]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
Add support for SSL client certificate authentication (#3435)
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 #!/usr/bin/env python3
2 import collections
3 import contextlib
4 import datetime
5 import errno
6 import fileinput
7 import functools
8 import io
9 import itertools
10 import json
11 import locale
12 import operator
13 import os
14 import platform
15 import random
16 import re
17 import shutil
18 import subprocess
19 import sys
20 import tempfile
21 import time
22 import tokenize
23 import traceback
24 import unicodedata
25 import urllib.request
26 from string import ascii_letters
27
28 from .cache import Cache
29 from .compat import (
30 compat_get_terminal_size,
31 compat_os_name,
32 compat_shlex_quote,
33 compat_str,
34 compat_urllib_error,
35 compat_urllib_request,
36 windows_enable_vt_mode,
37 )
38 from .cookies import load_cookies
39 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
40 from .downloader.rtmp import rtmpdump_version
41 from .extractor import _LAZY_LOADER
42 from .extractor import _PLUGIN_CLASSES as plugin_extractors
43 from .extractor import gen_extractor_classes, get_info_extractor
44 from .extractor.openload import PhantomJSwrapper
45 from .minicurses import format_text
46 from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
47 from .postprocessor import (
48 EmbedThumbnailPP,
49 FFmpegFixupDuplicateMoovPP,
50 FFmpegFixupDurationPP,
51 FFmpegFixupM3u8PP,
52 FFmpegFixupM4aPP,
53 FFmpegFixupStretchedPP,
54 FFmpegFixupTimestampPP,
55 FFmpegMergerPP,
56 FFmpegPostProcessor,
57 MoveFilesAfterDownloadPP,
58 get_postprocessor,
59 )
60 from .update import detect_variant
61 from .utils import (
62 DEFAULT_OUTTMPL,
63 LINK_TEMPLATES,
64 NO_DEFAULT,
65 NUMBER_RE,
66 OUTTMPL_TYPES,
67 POSTPROCESS_WHEN,
68 STR_FORMAT_RE_TMPL,
69 STR_FORMAT_TYPES,
70 ContentTooShortError,
71 DateRange,
72 DownloadCancelled,
73 DownloadError,
74 EntryNotInPlaylist,
75 ExistingVideoReached,
76 ExtractorError,
77 GeoRestrictedError,
78 HEADRequest,
79 InAdvancePagedList,
80 ISO3166Utils,
81 LazyList,
82 MaxDownloadsReached,
83 Namespace,
84 PagedList,
85 PerRequestProxyHandler,
86 Popen,
87 PostProcessingError,
88 ReExtractInfo,
89 RejectedVideoReached,
90 SameFileError,
91 UnavailableVideoError,
92 YoutubeDLCookieProcessor,
93 YoutubeDLHandler,
94 YoutubeDLRedirectHandler,
95 age_restricted,
96 args_to_str,
97 date_from_str,
98 determine_ext,
99 determine_protocol,
100 encode_compat_str,
101 encodeFilename,
102 error_to_compat_str,
103 expand_path,
104 filter_dict,
105 float_or_none,
106 format_bytes,
107 format_decimal_suffix,
108 format_field,
109 formatSeconds,
110 get_domain,
111 int_or_none,
112 iri_to_uri,
113 join_nonempty,
114 locked_file,
115 make_dir,
116 make_HTTPS_handler,
117 merge_headers,
118 network_exceptions,
119 number_of_digits,
120 orderedSet,
121 parse_filesize,
122 platform_name,
123 preferredencoding,
124 prepend_extension,
125 register_socks_protocols,
126 remove_terminal_sequences,
127 render_table,
128 replace_extension,
129 sanitize_filename,
130 sanitize_path,
131 sanitize_url,
132 sanitized_Request,
133 std_headers,
134 str_or_none,
135 strftime_or_none,
136 subtitles_filename,
137 supports_terminal_sequences,
138 timetuple_from_msec,
139 to_high_limit_path,
140 traverse_obj,
141 try_get,
142 url_basename,
143 variadic,
144 version_tuple,
145 write_json_file,
146 write_string,
147 )
148 from .version import RELEASE_GIT_HEAD, __version__
149
150 if compat_os_name == 'nt':
151 import ctypes
152
153
154 class YoutubeDL:
155 """YoutubeDL class.
156
157 YoutubeDL objects are the ones responsible of downloading the
158 actual video file and writing it to disk if the user has requested
159 it, among some other tasks. In most cases there should be one per
160 program. As, given a video URL, the downloader doesn't know how to
161 extract all the needed information, task that InfoExtractors do, it
162 has to pass the URL to one of them.
163
164 For this, YoutubeDL objects have a method that allows
165 InfoExtractors to be registered in a given order. When it is passed
166 a URL, the YoutubeDL object handles it to the first InfoExtractor it
167 finds that reports being able to handle it. The InfoExtractor extracts
168 all the information about the video or videos the URL refers to, and
169 YoutubeDL process the extracted information, possibly using a File
170 Downloader to download the video.
171
172 YoutubeDL objects accept a lot of parameters. In order not to saturate
173 the object constructor with arguments, it receives a dictionary of
174 options instead. These options are available through the params
175 attribute for the InfoExtractors to use. The YoutubeDL also
176 registers itself as the downloader in charge for the InfoExtractors
177 that are added to it, so this is a "mutual registration".
178
179 Available options:
180
181 username: Username for authentication purposes.
182 password: Password for authentication purposes.
183 videopassword: Password for accessing a video.
