ie_result, ie, url)
1509 if process:
1510 self._wait_for_video(ie_result)
1511 return self.process_ie_result(ie_result, download, extra_info)
1512 else:
1513 return ie_result
1514
1515 def add_default_extra_info(self, ie_result, ie, url):
1516 if url is not None:
1517 self.add_extra_info(ie_result, {
1518 'webpage_url': url,
1519 'original_url': url,
1520 })
1521 webpage_url = ie_result.get('webpage_url')
1522 if webpage_url:
1523 self.add_extra_info(ie_result, {
1524 'webpage_url_basename': url_basename(webpage_url),
1525 'webpage_url_domain': get_domain(webpage_url),
1526 })
1527 if ie is not None:
1528 self.add_extra_info(ie_result, {
1529 'extractor': ie.IE_NAME,
1530 'extractor_key': ie.ie_key(),
1531 })
1532
1533 def process_ie_result(self, ie_result, download=True, extra_info=None):
1534 """
1535 Take the result of the ie(may be modified) and resolve all unresolved
1536 references (URLs, playlist items).
1537
1538 It will also download the videos if 'download'.
1539 Returns the resolved ie_result.
1540 """
1541 if extra_info is None:
1542 extra_info = {}
1543 result_type = ie_result.get('_type', 'video')
1544
1545 if result_type in ('url', 'url_transparent'):
1546 ie_result['url'] = sanitize_url(ie_result['url'])
1547 if ie_result.get('original_url'):
1548 extra_info.setdefault('original_url', ie_result['original_url'])
1549
1550 extract_flat = self.params.get('extract_flat', False)
1551 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1552 or extract_flat is True):
1553 info_copy = ie_result.copy()
1554 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1555 if ie and not ie_result.get('id'):
1556 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1557 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1558 self.add_extra_info(info_copy, extra_info)
1559 info_copy, _ = self.pre_process(info_copy)
1560 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1561 self._raise_pending_errors(info_copy)
1562 if self.params.get('force_write_download_archive', False):
1563 self.record_download_archive(info_copy)
1564 return ie_result
1565
1566 if result_type == 'video':
1567 self.add_extra_info(ie_result, extra_info)
1568 ie_result = self.process_video_result(ie_result, download=download)
1569 self._raise_pending_errors(ie_result)
1570 additional_urls = (ie_result or {}).get('additional_urls')
1571 if additional_urls:
1572 # TODO: Improve MetadataParserPP to allow setting a list
1573 if isinstance(additional_urls, compat_str):
1574 additional_urls = [additional_urls]
1575 self.to_screen(
1576 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1577 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1578 ie_result['additional_entries'] = [
1579 self.extract_info(
1580 url, download, extra_info=extra_info,
1581 force_generic_extractor=self.params.get('force_generic_extractor'))
1582 for url in additional_urls
1583 ]
1584 return ie_result
1585 elif result_type == 'url':
1586 # We have to add extra_info to the results because it may be
1587 # contained in a playlist
1588 return self.extract_info(
1589 ie_result['url'], download,
1590 ie_key=ie_result.get('ie_key'),
1591 extra_info=extra_info)
1592 elif result_type == 'url_transparent':
1593 # Use the information from the embedding page
1594 info = self.extract_info(
1595 ie_result['url'], ie_key=ie_result.get('ie_key'),
1596 extra_info=extra_info, download=False, process=False)
1597
1598 # extract_info may return None when ignoreerrors is enabled and
1599 # extraction failed with an error, don't crash and return early
1600 # in this case
1601 if not info:
1602 return info
1603
1604 exempted_fields = {'_type', 'url', 'ie_key'}
1605 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1606 # For video clips, the id etc of the clip extractor should be used
1607 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1608
1609 new_result = info.copy()
1610 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1611
1612 # Extracted info may not be a video result (i.e.
1613 # info.get('_type', 'video') != video) but rather an url or
1614 # url_transparent. In such cases outer metadata (from ie_result)
1615 # should be propagated to inner one (info). For this to happen
1616 # _type of info should be overridden with url_transparent. This
1617 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1618 if new_result.get('_type') == 'url':
1619 new_result['_type'] = 'url_transparent'
1620
1621 return self.process_ie_result(
1622 new_result, download=download, extra_info=extra_info)
1623 elif result_type in ('playlist', 'multi_video'):
1624 # Protect from infinite recursion due to recursively nested playlists
1625 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1626 webpage_url = ie_result['webpage_url']
1627 if webpage_url in self._playlist_urls:
1628 self.to_screen(
1629 '[download] Skipping already downloaded playlist: %s'
1630 % ie_result.get('title') or ie_result.get('id'))
1631 return
1632
1633 self._playlist_level += 1
1634 self._playlist_urls.add(webpage_url)
1635 self._fill_common_fields(ie_result, False)
1636 self._sanitize_thumbnails(ie_result)
1637 try:
1638 return self.__process_playlist(ie_result, download)
1639 finally:
1640 self._playlist_level -= 1
1641 if not self._playlist_level:
1642 self._playlist_urls.clear()
1643 elif result_type == 'compat_list':
1644 self.report_warning(
1645 'Extractor %s returned a compat_list result. '
1646 'It needs to be updated.' % ie_result.get('extractor'))
1647
1648 def _fixup(r):
1649 self.add_extra_info(r, {
1650 'extractor': ie_result['extractor'],
1651 'webpage_url': ie_result['webpage_url'],
1652 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1653 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1654 'extractor_key': ie_result['extractor_key'],
1655 })
1656 return r
1657 ie_result['entries'] = [
1658 self.process_ie_result(_fixup(r), download, extra_info)
1659 for r in ie_result['entries']
1660 ]
1661 return ie_result
1662 else:
1663 raise Exception('Invalid result type: %s' % result_type)
1664
1665 def _ensure_dir_exists(self, path):
1666 return make_dir(path, self.report_error)
1667
1668 @staticmethod
1669 def _playlist_infodict(ie_result, **kwargs):
1670 return {
1671 **ie_result,
1672 'playlist': ie_result.get('title') or ie_result.get('id'),
1673 'playlist_id': ie_result.get('id'),
1674 'playlist_title': ie_result.get('title'),
1675 'playlist_uploader': ie_result.get('uploader'),
1676 'playlist_uploader_id': ie_result.get('uploader_id'),
1677 'playlist_index': 0,
1678 **kwargs,
1679 }
1680
1681 def __process_playlist(self, ie_result, download):
1682 """Process each entry in the playlist"""
1683 title = ie_result.get('title') or ie_result.get('id') or '<Untitled>'
1684 self.to_screen(f'[download] Downloading playlist: {title}')
1685
1686 all_entries = PlaylistEntries(self, ie_result)
1687 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1688
1689 lazy = self.params.get('lazy_playlist')
1690 if lazy:
1691 resolved_entries, n_entries = [], 'N/A'
1692 ie_result['requested_entries'], ie_result['entries'] = None, None
1693 else:
1694 entries = resolved_entries = list(entries)
1695 n_entries = len(resolved_entries)
1696 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1697 if not ie_result.get('playlist_count'):
1698 # Better to do this after potentially exhausting entries
1699 ie_result['playlist_count'] = all_entries.get_full_count()
1700
1701 _infojson_written = False
1702 write_playlist_files = self.params.get('allow_playlist_files', True)
1703 if write_playlist_files and self.params.get('list_thumbnails'):
1704 self.list_thumbnails(ie_result)
1705 if write_playlist_files and not self.params.get('simulate'):
1706 ie_copy = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1707 _infojson_written = self._write_info_json(
1708 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1709 if _infojson_written is None:
1710 return
1711 if self._write_description('playlist', ie_result,
1712 self.prepare_filename(ie_copy, 'pl_description')) is None:
1713 return
1714 # TODO: This should be passed to ThumbnailsConvertor if necessary
1715 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1716
1717 if lazy:
1718 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1719 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1720 elif self.params.get('playlistreverse'):
1721 entries.reverse()
1722 elif self.params.get('playlistrandom'):
1723 random.shuffle(entries)
1724
1725 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} videos'
1726 f'{format_field(ie_result, "playlist_count", " of %s")}')
1727
1728 failures = 0
1729 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1730 for i, (playlist_index, entry) in enumerate(entries):
1731 if lazy:
1732 resolved_entries.append((playlist_index, entry))
1733
1734 # TODO: Add auto-generated fields
1735 if not entry or self._match_entry(entry, incomplete=True) is not None:
1736 continue
1737
1738 self.to_screen('[download] Downloading video %s of %s' % (
1739 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1740
1741 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1742 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1743 playlist_index = ie_result['requested_entries'][i]
1744
1745 entry_result = self.__process_iterable_entry(entry, download, {
1746 'n_entries': int_or_none(n_entries),
1747 '__last_playlist_index': max(ie_result['requested_entries'] or (0, 0)),
1748 'playlist_count': ie_result.get('playlist_count'),
1749 'playlist_index': playlist_index,
1750 'playlist_autonumber': i + 1,
1751 'playlist': title,
1752 'playlist_id': ie_result.get('id'),
1753 'playlist_title': ie_result.get('title'),
1754 'playlist_uploader': ie_result.get('uploader'),
1755 'playlist_uploader_id': ie_result.get('uploader_id'),
1756 'extractor': ie_result['extractor'],
1757 'webpage_url': ie_result['webpage_url'],
1758 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1759 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1760 'extractor_key': ie_result['extractor_key'],
1761 })
1762 if not entry_result:
1763 failures += 1
1764 if failures >= max_failures:
1765 self.report_error(
1766 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1767 break
1768 resolved_entries[i] = (playlist_index, entry_result)
1769
1770 # Update with processed data
1771 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1772
1773 # Write the updated info to json
1774 if _infojson_written is True and self._write_info_json(
1775 'updated playlist', ie_result,
1776 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1777 return
1778
1779 ie_result = self.run_all_pps('playlist', ie_result)
1780 self.to_screen(f'[download] Finished downloading playlist: {title}')
1781 return ie_result
1782
1783 @_handle_extraction_exceptions
1784 def __process_iterable_entry(self, entry, download, extra_info):
1785 return self.process_ie_result(
1786 entry, download=download, extra_info=extra_info)
1787
1788 def _build_format_filter(self, filter_spec):
1789 " Returns a function to filter the formats according to the filter_spec "
1790
1791 OPERATORS = {
1792 '<': operator.lt,
1793 '<=': operator.le,
1794 '>': operator.gt,
1795 '>=': operator.ge,
1796 '=': operator.eq,
1797 '!=': operator.ne,
1798 }
1799 operator_rex = re.compile(r'''(?x)\s*
1800 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1801 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1802 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1803 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1804 m = operator_rex.fullmatch(filter_spec)
1805 if m:
1806 try:
1807 comparison_value = int(m.group('value'))
1808 except ValueError:
1809 comparison_value = parse_filesize(m.group('value'))
1810 if comparison_value is None:
1811 comparison_value = parse_filesize(m.group('value') + 'B')
1812 if comparison_value is None:
1813 raise ValueError(
1814 'Invalid value %r in format specification %r' % (
1815 m.group('value'), filter_spec))
1816 op = OPERATORS[m.group('op')]
1817
1818 if not m:
1819 STR_OPERATORS = {
1820 '=': operator.eq,
1821 '^=': lambda attr, value: attr.startswith(value),
1822 '$=': lambda attr, value: attr.endswith(value),
1823 '*=': lambda attr, value: value in attr,
1824 '~=': lambda attr, value: value.search(attr) is not None
1825 }
1826 str_operator_rex = re.compile(r'''(?x)\s*
1827 (?P<key>[a-zA-Z0-9._-]+)\s*
1828 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1829 (?P<quote>["'])?
