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