184 ap_mso: Adobe Pass multiple-system operator identifier.
185 ap_username: Multiple-system operator account username.
186 ap_password: Multiple-system operator account password.
187 usenetrc: Use netrc for authentication instead.
188 verbose: Print additional info to stdout.
189 quiet: Do not print messages to stdout.
190 no_warnings: Do not print out anything for warnings.
191 forceprint: A dict with keys WHEN mapped to a list of templates to
192 print to stdout. The allowed keys are video or any of the
193 items in utils.POSTPROCESS_WHEN.
194 For compatibility, a single list is also accepted
195 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
196 a list of tuples with (template, filename)
197 forceurl: Force printing final URL. (Deprecated)
198 forcetitle: Force printing title. (Deprecated)
199 forceid: Force printing ID. (Deprecated)
200 forcethumbnail: Force printing thumbnail URL. (Deprecated)
201 forcedescription: Force printing description. (Deprecated)
202 forcefilename: Force printing final filename. (Deprecated)
203 forceduration: Force printing duration. (Deprecated)
204 forcejson: Force printing info_dict as JSON.
205 dump_single_json: Force printing the info_dict of the whole playlist
206 (or video) as a single JSON line.
207 force_write_download_archive: Force writing download archive regardless
208 of 'skip_download' or 'simulate'.
209 simulate: Do not download the video files. If unset (or None),
210 simulate only if listsubtitles, listformats or list_thumbnails is used
211 format: Video format code. see "FORMAT SELECTION" for more details.
212 You can also pass a function. The function takes 'ctx' as
213 argument and returns the formats to download.
214 See "build_format_selector" for an implementation
215 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
216 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
217 extracting metadata even if the video is not actually
218 available for download (experimental)
219 format_sort: A list of fields by which to sort the video formats.
220 See "Sorting Formats" for more details.
221 format_sort_force: Force the given format_sort. see "Sorting Formats"
222 for more details.
223 prefer_free_formats: Whether to prefer video formats with free containers
224 over non-free ones of same quality.
225 allow_multiple_video_streams: Allow multiple video streams to be merged
226 into a single file
227 allow_multiple_audio_streams: Allow multiple audio streams to be merged
228 into a single file
229 check_formats Whether to test if the formats are downloadable.
230 Can be True (check all), False (check none),
231 'selected' (check selected formats),
232 or None (check only if requested by extractor)
233 paths: Dictionary of output paths. The allowed keys are 'home'
234 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
235 outtmpl: Dictionary of templates for output names. Allowed keys
236 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
237 For compatibility with youtube-dl, a single string can also be used
238 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
239 restrictfilenames: Do not allow "&" and spaces in file names
240 trim_file_name: Limit length of filename (extension excluded)
241 windowsfilenames: Force the filenames to be windows compatible
242 ignoreerrors: Do not stop on download/postprocessing errors.
243 Can be 'only_download' to ignore only download errors.
244 Default is 'only_download' for CLI, but False for API
245 skip_playlist_after_errors: Number of allowed failures until the rest of
246 the playlist is skipped
247 force_generic_extractor: Force downloader to use the generic extractor
248 overwrites: Overwrite all video and metadata files if True,
249 overwrite only non-video files if None
250 and don't overwrite any file if False
251 For compatibility with youtube-dl,
252 "nooverwrites" may also be used instead
253 playliststart: Playlist item to start at.
254 playlistend: Playlist item to end at.
255 playlist_items: Specific indices of playlist to download.
256 playlistreverse: Download playlist items in reverse order.
257 playlistrandom: Download playlist items in random order.
258 matchtitle: Download only matching titles.
259 rejecttitle: Reject downloads for matching titles.
260 logger: Log messages to a logging.Logger instance.
261 logtostderr: Log messages to stderr instead of stdout.
262 consoletitle: Display progress in console window's titlebar.
263 writedescription: Write the video description to a .description file
264 writeinfojson: Write the video description to a .info.json file
265 clean_infojson: Remove private fields from the infojson
266 getcomments: Extract video comments. This will not be written to disk
267 unless writeinfojson is also given
268 writeannotations: Write the video annotations to a .annotations.xml file
269 writethumbnail: Write the thumbnail image to a file
270 allow_playlist_files: Whether to write playlists' description, infojson etc
271 also to disk when using the 'write*' options
272 write_all_thumbnails: Write all thumbnail formats to files
273 writelink: Write an internet shortcut file, depending on the
274 current platform (.url/.webloc/.desktop)
275 writeurllink: Write a Windows internet shortcut file (.url)
276 writewebloclink: Write a macOS internet shortcut file (.webloc)
277 writedesktoplink: Write a Linux internet shortcut file (.desktop)
278 writesubtitles: Write the video subtitles to a file
279 writeautomaticsub: Write the automatically generated subtitles to a file
280 allsubtitles: Deprecated - Use subtitleslangs = ['all']
281 Downloads all the subtitles of the video
282 (requires writesubtitles or writeautomaticsub)
283 listsubtitles: Lists all available subtitles for the video
284 subtitlesformat: The format code for subtitles
285 subtitleslangs: List of languages of the subtitles to download (can be regex).
286 The list may contain "all" to refer to all the available
287 subtitles. The language can be prefixed with a "-" to
288 exclude it from the requested languages. Eg: ['all', '-live_chat']
289 keepvideo: Keep the video file after post-processing
290 daterange: A DateRange object, download only if the upload_date is in the range.
291 skip_download: Skip the actual download of the video file
292 cachedir: Location of the cache files in the filesystem.