1830 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1831 (?(quote)(?P=quote))\s*
1832 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1833 m = str_operator_rex.fullmatch(filter_spec)
1834 if m:
1835 if m.group('op') == '~=':
1836 comparison_value = re.compile(m.group('value'))
1837 else:
1838 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1839 str_op = STR_OPERATORS[m.group('op')]
1840 if m.group('negation'):
1841 op = lambda attr, value: not str_op(attr, value)
1842 else:
1843 op = str_op
1844
1845 if not m:
1846 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1847
1848 def _filter(f):
1849 actual_value = f.get(m.group('key'))
1850 if actual_value is None:
1851 return m.group('none_inclusive')
1852 return op(actual_value, comparison_value)
1853 return _filter
1854
1855 def _check_formats(self, formats):
1856 for f in formats:
1857 self.to_screen('[info] Testing format %s' % f['format_id'])
1858 path = self.get_output_path('temp')
1859 if not self._ensure_dir_exists(f'{path}/'):
1860 continue
1861 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1862 temp_file.close()
1863 try:
1864 success, _ = self.dl(temp_file.name, f, test=True)
1865 except (DownloadError, OSError, ValueError) + network_exceptions:
1866 success = False
1867 finally:
1868 if os.path.exists(temp_file.name):
1869 try:
1870 os.remove(temp_file.name)
1871 except OSError:
1872 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1873 if success:
1874 yield f
1875 else:
1876 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1877
1878 def _default_format_spec(self, info_dict, download=True):
1879
1880 def can_merge():
1881 merger = FFmpegMergerPP(self)
1882 return merger.available and merger.can_merge()
1883
1884 prefer_best = (
1885 not self.params.get('simulate')
1886 and download
1887 and (
1888 not can_merge()
1889 or info_dict.get('is_live') and not self.params.get('live_from_start')
1890 or self.params['outtmpl']['default'] == '-'))
1891 compat = (
1892 prefer_best
1893 or self.params.get('allow_multiple_audio_streams', False)
1894 or 'format-spec' in self.params['compat_opts'])
1895
1896 return (
1897 'best/bestvideo+bestaudio' if prefer_best
1898 else 'bestvideo*+bestaudio/best' if not compat
1899 else 'bestvideo+bestaudio/best')
1900
1901 def build_format_selector(self, format_spec):
1902 def syntax_error(note, start):
1903 message = (
1904 'Invalid format specification: '
1905 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1906 return SyntaxError(message)
1907
1908 PICKFIRST = 'PICKFIRST'
1909 MERGE = 'MERGE'
1910 SINGLE = 'SINGLE'
1911 GROUP = 'GROUP'
1912 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1913
1914 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1915 'video': self.params.get('allow_multiple_video_streams', False)}
1916
1917 check_formats = self.params.get('check_formats') == 'selected'
1918
1919 def _parse_filter(tokens):
1920 filter_parts = []
1921 for type, string, start, _, _ in tokens:
1922 if type == tokenize.OP and string == ']':
1923 return ''.join(filter_parts)
1924 else:
1925 filter_parts.append(string)
1926
1927 def _remove_unused_ops(tokens):
1928 # Remove operators that we don't use and join them with the surrounding strings
1929 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1930 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1931 last_string, last_start, last_end, last_line = None, None, None, None
1932 for type, string, start, end, line in tokens:
1933 if type == tokenize.OP and string == '[':
1934 if last_string:
1935 yield tokenize.NAME, last_string, last_start, last_end, last_line
1936 last_string = None
1937 yield type, string, start, end, line
1938 # everything inside brackets will be handled by _parse_filter
1939 for type, string, start, end, line in tokens:
1940 yield type, string, start, end, line
1941 if type == tokenize.OP and string == ']':
1942 break
1943 elif type == tokenize.OP and string in ALLOWED_OPS:
1944 if last_string:
1945 yield tokenize.NAME, last_string, last_start, last_end, last_line
1946 last_string = None
1947 yield type, string, start, end, line
1948 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1949 if not last_string:
1950 last_string = string
1951 last_start = start
1952 last_end = end
1953 else:
1954 last_string += string
1955 if last_string:
1956 yield tokenize.NAME, last_string, last_start, last_end, last_line
1957
1958 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1959 selectors = []
1960 current_selector = None
1961 for type, string, start, _, _ in tokens:
1962 # ENCODING is only defined in python 3.x
1963 if type == getattr(tokenize, 'ENCODING', None):
1964 continue
1965 elif type in [tokenize.NAME, tokenize.NUMBER]:
1966 current_selector = FormatSelector(SINGLE, string, [])
1967 elif type == tokenize.OP:
1968 if string == ')':
1969 if not inside_group:
1970 # ')' will be handled by the parentheses group
1971 tokens.restore_last_token()
1972 break
1973 elif inside_merge and string in ['/', ',']:
1974 tokens.restore_last_token()
1975 break
1976 elif inside_choice and string == ',':
1977 tokens.restore_last_token()
1978 break
1979 elif string == ',':
1980 if not current_selector:
1981 raise syntax_error('"," must follow a format selector', start)
1982 selectors.append(current_selector)
1983 current_selector = None
1984 elif string == '/':
1985 if not current_selector:
1986 raise syntax_error('"/" must follow a format selector', start)
1987 first_choice = current_selector
1988 second_choice = _parse_format_selection(tokens, inside_choice=True)
1989 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1990 elif string == '[':
1991 if not current_selector:
1992 current_selector = FormatSelector(SINGLE, 'best', [])
1993 format_filter = _parse_filter(tokens)
1994 current_selector.filters.append(format_filter)
1995 elif string == '(':
1996 if current_selector:
1997 raise syntax_error('Unexpected "("', start)
1998 group = _parse_format_selection(tokens, inside_group=True)
1999 current_selector = FormatSelector(GROUP, group, [])
2000 elif string == '+':
2001 if not current_selector:
2002 raise syntax_error('Unexpected "+"', start)
2003 selector_1 = current_selector
2004 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2005 if not selector_2:
2006 raise syntax_error('Expected a selector', start)
2007 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2008 else:
2009 raise syntax_error(f'Operator not recognized: "{string}"', start)
2010 elif type == tokenize.ENDMARKER:
2011 break
2012 if current_selector:
2013 selectors.append(current_selector)
2014 return selectors
2015
2016 def _merge(formats_pair):
2017 format_1, format_2 = formats_pair
2018
2019 formats_info = []
2020 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2021 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2022
2023 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2024 get_no_more = {'video': False, 'audio': False}
2025 for (i, fmt_info) in enumerate(formats_info):
2026 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2027 formats_info.pop(i)
2028 continue
2029 for aud_vid in ['audio', 'video']:
2030 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2031 if get_no_more[aud_vid]:
2032 formats_info.pop(i)
2033 break
2034 get_no_more[aud_vid] = True
2035
2036 if len(formats_info) == 1:
2037 return formats_info[0]
2038
2039 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2040 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2041
2042 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2043 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2044
2045 output_ext = self.params.get('merge_output_format')
2046 if not output_ext:
2047 if the_only_video:
2048 output_ext = the_only_video['ext']
2049 elif the_only_audio and not video_fmts:
2050 output_ext = the_only_audio['ext']
2051 else:
2052 output_ext = 'mkv'
2053
2054 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2055
2056 new_dict = {
2057 'requested_formats': formats_info,
2058 'format': '+'.join(filtered('format')),
2059 'format_id': '+'.join(filtered('format_id')),
2060 'ext': output_ext,
2061 'protocol': '+'.join(map(determine_protocol, formats_info)),
2062 'language': '+'.join(orderedSet(filtered('language'))) or None,
2063 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2064 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2065 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2066 }
2067
2068 if the_only_video:
2069 new_dict.update({
2070 'width': the_only_video.get('width'),
2071 'height': the_only_video.get('height'),
2072 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2073 'fps': the_only_video.get('fps'),
2074 'dynamic_range': the_only_video.get('dynamic_range'),
2075 'vcodec': the_only_video.get('vcodec'),
2076 'vbr': the_only_video.get('vbr'),
2077 'stretched_ratio': the_only_video.get('stretched_ratio'),
2078 })
2079
2080 if the_only_audio:
2081 new_dict.update({
2082 'acodec': the_only_audio.get('acodec'),