293 False to disable filesystem cache.
294 noplaylist: Download single video instead of a playlist if in doubt.
295 age_limit: An integer representing the user's age in years.
296 Unsuitable videos for the given age are skipped.
297 min_views: An integer representing the minimum view count the video
298 must have in order to not be skipped.
299 Videos without view count information are always
300 downloaded. None for no limit.
301 max_views: An integer representing the maximum view count.
302 Videos that are more popular than that are not
303 downloaded.
304 Videos without view count information are always
305 downloaded. None for no limit.
306 download_archive: File name of a file where all downloads are recorded.
307 Videos already present in the file are not downloaded
308 again.
309 break_on_existing: Stop the download process after attempting to download a
310 file that is in the archive.
311 break_on_reject: Stop the download process when encountering a video that
312 has been filtered out.
313 break_per_url: Whether break_on_reject and break_on_existing
314 should act on each input URL as opposed to for the entire queue
315 cookiefile: File name where cookies should be read from and dumped to
316 cookiesfrombrowser: A tuple containing the name of the browser, the profile
317 name/pathfrom where cookies are loaded, and the name of the
318 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
319 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
320 support RFC 5746 secure renegotiation
321 nocheckcertificate: Do not verify SSL certificates
322 client_certificate: Path to client certificate file in PEM format. May include the private key
323 client_certificate_key: Path to private key file for client certificate
324 client_certificate_password: Password for client certificate private key, if encrypted.
325 If not provided and the key is encrypted, yt-dlp will ask interactively
326 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
327 At the moment, this is only supported by YouTube.
328 http_headers: A dictionary of custom headers to be used for all requests
329 proxy: URL of the proxy server to use
330 geo_verification_proxy: URL of the proxy to use for IP address verification
331 on geo-restricted sites.
332 socket_timeout: Time to wait for unresponsive hosts, in seconds
333 bidi_workaround: Work around buggy terminals without bidirectional text
334 support, using fridibi
335 debug_printtraffic:Print out sent and received HTTP traffic
336 include_ads: Download ads as well (deprecated)
337 default_search: Prepend this string if an input url is not valid.
338 'auto' for elaborate guessing
339 encoding: Use this encoding instead of the system-specified.
340 extract_flat: Do not resolve URLs, return the immediate result.
341 Pass in 'in_playlist' to only show this behavior for
342 playlist items.
343 wait_for_video: If given, wait for scheduled streams to become available.
344 The value should be a tuple containing the range
345 (min_secs, max_secs) to wait between retries
346 postprocessors: A list of dictionaries, each with an entry
347 * key: The name of the postprocessor. See
348 yt_dlp/postprocessor/__init__.py for a list.
349 * when: When to run the postprocessor. Allowed values are
350 the entries of utils.POSTPROCESS_WHEN
351 Assumed to be 'post_process' if not given
352 post_hooks: Deprecated - Register a custom postprocessor instead
353 A list of functions that get called as the final step
354 for each video file, after all postprocessors have been
355 called. The filename will be passed as the only argument.
356 progress_hooks: A list of functions that get called on download
357 progress, with a dictionary with the entries
358 * status: One of "downloading", "error", or "finished".
359 Check this first and ignore unknown values.
360 * info_dict: The extracted info_dict
361
362 If status is one of "downloading", or "finished", the
363 following properties may also be present:
364 * filename: The final filename (always present)
365 * tmpfilename: The filename we're currently writing to
366 * downloaded_bytes: Bytes on disk
367 * total_bytes: Size of the whole file, None if unknown
368 * total_bytes_estimate: Guess of the eventual file size,
369 None if unavailable.
370 * elapsed: The number of seconds since download started.
371 * eta: The estimated time in seconds, None if unknown
372 * speed: The download speed in bytes/second, None if
373 unknown
374 * fragment_index: The counter of the currently
375 downloaded video fragment.
376 * fragment_count: The number of fragments (= individual
377 files that will be merged)
378
379 Progress hooks are guaranteed to be called at least once
380 (with status "finished") if the download is successful.
381 postprocessor_hooks: A list of functions that get called on postprocessing
382 progress, with a dictionary with the entries
383 * status: One of "started", "processing", or "finished".
384 Check this first and ignore unknown values.
385 * postprocessor: Name of the postprocessor
386 * info_dict: The extracted info_dict
387
388 Progress hooks are guaranteed to be called at least twice
389 (with status "started" and "finished") if the processing is successful.
390 merge_output_format: Extension to use when merging formats.
391 final_ext: Expected final extension; used to detect when the file was
392 already downloaded and converted
393 fixup: Automatically correct known faults of the file.
394 One of:
395 - "never": do nothing
396 - "warn": only emit a warning
397 - "detect_or_warn": check whether we can do anything
398 about it, warn otherwise (default)
399 source_address: Client-side IP address to bind to.
400 call_home: Boolean, true iff we are allowed to contact the
401 yt-dlp servers for debugging. (BROKEN)
402 sleep_interval_requests: Number of seconds to sleep between requests
403 during extraction
404 sleep_interval: Number of seconds to sleep before each download when
405 used alone or a lower bound of a range for randomized
406 sleep before each download (minimum possible number
407 of seconds to sleep) when used along with
408 max_sleep_interval.
409 max_sleep_interval:Upper bound of a range for randomized sleep before each
410 download (maximum possible number of seconds to sleep).
411 Must only be used along with sleep_interval.
412 Actual sleep time will be a random float from range
413 [sleep_interval; max_sleep_interval].