2083 'abr': the_only_audio.get('abr'),
2084 'asr': the_only_audio.get('asr'),
2085 })
2086
2087 return new_dict
2088
2089 def _check_formats(formats):
2090 if not check_formats:
2091 yield from formats
2092 return
2093 yield from self._check_formats(formats)
2094
2095 def _build_selector_function(selector):
2096 if isinstance(selector, list): # ,
2097 fs = [_build_selector_function(s) for s in selector]
2098
2099 def selector_function(ctx):
2100 for f in fs:
2101 yield from f(ctx)
2102 return selector_function
2103
2104 elif selector.type == GROUP: # ()
2105 selector_function = _build_selector_function(selector.selector)
2106
2107 elif selector.type == PICKFIRST: # /
2108 fs = [_build_selector_function(s) for s in selector.selector]
2109
2110 def selector_function(ctx):
2111 for f in fs:
2112 picked_formats = list(f(ctx))
2113 if picked_formats:
2114 return picked_formats
2115 return []
2116
2117 elif selector.type == MERGE: # +
2118 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2119
2120 def selector_function(ctx):
2121 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2122 yield _merge(pair)
2123
2124 elif selector.type == SINGLE: # atom
2125 format_spec = selector.selector or 'best'
2126
2127 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2128 if format_spec == 'all':
2129 def selector_function(ctx):
2130 yield from _check_formats(ctx['formats'][::-1])
2131 elif format_spec == 'mergeall':
2132 def selector_function(ctx):
2133 formats = list(_check_formats(
2134 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2135 if not formats:
2136 return
2137 merged_format = formats[-1]
2138 for f in formats[-2::-1]:
2139 merged_format = _merge((merged_format, f))
2140 yield merged_format
2141
2142 else:
2143 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2144 mobj = re.match(
2145 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2146 format_spec)
2147 if mobj is not None:
2148 format_idx = int_or_none(mobj.group('n'), default=1)
2149 format_reverse = mobj.group('bw')[0] == 'b'
2150 format_type = (mobj.group('type') or [None])[0]
2151 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2152 format_modified = mobj.group('mod') is not None
2153
2154 format_fallback = not format_type and not format_modified # for b, w
2155 _filter_f = (
2156 (lambda f: f.get('%scodec' % format_type) != 'none')
2157 if format_type and format_modified # bv*, ba*, wv*, wa*
2158 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2159 if format_type # bv, ba, wv, wa
2160 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2161 if not format_modified # b, w
2162 else lambda f: True) # b*, w*
2163 filter_f = lambda f: _filter_f(f) and (
2164 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2165 else:
2166 if format_spec in self._format_selection_exts['audio']:
2167 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2168 elif format_spec in self._format_selection_exts['video']:
2169 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2170 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2171 elif format_spec in self._format_selection_exts['storyboards']:
2172 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2173 else:
2174 filter_f = lambda f: f.get('format_id') == format_spec # id
2175
2176 def selector_function(ctx):
2177 formats = list(ctx['formats'])
2178 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2179 if not matches:
2180 if format_fallback and ctx['incomplete_formats']:
2181 # for extractors with incomplete formats (audio only (soundcloud)
2182 # or video only (imgur)) best/worst will fallback to
2183 # best/worst {video,audio}-only format
2184 matches = formats
2185 elif seperate_fallback and not ctx['has_merged_format']:
2186 # for compatibility with youtube-dl when there is no pre-merged format
2187 matches = list(filter(seperate_fallback, formats))
2188 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2189 try:
2190 yield matches[format_idx - 1]
2191 except LazyList.IndexError:
2192 return
2193
2194 filters = [self._build_format_filter(f) for f in selector.filters]
2195
2196 def final_selector(ctx):
2197 ctx_copy = dict(ctx)
2198 for _filter in filters:
2199 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2200 return selector_function(ctx_copy)
2201 return final_selector
2202
2203 stream = io.BytesIO(format_spec.encode())
2204 try:
2205 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2206 except tokenize.TokenError:
2207 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2208
2209 class TokenIterator:
2210 def __init__(self, tokens):
2211 self.tokens = tokens
2212 self.counter = 0
2213
2214 def __iter__(self):
2215 return self
2216
2217 def __next__(self):
2218 if self.counter >= len(self.tokens):
2219 raise StopIteration()
2220 value = self.tokens[self.counter]
2221 self.counter += 1
2222 return value
2223
2224 next = __next__
2225
2226 def restore_last_token(self):
2227 self.counter -= 1
2228
2229 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2230 return _build_selector_function(parsed_selector)
2231
2232 def _calc_headers(self, info_dict):
2233 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2234
2235 cookies = self._calc_cookies(info_dict['url'])
2236 if cookies:
2237 res['Cookie'] = cookies
2238
2239 if 'X-Forwarded-For' not in res:
2240 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2241 if x_forwarded_for_ip:
2242 res['X-Forwarded-For'] = x_forwarded_for_ip
2243
2244 return res
2245
2246 def _calc_cookies(self, url):
2247 pr = sanitized_Request(url)
2248 self.cookiejar.add_cookie_header(pr)
2249 return pr.get_header('Cookie')
2250
2251 def _sort_thumbnails(self, thumbnails):
2252 thumbnails.sort(key=lambda t: (
2253 t.get('preference') if t.get('preference') is not None else -1,
2254 t.get('width') if t.get('width') is not None else -1,
2255 t.get('height') if t.get('height') is not None else -1,
2256 t.get('id') if t.get('id') is not None else '',
2257 t.get('url')))
2258
2259 def _sanitize_thumbnails(self, info_dict):
2260 thumbnails = info_dict.get('thumbnails')
2261 if thumbnails is None:
2262 thumbnail = info_dict.get('thumbnail')
2263 if thumbnail:
2264 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2265 if not thumbnails:
2266 return
2267
2268 def check_thumbnails(thumbnails):
2269 for t in thumbnails:
2270 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2271 try:
2272 self.urlopen(HEADRequest(t['url']))
2273 except network_exceptions as err:
2274 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2275 continue
2276 yield t
2277
2278 self._sort_thumbnails(thumbnails)
2279 for i, t in enumerate(thumbnails):
2280 if t.get('id') is None:
2281 t['id'] = '%d' % i
2282 if t.get('width') and t.get('height'):
2283 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2284 t['url'] = sanitize_url(t['url'])
2285
2286 if self.params.get('check_formats') is True:
2287 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2288 else:
2289 info_dict['thumbnails'] = thumbnails
2290
2291 def _fill_common_fields(self, info_dict, is_video=True):
2292 # TODO: move sanitization here
2293 if is_video:
2294 # playlists are allowed to lack "title"
2295 title = info_dict.get('title', NO_DEFAULT)
2296 if title is NO_DEFAULT:
2297 raise ExtractorError('Missing "title" field in extractor result',
2298 video_id=info_dict['id'], ie=info_dict['extractor'])
2299 info_dict['fulltitle'] = title
2300 if not title:
2301 if title == '':
2302 self.write_debug('Extractor gave empty title. Creating a generic title')
2303 else:
2304 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2305 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2306
2307 if info_dict.get('duration') is not None:
2308 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2309
2310 for ts_key, date_key in (
2311 ('timestamp', 'upload_date'),
2312 ('release_timestamp', 'release_date'),
2313 ('modified_timestamp', 'modified_date'),
2314 ):
2315 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2316 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2317 # see http://bugs.python.org/issue1646728)
2318 with contextlib.suppress(ValueError, OverflowError, OSError):
2319 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2320 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2321
2322 live_keys = ('is_live', 'was_live')
2323 live_status = info_dict.get('live_status')
2324 if live_status is None:
2325 for key in live_keys:
2326 if info_dict.get(key) is False:
2327 continue
2328 if info_dict.get(key):
2329 live_status = key
2330 break
2331 if all(info_dict.get(key) is False for key in live_keys):
2332 live_status = 'not_live'
2333 if live_status:
2334 info_dict['live_status'] = live_status
2335 for key in live_keys:
2336 if info_dict.get(key) is None:
2337 info_dict[key] = (live_status == key)
2338
2339 # Auto generate title fields corresponding to the *_number fields when missing
2340 # in order to always have clean titles. This is very common for TV series.
2341 for field in ('chapter', 'season', 'episode'):
2342 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2343 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2344
2345 def _raise_pending_errors(self, info):
2346 err = info.pop('__pending_error', None)
2347 if err:
2348 self.report_error(err, tb=False)
2349
2350 def process_video_result(self, info_dict, download=True):
2351 assert info_dict.get('_type', 'video') == 'video'
2352 self._num_videos += 1
2353
2354 if 'id' not in info_dict:
2355 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2356 elif not info_dict.get('id'):
2357 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2358
2359 def report_force_conversion(field, field_not, conversion):
2360 self.report_warning(
2361 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2362 % (field, field_not, conversion))
2363
2364 def sanitize_string_field(info, string_field):
2365 field = info.get(string_field)
2366 if field is None or isinstance(field, compat_str):
2367 return
2368 report_force_conversion(string_field, 'a string', 'string')
2369 info[string_field] = compat_str(field)
2370
2371 def sanitize_numeric_fields(info):
2372 for numeric_field in self._NUMERIC_FIELDS:
2373 field = info.get(numeric_field)
2374 if field is None or isinstance(field, (int, float)):
2375 continue
2376 report_force_conversion(numeric_field, 'numeric', 'int')
2377 info[numeric_field] = int_or_none(field)
2378
2379 sanitize_string_field(info_dict, 'id')
2380 sanitize_numeric_fields(info_dict)
2381 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2382 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2383 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2384 self.report_warning('"duration" field is negative, there is an error in extractor')
2385
2386 if 'playlist' not in info_dict:
2387 # It isn't part of a playlist
2388 info_dict['playlist'] = None
2389 info_dict['playlist_index'] = None
2390
2391 self._sanitize_thumbnails(info_dict)
2392
2393 thumbnail = info_dict.get('thumbnail')
2394 thumbnails = info_dict.get('thumbnails')
2395 if thumbnail:
2396 info_dict['thumbnail'] = sanitize_url(thumbnail)
2397 elif thumbnails:
2398 info_dict['thumbnail'] = thumbnails[-1]['url']
2399
2400 if info_dict.get('display_id') is None and 'id' in info_dict:
2401 info_dict['display_id'] = info_dict['id']
2402
2403 self._fill_common_fields(info_dict)
2404
2405 for cc_kind in ('subtitles', 'automatic_captions'):
2406 cc = info_dict.get(cc_kind)
2407 if cc:
2408 for _, subtitle in cc.items():
2409 for subtitle_format in subtitle:
2410 if subtitle_format.get('url'):
2411 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2412 if subtitle_format.get('ext') is None:
2413 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2414
2415 automatic_captions = info_dict.get('automatic_captions')
2416 subtitles = info_dict.get('subtitles')
2417
2418 info_dict['requested_subtitles'] = self.process_subtitles(
2419 info_dict['id'], subtitles, automatic_captions)
2420
2421 if info_dict.get('formats') is None:
2422 # There's only one format available
2423 formats = [info_dict]
2424 else:
2425 formats = info_dict['formats']
2426
2427 # or None ensures --clean-infojson removes it
2428 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2429 if not self.params.get('allow_unplayable_formats'):
2430 formats = [f for f in formats if not f.get('has_drm')]
2431 if info_dict['_has_drm'] and all(
2432 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2433 self.report_warning(
2434 'This video is DRM protected and only images are available for download. '
2435 'Use --list-formats to see them')
2436
2437 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2438 if not get_from_start:
2439 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2440 if info_dict.get('is_live') and formats:
2441 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2442 if get_from_start and not formats:
2443 self.raise_no_formats(info_dict, msg=(
2444 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2445 'If you want to download from the current time, use --no-live-from-start'))
2446
2447 if not formats:
2448 self.raise_no_formats(info_dict)
2449
2450 def is_wellformed(f):
2451 url = f.get('url')
2452 if not url:
2453 self.report_warning(
2454 '"url" field is missing or empty - skipping format, '
2455 'there is an error in extractor')
2456 return False
2457 if isinstance(url, bytes):
2458 sanitize_string_field(f, 'url')
2459 return True
2460
2461 # Filter out malformed formats for better extraction robustness
2462 formats = list(filter(is_wellformed, formats))
2463
2464 formats_dict = {}
2465
2466 # We check that all the formats have the format and format_id fields
2467 for i, format in enumerate(formats):
2468 sanitize_string_field(format, 'format_id')
2469 sanitize_numeric_fields(format)
2470 format['url'] = sanitize_url(format['url'])
2471 if not format.get('format_id'):
2472 format['format_id'] = compat_str(i)
2473 else:
2474 # Sanitize format_id from characters used in format selector expression
2475 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2476 format_id = format['format_id']
2477 if format_id not in formats_dict:
2478 formats_dict[format_id] = []
2479 formats_dict[format_id].append(format)
2480
2481 # Make sure all formats have unique format_id
2482 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2483 for format_id, ambiguous_formats in formats_dict.items():
2484 ambigious_id = len(ambiguous_formats) > 1
2485 for i, format in enumerate(ambiguous_formats):
2486 if ambigious_id:
2487 format['format_id'] = '%s-%d' % (format_id, i)
2488 if format.get('ext') is None:
2489 format['ext'] = determine_ext(format['url']).lower()
2490 # Ensure there is no conflict between id and ext in format selection
2491 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2492 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2493 format['format_id'] = 'f%s' % format['format_id']
2494
2495 for i, format in enumerate(formats):
2496 if format.get('format') is None:
2497 format['format'] = '{id} - {res}{note}'.format(
2498 id=format['format_id'],
2499 res=self.format_resolution(format),
2500 note=format_field(format, 'format_note', ' (%s)'),
2501 )
2502 if format.get('protocol') is None:
2503 format['protocol'] = determine_protocol(format)
2504 if format.get('resolution') is None:
2505 format['resolution'] = self.format_resolution(format, default=None)
2506 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2507 format['dynamic_range'] = 'SDR'
2508 if (info_dict.get('duration') and format.get('tbr')
2509 and not format.get('filesize') and not format.get('filesize_approx')):
2510 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2511
2512 # Add HTTP headers, so that external programs can use them from the
2513 # json output
2514 full_format_info = info_dict.copy()
2515 full_format_info.update(format)
2516 format['http_headers'] = self._calc_headers(full_format_info)
2517 # Remove private housekeeping stuff
2518 if '__x_forwarded_for_ip' in info_dict:
2519 del info_dict['__x_forwarded_for_ip']
2520
2521 if self.params.get('check_formats') is True:
2522 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2523
2524 if not formats or formats[0] is not info_dict:
2525 # only set the 'formats' fields if the original info_dict list them
2526 # otherwise we end up with a circular reference, the first (and unique)
2527 # element in the 'formats' field in info_dict is info_dict itself,
2528 # which can't be exported to json
2529 info_dict['formats'] = formats
2530
2531 info_dict, _ = self.pre_process(info_dict)
2532
2533 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2534 return info_dict
2535
2536 self.post_extract(info_dict)
2537 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2538
2539 # The pre-processors may have modified the formats
2540 formats = info_dict.get('formats', [info_dict])
2541
2542 list_only = self.params.get('simulate') is None and (
2543 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2544 interactive_format_selection = not list_only and self.format_selector == '-'
2545 if self.params.get('list_thumbnails'):
2546 self.list_thumbnails(info_dict)
2547 if self.params.get('listsubtitles'):
2548 if 'automatic_captions' in info_dict:
2549 self.list_subtitles(
2550 info_dict['id'], automatic_captions, 'automatic captions')
2551 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2552 if self.params.get('listformats') or interactive_format_selection:
2553 self.list_formats(info_dict)
2554 if list_only:
2555 # Without this printing, -F --print-json will not work
2556 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2557 return info_dict
2558
2559 format_selector = self.format_selector
2560 if format_selector is None:
2561 req_format = self._default_format_spec(info_dict, download=download)
2562 self.write_debug('Default format spec: %s' % req_format)
2563 format_selector = self.build_format_selector(req_format)
2564
2565 while True:
2566 if interactive_format_selection:
2567 req_format = input(
2568 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2569 try:
2570 format_selector = self.build_format_selector(req_format)
2571 except SyntaxError as err:
2572 self.report_error(err, tb=False, is_error=False)
2573 continue
2574
2575 formats_to_download = list(format_selector({
2576 'formats': formats,
2577 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2578 'incomplete_formats': (
2579 # All formats are video-only or
2580 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2581 # all formats are audio-only
2582 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2583 }))
2584 if interactive_format_selection and not formats_to_download:
2585 self.report_error('Requested format is not available', tb=False, is_error=False)
2586 continue
2587 break
2588
2589 if not formats_to_download:
2590 if not self.params.get('ignore_no_formats_error'):
2591 raise ExtractorError(
2592 'Requested format is not available. Use --list-formats for a list of available formats',
2593 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2594 self.report_warning('Requested format is not available')
2595 # Process what we can, even without any available formats.