414 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
415 listformats: Print an overview of available video formats and exit.
416 list_thumbnails: Print a table of all thumbnails and exit.
417 match_filter: A function that gets called for every video with the signature
418 (info_dict, *, incomplete: bool) -> Optional[str]
419 For backward compatibility with youtube-dl, the signature
420 (info_dict) -> Optional[str] is also allowed.
421 - If it returns a message, the video is ignored.
422 - If it returns None, the video is downloaded.
423 - If it returns utils.NO_DEFAULT, the user is interactively
424 asked whether to download the video.
425 match_filter_func in utils.py is one example for this.
426 no_color: Do not emit color codes in output.
427 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
428 HTTP header
429 geo_bypass_country:
430 Two-letter ISO 3166-2 country code that will be used for
431 explicit geographic restriction bypassing via faking
432 X-Forwarded-For HTTP header
433 geo_bypass_ip_block:
434 IP range in CIDR notation that will be used similarly to
435 geo_bypass_country
436
437 The following options determine which downloader is picked:
438 external_downloader: A dictionary of protocol keys and the executable of the
439 external downloader to use for it. The allowed protocols
440 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
441 Set the value to 'native' to use the native downloader
442 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
443 or {'m3u8': 'ffmpeg'} instead.
444 Use the native HLS downloader instead of ffmpeg/avconv
445 if True, otherwise use ffmpeg/avconv if False, otherwise
446 use downloader suggested by extractor if None.
447 compat_opts: Compatibility options. See "Differences in default behavior".
448 The following options do not work when used through the API:
449 filename, abort-on-error, multistreams, no-live-chat, format-sort
450 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
451 Refer __init__.py for their implementation
452 progress_template: Dictionary of templates for progress outputs.
453 Allowed keys are 'download', 'postprocess',
454 'download-title' (console title) and 'postprocess-title'.
455 The template is mapped on a dictionary with keys 'progress' and 'info'
456
457 The following parameters are not used by YoutubeDL itself, they are used by
458 the downloader (see yt_dlp/downloader/common.py):
459 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
460 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
461 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
462 external_downloader_args, concurrent_fragment_downloads.
463
464 The following options are used by the post processors:
465 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
466 otherwise prefer ffmpeg. (avconv support is deprecated)
467 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
468 to the binary or its containing directory.
469 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
470 and a list of additional command-line arguments for the
471 postprocessor/executable. The dict can also have "PP+EXE" keys
472 which are used when the given exe is used by the given PP.
473 Use 'default' as the name for arguments to passed to all PP
474 For compatibility with youtube-dl, a single list of args
475 can also be used
476
477 The following options are used by the extractors:
478 extractor_retries: Number of times to retry for known errors
479 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
480 hls_split_discontinuity: Split HLS playlists to different formats at
481 discontinuities such as ad breaks (default: False)
482 extractor_args: A dictionary of arguments to be passed to the extractors.
483 See "EXTRACTOR ARGUMENTS" for details.
484 Eg: {'youtube': {'skip': ['dash', 'hls']}}
485 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
486 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
487 If True (default), DASH manifests and related
488 data will be downloaded and processed by extractor.
489 You can reduce network I/O by disabling it if you don't
490 care about DASH. (only for youtube)
491 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
492 If True (default), HLS manifests and related
493 data will be downloaded and processed by extractor.
494 You can reduce network I/O by disabling it if you don't
495 care about HLS. (only for youtube)
496 """
497
498 _NUMERIC_FIELDS = {
499 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
500 'timestamp', 'release_timestamp',
501 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
502 'average_rating', 'comment_count', 'age_limit',
503 'start_time', 'end_time',
504 'chapter_number', 'season_number', 'episode_number',
505 'track_number', 'disc_number', 'release_year',
506 }
507
508 _format_fields = {
509 # NB: Keep in sync with the docstring of extractor/common.py
510 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
511 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
512 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
513 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
514 'preference', 'language', 'language_preference', 'quality', 'source_preference',
515 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
516 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
517 }
518 _format_selection_exts = {
519 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
520 'video': {'mp4', 'flv', 'webm', '3gp'},
521 'storyboards': {'mhtml'},
522 }
523
524 def __init__(self, params=None, auto_init=True):
525 """Create a FileDownloader object with the given options.
526 @param auto_init Whether to load the default extractors and print header (if verbose).