2596 formats_to_download = [{}]
2597
2598 requested_ranges = self.params.get('download_ranges')
2599 if requested_ranges:
2600 requested_ranges = tuple(requested_ranges(info_dict, self))
2601
2602 best_format, downloaded_formats = formats_to_download[-1], []
2603 if download:
2604 if best_format:
2605 def to_screen(*msg):
2606 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2607
2608 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2609 (f['format_id'] for f in formats_to_download))
2610 if requested_ranges:
2611 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2612 (f'{int(c["start_time"])}-{int(c["end_time"])}' for c in requested_ranges))
2613 max_downloads_reached = False
2614
2615 for fmt, chapter in itertools.product(formats_to_download, requested_ranges or [{}]):
2616 new_info = self._copy_infodict(info_dict)
2617 new_info.update(fmt)
2618 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2619 if chapter or offset:
2620 new_info.update({
2621 'section_start': offset + chapter.get('start_time', 0),
2622 'section_end': offset + min(chapter.get('end_time', duration), duration),
2623 'section_title': chapter.get('title'),
2624 'section_number': chapter.get('index'),
2625 })
2626 downloaded_formats.append(new_info)
2627 try:
2628 self.process_info(new_info)
2629 except MaxDownloadsReached:
2630 max_downloads_reached = True
2631 self._raise_pending_errors(new_info)
2632 # Remove copied info
2633 for key, val in tuple(new_info.items()):
2634 if info_dict.get(key) == val:
2635 new_info.pop(key)
2636 if max_downloads_reached:
2637 break
2638
2639 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2640 assert write_archive.issubset({True, False, 'ignore'})
2641 if True in write_archive and False not in write_archive:
2642 self.record_download_archive(info_dict)
2643
2644 info_dict['requested_downloads'] = downloaded_formats
2645 info_dict = self.run_all_pps('after_video', info_dict)
2646 if max_downloads_reached:
2647 raise MaxDownloadsReached()
2648
2649 # We update the info dict with the selected best quality format (backwards compatibility)
2650 info_dict.update(best_format)
2651 return info_dict
2652
2653 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2654 """Select the requested subtitles and their format"""
2655 available_subs, normal_sub_langs = {}, []
2656 if normal_subtitles and self.params.get('writesubtitles'):
2657 available_subs.update(normal_subtitles)
2658 normal_sub_langs = tuple(normal_subtitles.keys())
2659 if automatic_captions and self.params.get('writeautomaticsub'):
2660 for lang, cap_info in automatic_captions.items():
2661 if lang not in available_subs:
2662 available_subs[lang] = cap_info
2663
2664 if (not self.params.get('writesubtitles') and not
2665 self.params.get('writeautomaticsub') or not
2666 available_subs):
2667 return None
2668
2669 all_sub_langs = tuple(available_subs.keys())
2670 if self.params.get('allsubtitles', False):
2671 requested_langs = all_sub_langs
2672 elif self.params.get('subtitleslangs', False):
2673 # A list is used so that the order of languages will be the same as
2674 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2675 requested_langs = []
2676 for lang_re in self.params.get('subtitleslangs'):
2677 discard = lang_re[0] == '-'
2678 if discard:
2679 lang_re = lang_re[1:]
2680 if lang_re == 'all':
2681 if discard:
2682 requested_langs = []
2683 else:
2684 requested_langs.extend(all_sub_langs)
2685 continue
2686 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2687 if discard:
2688 for lang in current_langs:
2689 while lang in requested_langs:
2690 requested_langs.remove(lang)
2691 else:
2692 requested_langs.extend(current_langs)
2693 requested_langs = orderedSet(requested_langs)
2694 elif normal_sub_langs:
2695 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2696 else:
2697 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2698 if requested_langs:
2699 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2700
2701 formats_query = self.params.get('subtitlesformat', 'best')
2702 formats_preference = formats_query.split('/') if formats_query else []
2703 subs = {}
2704 for lang in requested_langs:
2705 formats = available_subs.get(lang)
2706 if formats is None:
2707 self.report_warning(f'{lang} subtitles not available for {video_id}')
2708 continue
2709 for ext in formats_preference:
2710 if ext == 'best':
2711 f = formats[-1]
2712 break
2713 matches = list(filter(lambda f: f['ext'] == ext, formats))
2714 if matches:
2715 f = matches[-1]
2716 break
2717 else:
2718 f = formats[-1]
2719 self.report_warning(
2720 'No subtitle format found matching "%s" for language %s, '
2721 'using %s' % (formats_query, lang, f['ext']))
2722 subs[lang] = f
2723 return subs
2724
2725 def _forceprint(self, key, info_dict):
2726 if info_dict is None:
2727 return
2728 info_copy = info_dict.copy()
2729 info_copy['formats_table'] = self.render_formats_table(info_dict)
2730 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2731 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2732 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2733
2734 def format_tmpl(tmpl):
2735 mobj = re.match(r'\w+(=?)$', tmpl)
2736 if mobj and mobj.group(1):
2737 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2738 elif mobj:
2739 return f'%({tmpl})s'
2740 return tmpl
2741
2742 for tmpl in self.params['forceprint'].get(key, []):
2743 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2744
2745 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2746 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2747 tmpl = format_tmpl(tmpl)
2748 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2749 if self._ensure_dir_exists(filename):
2750 with open(filename, 'a', encoding='utf-8') as f:
2751 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2752
2753 def __forced_printings(self, info_dict, filename, incomplete):
2754 def print_mandatory(field, actual_field=None):
2755 if actual_field is None:
2756 actual_field = field
2757 if (self.params.get('force%s' % field, False)
2758 and (not incomplete or info_dict.get(actual_field) is not None)):
2759 self.to_stdout(info_dict[actual_field])
2760
2761 def print_optional(field):
2762 if (self.params.get('force%s' % field, False)
2763 and info_dict.get(field) is not None):
2764 self.to_stdout(info_dict[field])
2765
2766 info_dict = info_dict.copy()
2767 if filename is not None:
2768 info_dict['filename'] = filename
2769 if info_dict.get('requested_formats') is not None:
2770 # For RTMP URLs, also include the playpath
2771 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2772 elif info_dict.get('url'):
2773 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2774
2775 if (self.params.get('forcejson')
2776 or self.params['forceprint'].get('video')
2777 or self.params['print_to_file'].get('video')):
2778 self.post_extract(info_dict)
2779 self._forceprint('video', info_dict)
2780
2781 print_mandatory('title')
2782 print_mandatory('id')
2783 print_mandatory('url', 'urls')
2784 print_optional('thumbnail')
2785 print_optional('description')
2786 print_optional('filename')
2787 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2788 self.to_stdout(formatSeconds(info_dict['duration']))
2789 print_mandatory('format')
2790
2791 if self.params.get('forcejson'):
2792 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2793
2794 def dl(self, name, info, subtitle=False, test=False):
2795 if not info.get('url'):
2796 self.raise_no_formats(info, True)
2797
2798 if test:
2799 verbose = self.params.get('verbose')
2800 params = {
2801 'test': True,
2802 'quiet': self.params.get('quiet') or not verbose,
2803 'verbose': verbose,
2804 'noprogress': not verbose,
2805 'nopart': True,
2806 'skip_unavailable_fragments': False,
2807 'keep_fragments': False,
2808 'overwrites': True,
2809 '_no_ytdl_file': True,
2810 }
2811 else:
2812 params = self.params
2813 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2814 if not test:
2815 for ph in self._progress_hooks:
2816 fd.add_progress_hook(ph)
2817 urls = '", "'.join(
2818 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2819 for f in info.get('requested_formats', []) or [info])
2820 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2821
2822 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2823 # But it may contain objects that are not deep-copyable
2824 new_info = self._copy_infodict(info)
2825 if new_info.get('http_headers') is None:
2826 new_info['http_headers'] = self._calc_headers(new_info)
2827 return fd.download(name, new_info, subtitle)
2828
2829 def existing_file(self, filepaths, *, default_overwrite=True):
2830 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2831 if existing_files and not self.params.get('overwrites', default_overwrite):
2832 return existing_files[0]
2833
2834 for file in existing_files:
2835 self.report_file_delete(file)
2836 os.remove(file)
2837 return None
2838
2839 def process_info(self, info_dict):
2840 """Process a single resolved IE result. (Modifies it in-place)"""
2841
2842 assert info_dict.get('_type', 'video') == 'video'
2843 original_infodict = info_dict
2844
2845 if 'format' not in info_dict and 'ext' in info_dict:
2846 info_dict['format'] = info_dict['ext']
2847
2848 # This is mostly just for backward compatibility of process_info
2849 # As a side-effect, this allows for format-specific filters
2850 if self._match_entry(info_dict) is not None:
2851 info_dict['__write_download_archive'] = 'ignore'
2852 return
2853
2854 # Does nothing under normal operation - for backward compatibility of process_info
2855 self.post_extract(info_dict)
2856 self._num_downloads += 1
2857
2858 # info_dict['_filename'] needs to be set for backward compatibility
2859 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2860 temp_filename = self.prepare_filename(info_dict, 'temp')
2861 files_to_move = {}
2862
2863 # Forced printings
2864 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2865
2866 def check_max_downloads():
2867 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
2868 raise MaxDownloadsReached()
2869
2870 if self.params.get('simulate'):
2871 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2872 check_max_downloads()
2873 return
2874
2875 if full_filename is None:
2876 return
2877 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2878 return
2879 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2880 return
2881
2882 if self._write_description('video', info_dict,
2883 self.prepare_filename(info_dict, 'description')) is None:
2884 return
2885
2886 sub_files = self._write_subtitles(info_dict, temp_filename)
2887 if sub_files is None:
2888 return
2889 files_to_move.update(dict(sub_files))
2890
2891 thumb_files = self._write_thumbnails(
2892 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2893 if thumb_files is None:
2894 return
2895 files_to_move.update(dict(thumb_files))
2896
2897 infofn = self.prepare_filename(info_dict, 'infojson')
2898 _infojson_written = self._write_info_json('video', info_dict, infofn)
2899 if _infojson_written:
2900 info_dict['infojson_filename'] = infofn
2901 # For backward compatibility, even though it was a private field
2902 info_dict['__infojson_filename'] = infofn
2903 elif _infojson_written is None:
2904 return
2905
2906 # Note: Annotations are deprecated
2907 annofn = None
2908 if self.params.get('writeannotations', False):
2909 annofn = self.prepare_filename(info_dict, 'annotation')
2910 if annofn:
2911 if not self._ensure_dir_exists(encodeFilename(annofn)):
2912 return
2913 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2914 self.to_screen('[info] Video annotations are already present')
2915 elif not info_dict.get('annotations'):
2916 self.report_warning('There are no annotations to write.')
2917 else:
2918 try:
2919 self.to_screen('[info] Writing video annotations to: ' + annofn)
2920 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2921 annofile.write(info_dict['annotations'])
2922 except (KeyError, TypeError):
2923 self.report_warning('There are no annotations to write.')
2924 except OSError:
2925 self.report_error('Cannot write annotations file: ' + annofn)
2926 return
2927
2928 # Write internet shortcut files
2929 def _write_link_file(link_type):
2930 url = try_get(info_dict['webpage_url'], iri_to_uri)
2931 if not url:
2932 self.report_warning(
2933 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2934 return True
2935 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2936 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2937 return False
2938 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2939 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2940 return True
2941 try:
2942 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2943 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2944 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2945 template_vars = {'url': url}
2946 if link_type == 'desktop':
2947 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2948 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2949 except OSError:
2950 self.report_error(f'Cannot write internet shortcut {linkfn}')
2951 return False
2952 return True
2953
2954 write_links = {
2955 'url': self.params.get('writeurllink'),
2956 'webloc': self.params.get('writewebloclink'),
2957 'desktop': self.params.get('writedesktoplink'),
2958 }
2959 if self.params.get('writelink'):
2960 link_type = ('webloc' if sys.platform == 'darwin'
2961 else 'desktop' if sys.platform.startswith('linux')
2962 else 'url')
2963 write_links[link_type] = True
2964
2965 if any(should_write and not _write_link_file(link_type)
2966 for link_type, should_write in write_links.items()):
2967 return
2968
2969 def replace_info_dict(new_info):
2970 nonlocal info_dict
2971 if new_info == info_dict:
2972 return
2973 info_dict.clear()
2974 info_dict.update(new_info)
2975
2976 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2977 replace_info_dict(new_info)
2978
2979 if self.params.get('skip_download'):
2980 info_dict['filepath'] = temp_filename
2981 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2982 info_dict['__files_to_move'] = files_to_move
2983 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2984 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2985 else:
2986 # Download
2987 info_dict.setdefault('__postprocessors', [])
2988 try:
2989
2990 def existing_video_file(*filepaths):
2991 ext = info_dict.get('ext')
2992 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
2993 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
2994 default_overwrite=False)
2995 if file:
2996 info_dict['ext'] = os.path.splitext(file)[1][1:]
2997 return file
2998
2999 fd, success = None, True
3000 if info_dict.get('protocol') or info_dict.get('url'):
3001 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3002 if fd is not FFmpegFD and (
3003 info_dict.get('section_start') or info_dict.get('section_end')):
3004 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3005 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3006 self.report_error(f'{msg}. Aborting')
3007 return
3008
3009 if info_dict.get('requested_formats') is not None:
3010
3011 def compatible_formats(formats):
3012 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3013 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3014 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3015 if len(video_formats) > 2 or len(audio_formats) > 2:
3016 return False
3017
3018 # Check extension
3019 exts = {format.get('ext') for format in formats}
3020 COMPATIBLE_EXTS = (
3021 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3022 {'webm'},
3023 )
3024 for ext_sets in COMPATIBLE_EXTS:
3025 if ext_sets.issuperset(exts):
3026 return True
3027 # TODO: Check acodec/vcodec
3028 return False
3029
3030 requested_formats = info_dict['requested_formats']
3031 old_ext = info_dict['ext']
3032 if self.params.get('merge_output_format') is None:
3033 if not compatible_formats(requested_formats):
3034 info_dict['ext'] = 'mkv'
3035 self.report_warning(
3036 'Requested formats are incompatible for merge and will be merged into mkv')
3037 if (info_dict['ext'] == 'webm'
3038 and info_dict.get('thumbnails')
3039 # check with type instead of pp_key, __name__, or isinstance
3040 # since we dont want any custom PPs to trigger this
3041 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3042 info_dict['ext'] = 'mkv'
3043 self.report_warning(
3044 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3045 new_ext = info_dict['ext']
3046
3047 def correct_ext(filename, ext=new_ext):
3048 if filename == '-':
3049 return filename
3050 filename_real_ext = os.path.splitext(filename)[1][1:]
3051 filename_wo_ext = (
3052 os.path.splitext(filename)[0]
3053 if filename_real_ext in (old_ext, new_ext)
3054 else filename)
3055 return f'{filename_wo_ext}.{ext}'
3056
3057 # Ensure filename always has a correct extension for successful merge
3058 full_filename = correct_ext(full_filename)
3059 temp_filename = correct_ext(temp_filename)
3060 dl_filename = existing_video_file(full_filename, temp_filename)
3061 info_dict['__real_download'] = False
3062
3063 merger = FFmpegMergerPP(self)
3064 downloaded = []
3065 if dl_filename is not None:
3066 self.report_file_already_downloaded(dl_filename)
3067 elif fd:
3068 for f in requested_formats if fd != FFmpegFD else []:
3069 f['filepath'] = fname = prepend_extension(
3070 correct_ext(temp_filename, info_dict['ext']),
3071 'f%s' % f['format_id'], info_dict['ext'])
3072 downloaded.append(fname)
3073 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3074 success, real_download = self.dl(temp_filename, info_dict)
3075 info_dict['__real_download'] = real_download
3076 else:
3077 if self.params.get('allow_unplayable_formats'):
3078 self.report_warning(
3079 'You have requested merging of multiple formats '
3080 'while also allowing unplayable formats to be downloaded. '
3081 'The formats won\'t be merged to prevent data corruption.')
3082 elif not merger.available:
3083 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3084 if not self.params.get('ignoreerrors'):
3085 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3086 return
3087 self.report_warning(f'{msg}. The formats won\'t be merged')
3088
3089 if temp_filename == '-':
3090 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3091 else 'but the formats are incompatible for simultaneous download' if merger.available
3092 else 'but ffmpeg is not installed')
3093 self.report_warning(
3094 f'You have requested downloading multiple formats to stdout {reason}. '
3095 'The formats will be streamed one after the other')
3096 fname = temp_filename
3097 for f in requested_formats:
3098 new_info = dict(info_dict)
3099 del new_info['requested_formats']
3100 new_info.update(f)
3101 if temp_filename != '-':
3102 fname = prepend_extension(
3103 correct_ext(temp_filename, new_info['ext']),
3104 'f%s' % f['format_id'], new_info['ext'])
3105 if not self._ensure_dir_exists(fname):
3106 return
3107 f['filepath'] = fname
3108 downloaded.append(fname)
3109 partial_success, real_download = self.dl(fname, new_info)
3110 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3111 success = success and partial_success
3112
3113 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3114 info_dict['__postprocessors'].append(merger)
3115 info_dict['__files_to_merge'] = downloaded
3116 # Even if there were no downloads, it is being merged only now
3117 info_dict['__real_download'] = True
3118 else:
3119 for file in downloaded:
3120 files_to_move[file] = None
3121 else:
3122 # Just a single file
3123 dl_filename = existing_video_file(full_filename, temp_filename)
3124 if dl_filename is None or dl_filename == temp_filename:
3125 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3126 # So we should try to resume the download
3127 success, real_download = self.dl(temp_filename, info_dict)
3128 info_dict['__real_download'] = real_download
3129 else:
3130 self.report_file_already_downloaded(dl_filename)
3131
3132 dl_filename = dl_filename or temp_filename
3133 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3134
3135 except network_exceptions as err:
3136 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3137 return
3138 except OSError as err:
3139 raise UnavailableVideoError(err)
3140 except (ContentTooShortError, ) as err:
3141 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3142 return
3143
3144 self._raise_pending_errors(info_dict)
3145 if success and full_filename != '-':
3146
3147 def fixup():
3148 do_fixup = True
3149 fixup_policy = self.params.get('fixup')
3150 vid = info_dict['id']
3151
3152 if fixup_policy in ('ignore', 'never'):
3153 return
3154 elif fixup_policy == 'warn':
3155 do_fixup = 'warn'
3156 elif fixup_policy != 'force':
3157 assert fixup_policy in ('detect_or_warn', None)
3158 if not info_dict.get('__real_download'):
3159 do_fixup = False
3160
3161 def ffmpeg_fixup(cndn, msg, cls):
3162 if not (do_fixup and cndn):
3163 return
3164 elif do_fixup == 'warn':
3165 self.report_warning(f'{vid}: {msg}')
3166 return
3167 pp = cls(self)
3168 if pp.available:
3169 info_dict['__postprocessors'].append(pp)
3170 else:
3171 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3172
3173 stretched_ratio = info_dict.get('stretched_ratio')
3174 ffmpeg_fixup(
3175 stretched_ratio not in (1, None),
3176 f'Non-uniform pixel ratio {stretched_ratio}',
3177 FFmpegFixupStretchedPP)
3178
3179 ffmpeg_fixup(
3180 (info_dict.get('requested_formats') is None
3181 and info_dict.get('container') == 'm4a_dash'
3182 and info_dict.get('ext') == 'm4a'),
3183 'writing DASH m4a. Only some players support this container',
3184 FFmpegFixupM4aPP)
3185
3186 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3187 downloader = downloader.FD_NAME if downloader else None
3188
3189 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3190 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3191 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3192 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3193 FFmpegFixupM3u8PP)
3194 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3195 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3196
3197 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3198 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3199
3200 fixup()
3201 try:
3202 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3203 except PostProcessingError as err:
3204 self.report_error('Postprocessing: %s' % str(err))
3205 return
3206 try:
3207 for ph in self._post_hooks:
3208 ph(info_dict['filepath'])
3209 except Exception as err:
3210 self.report_error('post hooks: %s' % str(err))
3211 return
3212 info_dict['__write_download_archive'] = True
3213
3214 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3215 if self.params.get('force_write_download_archive'):
3216 info_dict['__write_download_archive'] = True
3217 check_max_downloads()
3218
3219 def __download_wrapper(self, func):
3220 @functools.wraps(func)
3221 def wrapper(*args, **kwargs):
3222 try:
3223 res = func(*args, **kwargs)
3224 except UnavailableVideoError as e:
3225 self.report_error(e)
3226 except DownloadCancelled as e:
3227 self.to_screen(f'[info] {e}')
3228 if not self.params.get('break_per_url'):
3229 raise
3230 else:
3231 if self.params.get('dump_single_json', False):
3232 self.post_extract(res)
3233 self.to_stdout(json.dumps(self.sanitize_info(res)))
3234 return wrapper
3235
3236 def download(self, url_list):
3237 """Download a given list of URLs."""