527 Set to 'no_verbose_header' to not print the header
528 """
529 if params is None:
530 params = {}
531 self.params = params
532 self._ies = {}
533 self._ies_instances = {}
534 self._pps = {k: [] for k in POSTPROCESS_WHEN}
535 self._printed_messages = set()
536 self._first_webpage_request = True
537 self._post_hooks = []
538 self._progress_hooks = []
539 self._postprocessor_hooks = []
540 self._download_retcode = 0
541 self._num_downloads = 0
542 self._num_videos = 0
543 self._playlist_level = 0
544 self._playlist_urls = set()
545 self.cache = Cache(self)
546
547 windows_enable_vt_mode()
548 self._out_files = {
549 'error': sys.stderr,
550 'print': sys.stderr if self.params.get('logtostderr') else sys.stdout,
551 'console': None if compat_os_name == 'nt' else next(
552 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
553 }
554 self._out_files['screen'] = sys.stderr if self.params.get('quiet') else self._out_files['print']
555 self._allow_colors = {
556 type_: not self.params.get('no_color') and supports_terminal_sequences(self._out_files[type_])
557 for type_ in ('screen', 'error')
558 }
559
560 if sys.version_info < (3, 6):
561 self.report_warning(
562 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
563
564 if self.params.get('allow_unplayable_formats'):
565 self.report_warning(
566 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
567 'This is a developer option intended for debugging. \n'
568 ' If you experience any issues while using this option, '
569 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
570
571 def check_deprecated(param, option, suggestion):
572 if self.params.get(param) is not None:
573 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
574 return True
575 return False
576
577 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
578 if self.params.get('geo_verification_proxy') is None:
579 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
580
581 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
582 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
583 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
584
585 for msg in self.params.get('_warnings', []):
586 self.report_warning(msg)
587 for msg in self.params.get('_deprecation_warnings', []):
588 self.deprecation_warning(msg)
589
590 if 'list-formats' in self.params.get('compat_opts', []):
591 self.params['listformats_table'] = False
592
593 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
594 # nooverwrites was unnecessarily changed to overwrites
595 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
596 # This ensures compatibility with both keys
597 self.params['overwrites'] = not self.params['nooverwrites']
598 elif self.params.get('overwrites') is None:
599 self.params.pop('overwrites', None)
600 else:
601 self.params['nooverwrites'] = not self.params['overwrites']
602
603 self.params.setdefault('forceprint', {})
604 self.params.setdefault('print_to_file', {})
605
606 # Compatibility with older syntax
607 if not isinstance(params['forceprint'], dict):
608 self.params['forceprint'] = {'video': params['forceprint']}
609
610 if self.params.get('bidi_workaround', False):
611 try:
612 import pty
613 master, slave = pty.openpty()
614 width = compat_get_terminal_size().columns
615 if width is None:
616 width_args = []
617 else:
618 width_args = ['-w', str(width)]
619 sp_kwargs = dict(
620 stdin=subprocess.PIPE,
621 stdout=slave,
622 stderr=self._out_files['error'])
623 try:
624 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
625 except OSError:
626 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
627 self._output_channel = os.fdopen(master, 'rb')
628 except OSError as ose:
629 if ose.errno == errno.ENOENT:
630 self.report_warning(
631 'Could not find fribidi executable, ignoring --bidi-workaround. '
632 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
633 else:
634 raise
635
636 if auto_init:
637 if auto_init != 'no_verbose_header':
638 self.print_debug_header()
639 self.add_default_info_extractors()
640
641 if (sys.platform != 'win32'
642 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
643 and not self.params.get('restrictfilenames', False)):
644 # Unicode filesystem API will throw errors (#1474, #13027)
645 self.report_warning(
646 'Assuming --restrict-filenames since file system encoding '
647 'cannot encode all characters. '
648 'Set the LC_ALL environment variable to fix this.')
649 self.params['restrictfilenames'] = True
650
651 self.outtmpl_dict = self.parse_outtmpl()
652
653 # Creating format selector here allows us to catch syntax errors before the extraction
654 self.format_selector = (
655 self.params.get('format') if self.params.get('format') in (None, '-')
656 else self.params['format'] if callable(self.params['format'])
657 else self.build_format_selector(self.params['format']))
658
659 # Set http_headers defaults according to std_headers
660 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
661
662 hooks = {
663 'post_hooks': self.add_post_hook,
664 'progress_hooks': self.add_progress_hook,
665 'postprocessor_hooks': self.add_postprocessor_hook,
666 }
667 for opt, fn in hooks.items():
668 for ph in self.params.get(opt, []):
669 fn(ph)
670
671 for pp_def_raw in self.params.get('postprocessors', []):
672 pp_def = dict(pp_def_raw)
673 when = pp_def.pop('when', 'post_process')
674 self.add_post_processor(
675 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
676 when=when)
677
678 self._setup_opener()
679 register_socks_protocols()
680
681 def preload_download_archive(fn):
682 """Preload the archive, if any is specified"""
683 if fn is None:
684 return False
685 self.write_debug(f'Loading archive file {fn!r}')
686 try:
687 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
688 for line in archive_file:
689 self.archive.add(line.strip())
690 except OSError as ioe:
691 if ioe.errno != errno.ENOENT:
692 raise
693 return False
694 return True
695
696 self.archive = set()
697 preload_download_archive(self.params.get('download_archive'))
698
699 def warn_if_short_id(self, argv):
700 # short YouTube ID starting with dash?
701 idxs = [
702 i for i, a in enumerate(argv)
703 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
704 if idxs:
705 correct_argv = (
706 ['yt-dlp']
707 + [a for i, a in enumerate(argv) if i not in idxs]
708 + ['--'] + [argv[i] for i in idxs]
709 )
710 self.report_warning(
711 'Long argument string detected. '
712 'Use -- to separate parameters and URLs, like this:\n%s' %
713 args_to_str(correct_argv))
714
715 def add_info_extractor(self, ie):
716 """Add an InfoExtractor object to the end of the list."""
717 ie_key = ie.ie_key()
718 self._ies[ie_key] = ie
719 if not isinstance(ie, type):
720 self._ies_instances[ie_key] = ie
721 ie.set_downloader(self)
722
723 def _get_info_extractor_class(self, ie_key):
724 ie = self._ies.get(ie_key)
725 if ie is None:
726 ie = get_info_extractor(ie_key)
727 self.add_info_extractor(ie)
728 return ie
729
730 def get_info_extractor(self, ie_key):
731 """
732 Get an instance of an IE with name ie_key, it will try to get one from
733 the _ies list, if there's no instance it will create a new one and add
734 it to the extractor list.