3238 url_list = variadic(url_list) # Passing a single URL is a common mistake
3239 outtmpl = self.params['outtmpl']['default']
3240 if (len(url_list) > 1
3241 and outtmpl != '-'
3242 and '%' not in outtmpl
3243 and self.params.get('max_downloads') != 1):
3244 raise SameFileError(outtmpl)
3245
3246 for url in url_list:
3247 self.__download_wrapper(self.extract_info)(
3248 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3249
3250 return self._download_retcode
3251
3252 def download_with_info_file(self, info_filename):
3253 with contextlib.closing(fileinput.FileInput(
3254 [info_filename], mode='r',
3255 openhook=fileinput.hook_encoded('utf-8'))) as f:
3256 # FileInput doesn't have a read method, we can't call json.load
3257 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3258 try:
3259 self.__download_wrapper(self.process_ie_result)(info, download=True)
3260 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3261 if not isinstance(e, EntryNotInPlaylist):
3262 self.to_stderr('\r')
3263 webpage_url = info.get('webpage_url')
3264 if webpage_url is not None:
3265 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3266 return self.download([webpage_url])
3267 else:
3268 raise
3269 return self._download_retcode
3270
3271 @staticmethod
3272 def sanitize_info(info_dict, remove_private_keys=False):
3273 ''' Sanitize the infodict for converting to json '''
3274 if info_dict is None:
3275 return info_dict
3276 info_dict.setdefault('epoch', int(time.time()))
3277 info_dict.setdefault('_type', 'video')
3278
3279 if remove_private_keys:
3280 reject = lambda k, v: v is None or k.startswith('__') or k in {
3281 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3282 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3283 }
3284 else:
3285 reject = lambda k, v: False
3286
3287 def filter_fn(obj):
3288 if isinstance(obj, dict):
3289 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3290 elif isinstance(obj, (list, tuple, set, LazyList)):
3291 return list(map(filter_fn, obj))
3292 elif obj is None or isinstance(obj, (str, int, float, bool)):
3293 return obj
3294 else:
3295 return repr(obj)
3296
3297 return filter_fn(info_dict)
3298
3299 @staticmethod
3300 def filter_requested_info(info_dict, actually_filter=True):
3301 ''' Alias of sanitize_info for backward compatibility '''
3302 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3303
3304 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3305 for filename in set(filter(None, files_to_delete)):
3306 if msg:
3307 self.to_screen(msg % filename)
3308 try:
3309 os.remove(filename)
3310 except OSError:
3311 self.report_warning(f'Unable to delete file {filename}')
3312 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3313 del info['__files_to_move'][filename]
3314
3315 @staticmethod
3316 def post_extract(info_dict):
3317 def actual_post_extract(info_dict):
3318 if info_dict.get('_type') in ('playlist', 'multi_video'):
3319 for video_dict in info_dict.get('entries', {}):
3320 actual_post_extract(video_dict or {})
3321 return
3322
3323 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3324 info_dict.update(post_extractor())
3325
3326 actual_post_extract(info_dict or {})
3327
3328 def run_pp(self, pp, infodict):
3329 files_to_delete = []
3330 if '__files_to_move' not in infodict:
3331 infodict['__files_to_move'] = {}
3332 try:
3333 files_to_delete, infodict = pp.run(infodict)
3334 except PostProcessingError as e:
3335 # Must be True and not 'only_download'
3336 if self.params.get('ignoreerrors') is True:
3337 self.report_error(e)
3338 return infodict
3339 raise
3340
3341 if not files_to_delete:
3342 return infodict
3343 if self.params.get('keepvideo', False):
3344 for f in files_to_delete:
3345 infodict['__files_to_move'].setdefault(f, '')
3346 else:
3347 self._delete_downloaded_files(
3348 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3349 return infodict
3350
3351 def run_all_pps(self, key, info, *, additional_pps=None):
3352 self._forceprint(key, info)
3353 for pp in (additional_pps or []) + self._pps[key]:
3354 info = self.run_pp(pp, info)
3355 return info
3356
3357 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3358 info = dict(ie_info)
3359 info['__files_to_move'] = files_to_move or {}
3360 try:
3361 info = self.run_all_pps(key, info)
3362 except PostProcessingError as err:
3363 msg = f'Preprocessing: {err}'
3364 info.setdefault('__pending_error', msg)
3365 self.report_error(msg, is_error=False)
3366 return info, info.pop('__files_to_move', None)
3367
3368 def post_process(self, filename, info, files_to_move=None):
3369 """Run all the postprocessors on the given file."""
3370 info['filepath'] = filename
3371 info['__files_to_move'] = files_to_move or {}
3372 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3373 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3374 del info['__files_to_move']
3375 return self.run_all_pps('after_move', info)
3376
3377 def _make_archive_id(self, info_dict):
3378 video_id = info_dict.get('id')
3379 if not video_id:
3380 return
3381 # Future-proof against any change in case
3382 # and backwards compatibility with prior versions
3383 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3384 if extractor is None:
3385 url = str_or_none(info_dict.get('url'))
3386 if not url:
3387 return
3388 # Try to find matching extractor for the URL and take its ie_key
3389 for ie_key, ie in self._ies.items():
3390 if ie.suitable(url):
3391 extractor = ie_key
3392 break
3393 else:
3394 return
3395 return f'{extractor.lower()} {video_id}'
3396
3397 def in_download_archive(self, info_dict):
3398 fn = self.params.get('download_archive')
3399 if fn is None:
3400 return False
3401
3402 vid_id = self._make_archive_id(info_dict)
3403 if not vid_id:
3404 return False # Incomplete video information
3405
3406 return vid_id in self.archive
3407
3408 def record_download_archive(self, info_dict):
3409 fn = self.params.get('download_archive')
3410 if fn is None:
3411 return
3412 vid_id = self._make_archive_id(info_dict)
3413 assert vid_id
3414 self.write_debug(f'Adding to archive: {vid_id}')
3415 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3416 archive_file.write(vid_id + '\n')
3417 self.archive.add(vid_id)
3418
3419 @staticmethod
3420 def format_resolution(format, default='unknown'):
3421 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3422 return 'audio only'
3423 if format.get('resolution') is not None:
3424 return format['resolution']
3425 if format.get('width') and format.get('height'):
3426 return '%dx%d' % (format['width'], format['height'])
3427 elif format.get('height'):
3428 return '%sp' % format['height']
3429 elif format.get('width'):
3430 return '%dx?' % format['width']
3431 return default
3432
3433 def _list_format_headers(self, *headers):
3434 if self.params.get('listformats_table', True) is not False:
3435 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3436 return headers
3437
3438 def _format_note(self, fdict):
3439 res = ''
3440 if fdict.get('ext') in ['f4f', 'f4m']:
3441 res += '(unsupported)'
3442 if fdict.get('language'):
3443 if res:
3444 res += ' '
3445 res += '[%s]' % fdict['language']
3446 if fdict.get('format_note') is not None:
3447 if res:
3448 res += ' '
3449 res += fdict['format_note']
3450 if fdict.get('tbr') is not None:
3451 if res:
3452 res += ', '
3453 res += '%4dk' % fdict['tbr']
3454 if fdict.get('container') is not None:
3455 if res:
3456 res += ', '
3457 res += '%s container' % fdict['container']
3458 if (fdict.get('vcodec') is not None
3459 and fdict.get('vcodec') != 'none'):
3460 if res:
3461 res += ', '
3462 res += fdict['vcodec']
3463 if fdict.get('vbr') is not None:
3464 res += '@'
3465 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3466 res += 'video@'
3467 if fdict.get('vbr') is not None:
3468 res += '%4dk' % fdict['vbr']
3469 if fdict.get('fps') is not None:
3470 if res:
3471 res += ', '
3472 res += '%sfps' % fdict['fps']
3473 if fdict.get('acodec') is not None:
3474 if res:
3475 res += ', '
3476 if fdict['acodec'] == 'none':
3477 res += 'video only'
3478 else:
3479 res += '%-5s' % fdict['acodec']
3480 elif fdict.get('abr') is not None:
3481 if res:
3482 res += ', '
3483 res += 'audio'
3484 if fdict.get('abr') is not None:
3485 res += '@%3dk' % fdict['abr']
3486 if fdict.get('asr') is not None:
3487 res += ' (%5dHz)' % fdict['asr']
3488 if fdict.get('filesize') is not None:
3489 if res:
3490 res += ', '
3491 res += format_bytes(fdict['filesize'])
3492 elif fdict.get('filesize_approx') is not None:
3493 if res:
3494 res += ', '
3495 res += '~' + format_bytes(fdict['filesize_approx'])
3496 return res
3497
3498 def render_formats_table(self, info_dict):
3499 if not info_dict.get('formats') and not info_dict.get('url'):
3500 return None
3501
3502 formats = info_dict.get('formats', [info_dict])
3503 if not self.params.get('listformats_table', True) is not False:
3504 table = [
3505 [
3506 format_field(f, 'format_id'),
3507 format_field(f, 'ext'),
3508 self.format_resolution(f),
3509 self._format_note(f)
3510 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3511 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3512
3513 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3514 table = [
3515 [
3516 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3517 format_field(f, 'ext'),
3518 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3519 format_field(f, 'fps', '\t%d'),
3520 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3521 delim,
3522 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3523 format_field(f, 'tbr', '\t%dk'),
3524 shorten_protocol_name(f.get('protocol', '')),
3525 delim,
3526 format_field(f, 'vcodec', default='unknown').replace(
3527 'none', 'images' if f.get('acodec') == 'none'
3528 else self._format_out('audio only', self.Styles.SUPPRESS)),
3529 format_field(f, 'vbr', '\t%dk'),
3530 format_field(f, 'acodec', default='unknown').replace(
3531 'none', '' if f.get('vcodec') == 'none'
3532 else self._format_out('video only', self.Styles.SUPPRESS)),
3533 format_field(f, 'abr', '\t%dk'),
3534 format_field(f, 'asr', '\t%dHz'),
3535 join_nonempty(
3536 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3537 format_field(f, 'language', '[%s]'),
3538 join_nonempty(format_field(f, 'format_note'),
3539 format_field(f, 'container', ignore=(None, f.get('ext'))),
3540 delim=', '),
3541 delim=' '),
3542 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3543 header_line = self._list_format_headers(
3544 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3545 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3546
3547 return render_table(
3548 header_line, table, hide_empty=True,
3549 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3550
3551 def render_thumbnails_table(self, info_dict):
3552 thumbnails = list(info_dict.get('thumbnails') or [])
3553 if not thumbnails:
3554 return None
3555 return render_table(
3556 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3557 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3558
3559 def render_subtitles_table(self, video_id, subtitles):
3560 def _row(lang, formats):
3561 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3562 if len(set(names)) == 1:
3563 names = [] if names[0] == 'unknown' else names[:1]
3564 return [lang, ', '.join(names), ', '.join(exts)]
3565
3566 if not subtitles:
3567 return None
3568 return render_table(
3569 self._list_format_headers('Language', 'Name', 'Formats'),
3570 [_row(lang, formats) for lang, formats in subtitles.items()],
3571 hide_empty=True)
3572
3573 def __list_table(self, video_id, name, func, *args):
3574 table = func(*args)
3575 if not table:
3576 self.to_screen(f'{video_id} has no {name}')
3577 return
3578 self.to_screen(f'[info] Available {name} for {video_id}:')
3579 self.to_stdout(table)
3580
3581 def list_formats(self, info_dict):
3582 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3583
3584 def list_thumbnails(self, info_dict):
3585 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3586
3587 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3588 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3589
3590 def urlopen(self, req):
3591 """ Start an HTTP download """
3592 if isinstance(req, str):
3593 req = sanitized_Request(req)
3594 return self._opener.open(req, timeout=self._socket_timeout)