735 """
736 ie = self._ies_instances.get(ie_key)
737 if ie is None:
738 ie = get_info_extractor(ie_key)()
739 self.add_info_extractor(ie)
740 return ie
741
742 def add_default_info_extractors(self):
743 """
744 Add the InfoExtractors returned by gen_extractors to the end of the list
745 """
746 for ie in gen_extractor_classes():
747 self.add_info_extractor(ie)
748
749 def add_post_processor(self, pp, when='post_process'):
750 """Add a PostProcessor object to the end of the chain."""
751 self._pps[when].append(pp)
752 pp.set_downloader(self)
753
754 def add_post_hook(self, ph):
755 """Add the post hook"""
756 self._post_hooks.append(ph)
757
758 def add_progress_hook(self, ph):
759 """Add the download progress hook"""
760 self._progress_hooks.append(ph)
761
762 def add_postprocessor_hook(self, ph):
763 """Add the postprocessing progress hook"""
764 self._postprocessor_hooks.append(ph)
765 for pps in self._pps.values():
766 for pp in pps:
767 pp.add_progress_hook(ph)
768
769 def _bidi_workaround(self, message):
770 if not hasattr(self, '_output_channel'):
771 return message
772
773 assert hasattr(self, '_output_process')
774 assert isinstance(message, compat_str)
775 line_count = message.count('\n') + 1
776 self._output_process.stdin.write((message + '\n').encode('utf-8'))
777 self._output_process.stdin.flush()
778 res = ''.join(self._output_channel.readline().decode('utf-8')
779 for _ in range(line_count))
780 return res[:-len('\n')]
781
782 def _write_string(self, message, out=None, only_once=False):
783 if only_once:
784 if message in self._printed_messages:
785 return
786 self._printed_messages.add(message)
787 write_string(message, out=out, encoding=self.params.get('encoding'))
788
789 def to_stdout(self, message, skip_eol=False, quiet=None):
790 """Print message to stdout"""
791 if quiet is not None:
792 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
793 self._write_string(
794 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
795 self._out_files['print'])
796
797 def to_screen(self, message, skip_eol=False, quiet=None):
798 """Print message to screen if not in quiet mode"""
799 if self.params.get('logger'):
800 self.params['logger'].debug(message)
801 return
802 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
803 return
804 self._write_string(
805 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
806 self._out_files['screen'])
807
808 def to_stderr(self, message, only_once=False):
809 """Print message to stderr"""
810 assert isinstance(message, compat_str)
811 if self.params.get('logger'):
812 self.params['logger'].error(message)
813 else:
814 self._write_string('%s\n' % self._bidi_workaround(message), self._out_files['error'], only_once=only_once)
815
816 def _send_console_code(self, code):
817 if compat_os_name == 'nt' or not self._out_files['console']:
818 return
819 self._write_string(code, self._out_files['console'])
820
821 def to_console_title(self, message):
822 if not self.params.get('consoletitle', False):
823 return
824 message = remove_terminal_sequences(message)
825 if compat_os_name == 'nt':
826 if ctypes.windll.kernel32.GetConsoleWindow():
827 # c_wchar_p() might not be necessary if `message` is
828 # already of type unicode()
829 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
830 else:
831 self._send_console_code(f'\033]0;{message}\007')
832
833 def save_console_title(self):
834 if not self.params.get('consoletitle') or self.params.get('simulate'):
835 return
836 self._send_console_code('\033[22;0t') # Save the title on stack
837
838 def restore_console_title(self):
839 if not self.params.get('consoletitle') or self.params.get('simulate'):
840 return
841 self._send_console_code('\033[23;0t') # Restore the title from stack
842
843 def __enter__(self):
844 self.save_console_title()
845 return self
846
847 def __exit__(self, *args):
848 self.restore_console_title()
849
850 if self.params.get('cookiefile') is not None:
851 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
852
853 def trouble(self, message=None, tb=None, is_error=True):
854 """Determine action to take when a download problem appears.
855
856 Depending on if the downloader has been configured to ignore
857 download errors or not, this method may throw an exception or
858 not when errors are found, after printing the message.
859
860 @param tb If given, is additional traceback information
861 @param is_error Whether to raise error according to ignorerrors
862 """
863 if message is not None:
864 self.to_stderr(message)
865 if self.params.get('verbose'):
866 if tb is None:
867 if sys.exc_info()[0]: # if .trouble has been called from an except block
868 tb = ''
869 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
870 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
871 tb += encode_compat_str(traceback.format_exc())
872 else:
873 tb_data = traceback.format_list(traceback.extract_stack())
874 tb = ''.join(tb_data)
875 if tb:
876 self.to_stderr(tb)
877 if not is_error:
878 return
879 if not self.params.get('ignoreerrors'):
880 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
881 exc_info = sys.exc_info()[1].exc_info
882 else:
883 exc_info = sys.exc_info()
884 raise DownloadError(message, exc_info)
885 self._download_retcode = 1
886
887 Styles = Namespace(
888 HEADERS='yellow',
889 EMPHASIS='light blue',
890 FILENAME='green',
891 ID='green',
892 DELIM='blue',
893 ERROR='red',
894 WARNING='yellow',
895 SUPPRESS='light black',
896 )
897
898 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
899 text = str(text)
900 if test_encoding:
901 original_text = text
902 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
903 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
904 text = text.encode(encoding, 'ignore').decode(encoding)
905 if fallback is not None and text != original_text:
906 text = fallback
907 return format_text(text, f) if allow_colors else text if fallback is None else fallback
908
909 def _format_screen(self, *args, **kwargs):
910 return self._format_text(
911 self._out_files['screen'], self._allow_colors['screen'], *args, **kwargs)
912
913 def _format_err(self, *args, **kwargs):
914 return self._format_text(
915 self._out_files['error'], self._allow_colors['error'], *args, **kwargs)
916
917 def report_warning(self, message, only_once=False):
918 '''
919 Print the message to stderr, it will be prefixed with 'WARNING:'
920 If stderr is a tty file the 'WARNING:' will be colored
921 '''
922 if self.params.get('logger') is not None:
923 self.params['logger'].warning(message)
924 else:
925 if self.params.get('no_warnings'):
926 return
927 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
928
929 def deprecation_warning(self, message):
930 if self.params.get('logger') is not None:
931 self.params['logger'].warning(f'DeprecationWarning: {message}')
932 else:
933 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
934
935 def report_error(self, message, *args, **kwargs):
936 '''
937 Do the same as trouble, but prefixes the message with 'ERROR:', colored
938 in red if stderr is a tty file.