3595
3596 def print_debug_header(self):
3597 if not self.params.get('verbose'):
3598 return
3599
3600 # These imports can be slow. So import them only as needed
3601 from .extractor.extractors import _LAZY_LOADER
3602 from .extractor.extractors import _PLUGIN_CLASSES as plugin_extractors
3603
3604 def get_encoding(stream):
3605 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3606 if not supports_terminal_sequences(stream):
3607 from .utils import WINDOWS_VT_MODE # Must be imported locally
3608 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3609 return ret
3610
3611 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3612 locale.getpreferredencoding(),
3613 sys.getfilesystemencoding(),
3614 self.get_encoding(),
3615 ', '.join(
3616 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3617 if stream is not None and key != 'console')
3618 )
3619
3620 logger = self.params.get('logger')
3621 if logger:
3622 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3623 write_debug(encoding_str)
3624 else:
3625 write_string(f'[debug] {encoding_str}\n', encoding=None)
3626 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3627
3628 source = detect_variant()
3629 write_debug(join_nonempty(
3630 'yt-dlp version', __version__,
3631 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3632 '' if source == 'unknown' else f'({source})',
3633 delim=' '))
3634 if not _LAZY_LOADER:
3635 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3636 write_debug('Lazy loading extractors is forcibly disabled')
3637 else:
3638 write_debug('Lazy loading extractors is disabled')
3639 if plugin_extractors or plugin_postprocessors:
3640 write_debug('Plugins: %s' % [
3641 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3642 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3643 if self.params['compat_opts']:
3644 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3645
3646 if source == 'source':
3647 try:
3648 stdout, _, _ = Popen.run(
3649 ['git', 'rev-parse', '--short', 'HEAD'],
3650 text=True, cwd=os.path.dirname(os.path.abspath(__file__)),
3651 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
3652 if re.fullmatch('[0-9a-f]+', stdout.strip()):
3653 write_debug(f'Git HEAD: {stdout.strip()}')
3654 except Exception:
3655 with contextlib.suppress(Exception):
3656 sys.exc_clear()
3657
3658 def python_implementation():
3659 impl_name = platform.python_implementation()
3660 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3661 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3662 return impl_name
3663
3664 write_debug('Python version %s (%s %s) - %s' % (
3665 platform.python_version(),
3666 python_implementation(),
3667 platform.architecture()[0],
3668 platform_name()))
3669
3670 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3671 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3672 if ffmpeg_features:
3673 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3674
3675 exe_versions['rtmpdump'] = rtmpdump_version()
3676 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3677 exe_str = ', '.join(
3678 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3679 ) or 'none'
3680 write_debug('exe versions: %s' % exe_str)
3681
3682 from .compat.compat_utils import get_package_info
3683 from .dependencies import available_dependencies
3684
3685 write_debug('Optional libraries: %s' % (', '.join(sorted({
3686 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3687 })) or 'none'))
3688
3689 self._setup_opener()
3690 proxy_map = {}
3691 for handler in self._opener.handlers:
3692 if hasattr(handler, 'proxies'):
3693 proxy_map.update(handler.proxies)
3694 write_debug(f'Proxy map: {proxy_map}')
3695
3696 # Not implemented
3697 if False and self.params.get('call_home'):
3698 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3699 write_debug('Public IP address: %s' % ipaddr)
3700 latest_version = self.urlopen(
3701 'https://yt-dl.org/latest/version').read().decode()
3702 if version_tuple(latest_version) > version_tuple(__version__):
3703 self.report_warning(
3704 'You are using an outdated version (newest version: %s)! '
3705 'See https://yt-dl.org/update if you need help updating.' %
3706 latest_version)
3707
3708 def _setup_opener(self):
3709 if hasattr(self, '_opener'):
3710 return
3711 timeout_val = self.params.get('socket_timeout')
3712 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3713
3714 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3715 opts_cookiefile = self.params.get('cookiefile')
3716 opts_proxy = self.params.get('proxy')
3717
3718 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3719
3720 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3721 if opts_proxy is not None:
3722 if opts_proxy == '':
3723 proxies = {}
3724 else:
3725 proxies = {'http': opts_proxy, 'https': opts_proxy}
3726 else:
3727 proxies = compat_urllib_request.getproxies()
3728 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3729 if 'http' in proxies and 'https' not in proxies:
3730 proxies['https'] = proxies['http']
3731 proxy_handler = PerRequestProxyHandler(proxies)
3732
3733 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3734 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3735 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3736 redirect_handler = YoutubeDLRedirectHandler()
3737 data_handler = urllib.request.DataHandler()
3738
3739 # When passing our own FileHandler instance, build_opener won't add the
3740 # default FileHandler and allows us to disable the file protocol, which
3741 # can be used for malicious purposes (see
3742 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3743 file_handler = compat_urllib_request.FileHandler()
3744
3745 def file_open(*args, **kwargs):
3746 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3747 file_handler.file_open = file_open
3748
3749 opener = compat_urllib_request.build_opener(
3750 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3751
3752 # Delete the default user-agent header, which would otherwise apply in
3753 # cases where our custom HTTP handler doesn't come into play
3754 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3755 opener.addheaders = []
3756 self._opener = opener
3757
3758 def encode(self, s):
3759 if isinstance(s, bytes):
3760 return s # Already encoded
3761
3762 try:
3763 return s.encode(self.get_encoding())
3764 except UnicodeEncodeError as err:
3765 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3766 raise
3767
3768 def get_encoding(self):
3769 encoding = self.params.get('encoding')
3770 if encoding is None:
3771 encoding = preferredencoding()
3772 return encoding
3773
3774 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3775 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3776 if overwrite is None:
3777 overwrite = self.params.get('overwrites', True)
3778 if not self.params.get('writeinfojson'):
3779 return False
3780 elif not infofn:
3781 self.write_debug(f'Skipping writing {label} infojson')
3782 return False
3783 elif not self._ensure_dir_exists(infofn):
3784 return None
3785 elif not overwrite and os.path.exists(infofn):
3786 self.to_screen(f'[info] {label.title()} metadata is already present')
3787 return 'exists'
3788
3789 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3790 try:
3791 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3792 return True
3793 except OSError:
3794 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3795 return None
3796
3797 def _write_description(self, label, ie_result, descfn):
3798 ''' Write description and returns True = written, False = skip, None = error '''
3799 if not self.params.get('writedescription'):
3800 return False
3801 elif not descfn:
3802 self.write_debug(f'Skipping writing {label} description')
3803 return False
3804 elif not self._ensure_dir_exists(descfn):
3805 return None
3806 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3807 self.to_screen(f'[info] {label.title()} description is already present')
3808 elif ie_result.get('description') is None:
3809 self.report_warning(f'There\'s no {label} description to write')
3810 return False
3811 else:
3812 try:
3813 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3814 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3815 descfile.write(ie_result['description'])
3816 except OSError:
3817 self.report_error(f'Cannot write {label} description file {descfn}')
3818 return None
3819 return True
3820
3821 def _write_subtitles(self, info_dict, filename):
3822 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3823 ret = []
3824 subtitles = info_dict.get('requested_subtitles')
3825 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3826 # subtitles download errors are already managed as troubles in relevant IE
3827 # that way it will silently go on when used with unsupporting IE
3828 return ret
3829
3830 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3831 if not sub_filename_base:
3832 self.to_screen('[info] Skipping writing video subtitles')
3833 return ret
3834 for sub_lang, sub_info in subtitles.items():
3835 sub_format = sub_info['ext']
3836 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3837 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3838 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3839 if existing_sub:
3840 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3841 sub_info['filepath'] = existing_sub
3842 ret.append((existing_sub, sub_filename_final))
3843 continue
3844
3845 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3846 if sub_info.get('data') is not None:
3847 try:
3848 # Use newline='' to prevent conversion of newline characters
3849 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3850 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3851 subfile.write(sub_info['data'])
3852 sub_info['filepath'] = sub_filename
3853 ret.append((sub_filename, sub_filename_final))
3854 continue
3855 except OSError:
3856 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3857 return None
3858
3859 try:
3860 sub_copy = sub_info.copy()
3861 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3862 self.dl(sub_filename, sub_copy, subtitle=True)
3863 sub_info['filepath'] = sub_filename
3864 ret.append((sub_filename, sub_filename_final))
3865 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3866 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3867 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3868 if not self.params.get('ignoreerrors'):
3869 self.report_error(msg)
3870 raise DownloadError(msg)
3871 self.report_warning(msg)
3872 return ret
3873
3874 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3875 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3876 write_all = self.params.get('write_all_thumbnails', False)
3877 thumbnails, ret = [], []
3878 if write_all or self.params.get('writethumbnail', False):
3879 thumbnails = info_dict.get('thumbnails') or []
3880 multiple = write_all and len(thumbnails) > 1
3881
3882 if thumb_filename_base is None:
3883 thumb_filename_base = filename
3884 if thumbnails and not thumb_filename_base:
3885 self.write_debug(f'Skipping writing {label} thumbnail')
3886 return ret
3887
3888 for idx, t in list(enumerate(thumbnails))[::-1]:
3889 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3890 thumb_display_id = f'{label} thumbnail {t["id"]}'
3891 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3892 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3893
3894 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3895 if existing_thumb:
3896 self.to_screen('[info] %s is already present' % (
3897 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3898 t['filepath'] = existing_thumb
3899 ret.append((existing_thumb, thumb_filename_final))
3900 else:
3901 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3902 try:
3903 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3904 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3905 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3906 shutil.copyfileobj(uf, thumbf)
3907 ret.append((thumb_filename, thumb_filename_final))
3908 t['filepath'] = thumb_filename
3909 except network_exceptions as err:
3910 thumbnails.pop(idx)
3911 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3912 if ret and not write_all:
3913 break
3914 return ret