939 '''
940 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
941
942 def write_debug(self, message, only_once=False):
943 '''Log debug message or Print message to stderr'''
944 if not self.params.get('verbose', False):
945 return
946 message = '[debug] %s' % message
947 if self.params.get('logger'):
948 self.params['logger'].debug(message)
949 else:
950 self.to_stderr(message, only_once)
951
952 def report_file_already_downloaded(self, file_name):
953 """Report file has already been fully downloaded."""
954 try:
955 self.to_screen('[download] %s has already been downloaded' % file_name)
956 except UnicodeEncodeError:
957 self.to_screen('[download] The file has already been downloaded')
958
959 def report_file_delete(self, file_name):
960 """Report that existing file will be deleted."""
961 try:
962 self.to_screen('Deleting existing file %s' % file_name)
963 except UnicodeEncodeError:
964 self.to_screen('Deleting existing file')
965
966 def raise_no_formats(self, info, forced=False, *, msg=None):
967 has_drm = info.get('_has_drm')
968 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
969 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
970 if forced or not ignored:
971 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
972 expected=has_drm or ignored or expected)
973 else:
974 self.report_warning(msg)
975
976 def parse_outtmpl(self):
977 outtmpl_dict = self.params.get('outtmpl', {})
978 if not isinstance(outtmpl_dict, dict):
979 outtmpl_dict = {'default': outtmpl_dict}
980 # Remove spaces in the default template
981 if self.params.get('restrictfilenames'):
982 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
983 else:
984 sanitize = lambda x: x
985 outtmpl_dict.update({
986 k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items()
987 if outtmpl_dict.get(k) is None})
988 for _, val in outtmpl_dict.items():
989 if isinstance(val, bytes):
990 self.report_warning('Parameter outtmpl is bytes, but should be a unicode string')
991 return outtmpl_dict
992
993 def get_output_path(self, dir_type='', filename=None):
994 paths = self.params.get('paths', {})
995 assert isinstance(paths, dict)
996 path = os.path.join(
997 expand_path(paths.get('home', '').strip()),
998 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
999 filename or '')
1000 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1001
1002 @staticmethod
1003 def _outtmpl_expandpath(outtmpl):
1004 # expand_path translates '%%' into '%' and '$$' into '$'
1005 # correspondingly that is not what we want since we need to keep
1006 # '%%' intact for template dict substitution step. Working around
1007 # with boundary-alike separator hack.
1008 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1009 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1010
1011 # outtmpl should be expand_path'ed before template dict substitution
1012 # because meta fields may contain env variables we don't want to
1013 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1014 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1015 return expand_path(outtmpl).replace(sep, '')
1016
1017 @staticmethod
1018 def escape_outtmpl(outtmpl):
1019 ''' Escape any remaining strings like %s, %abc% etc. '''
1020 return re.sub(
1021 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1022 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1023 outtmpl)
1024
1025 @classmethod
1026 def validate_outtmpl(cls, outtmpl):
1027 ''' @return None or Exception object '''
1028 outtmpl = re.sub(
1029 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1030 lambda mobj: f'{mobj.group(0)[:-1]}s',
1031 cls._outtmpl_expandpath(outtmpl))
1032 try:
1033 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1034 return None
1035 except ValueError as err:
1036 return err
1037
1038 @staticmethod
1039 def _copy_infodict(info_dict):
1040 info_dict = dict(info_dict)
1041 info_dict.pop('__postprocessors', None)
1042 return info_dict
1043
1044 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1045 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1046 @param sanitize Whether to sanitize the output as a filename.
1047 For backward compatibility, a function can also be passed
1048 """
1049
1050 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1051
1052 info_dict = self._copy_infodict(info_dict)
1053 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1054 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1055 if info_dict.get('duration', None) is not None
1056 else None)
1057 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1058 info_dict['video_autonumber'] = self._num_videos
1059 if info_dict.get('resolution') is None:
1060 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1061
1062 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1063 # of %(field)s to %(field)0Nd for backward compatibility
1064 field_size_compat_map = {
1065 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1066 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1067 'autonumber': self.params.get('autonumber_size') or 5,
1068 }
1069
1070 TMPL_DICT = {}
1071 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1072 MATH_FUNCTIONS = {
1073 '+': float.__add__,
1074 '-': float.__sub__,
1075 }
1076 # Field is of the form key1.key2...
1077 # where keys (except first) can be string, int or slice
1078 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1079 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1080 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1081 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1082 (?P<negate>-)?
1083 (?P<fields>{FIELD_RE})
1084 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1085 (?:>(?P<strf_format>.+?))?
1086 (?P<remaining>
1087 (?P<alternate>(?<!\\),[^|&)]+)?
1088 (?:&(?P<replacement>.*?))?
1089 (?:\|(?P<default>.*?))?
1090 )$''')
1091
1092 def _traverse_infodict(k):
1093 k = k.split('.')
1094 if k[0] == '':
1095 k.pop(0)
1096 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1097
1098 def get_value(mdict):
1099 # Object traversal
1100 value = _traverse_infodict(mdict['fields'])
1101 # Negative
1102 if mdict['negate']:
1103 value = float_or_none(value)
1104 if value is not None:
1105 value *= -1
1106 # Do maths
1107 offset_key = mdict['maths']
1108 if offset_key:
1109 value = float_or_none(value)
1110 operator = None
1111 while offset_key:
1112 item = re.match(
1113 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1114 offset_key).group(0)
1115 offset_key = offset_key[len(item):]
1116 if operator is None:
1117 operator = MATH_FUNCTIONS[item]
1118 continue
1119 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1120 offset = float_or_none(item)
1121 if offset is None:
1122 offset = float_or_none(_traverse_infodict(item))
1123 try:
1124 value = operator(value, multiplier * offset)
1125 except (TypeError, ZeroDivisionError):
1126 return None
1127 operator = None
1128 # Datetime formatting
1129 if mdict['strf_format']:
1130 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1131
1132 return value
1133
1134 na = self.params.get('outtmpl_na_placeholder', 'NA')
1135
1136 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1137 return sanitize_filename(str(value), restricted=restricted, is_id=(
1138 bool(re.search(r'(^|[_.])id(\.|$)', key))
1139 if 'filename-sanitization' in self.params.get('compat_opts', [])
1140 else NO_DEFAULT))
1141
1142 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1143 sanitize = bool(sanitize)
1144
1145 def _dumpjson_default(obj):
1146 if isinstance(obj, (set, LazyList)):
1147 return list(obj)
1148 return repr(obj)
1149
1150 def create_key(outer_mobj):
1151 if not outer_mobj.group('has_key'):
1152 return outer_mobj.group(0)
1153 key = outer_mobj.group('key')
1154 mobj = re.match(INTERNAL_FORMAT_RE, key)
1155 initial_field = mobj.group('fields') if mobj else ''
1156 value, replacement, default = None, None, na
1157 while mobj:
1158 mobj = mobj.groupdict()
1159 default = mobj['default'] if mobj['default'] is not None else default
1160 value = get_value(mobj)
1161 replacement = mobj['replacement']
1162 if value is None and mobj['alternate']:
1163 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1164 else:
1165 break
1166
1167 fmt = outer_mobj.group('format')
1168 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1169 fmt = f'0{field_size_compat_map[key]:d}d'
1170
1171 value = default if value is None else value if replacement is None else replacement
1172
1173 flags = outer_mobj.group('conversion') or ''
1174 str_fmt = f'{fmt[:-1]}s'
1175 if fmt[-1] == 'l': # list
1176 delim = '\n' if '#' in flags else ', '
1177 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1178 elif fmt[-1] == 'j': # json
1179 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1180 elif fmt[-1] == 'q': # quoted
1181 value = map(str, variadic(value) if '#' in flags else [value])
1182 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1183 elif fmt[-1] == 'B': # bytes
1184 value = f'%{str_fmt}'.encode() % str(value).encode('utf-8')
1185 value, fmt = value.decode('utf-8', 'ignore'), 's'
1186 elif fmt[-1] == 'U': # unicode normalized
1187 value, fmt = unicodedata.normalize(
1188 # "+" = compatibility equivalence, "#" = NFD
1189 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1190 value), str_fmt
1191 elif fmt[-1] == 'D': # decimal suffix
1192 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1193 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1194 factor=1024 if '#' in flags else 1000)
1195 elif fmt[-1] == 'S': # filename sanitization
1196 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1197 elif fmt[-1] == 'c':
1198 if value:
1199 value = str(value)[0]
1200 else:
1201 fmt = str_fmt
1202 elif fmt[-1] not in 'rs': # numeric
1203 value = float_or_none(value)
1204 if value is None:
1205 value, fmt = default, 's'
1206
1207 if sanitize:
1208 if fmt[-1] == 'r':
1209 # If value is an object, sanitize might convert it to a string
1210 # So we convert it to repr first
1211 value, fmt = repr(value), str_fmt
1212 if fmt[-1] in 'csr':
1213 value = sanitizer(initial_field, value)
1214
1215 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1216 TMPL_DICT[key] = value
1217 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1218
1219 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1220
1221 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1222 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1223 return self.escape_outtmpl(outtmpl) % info_dict
1224
1225 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1226 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1227 if outtmpl is None:
1228 outtmpl = self.outtmpl_dict.get(tmpl_type or 'default', self.outtmpl_dict['default'])
1229 try:
1230 outtmpl = self._outtmpl_expandpath(outtmpl)
1231 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1232 if not filename:
1233 return None
1234
1235 if tmpl_type in ('', 'temp'):
1236 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1237 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1238 filename = replace_extension(filename, ext, final_ext)
1239 elif tmpl_type:
1240 force_ext = OUTTMPL_TYPES[tmpl_type]
1241 if force_ext:
1242 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1243
1244 # https://github.com/blackjack4494/youtube-dlc/issues/85
1245 trim_file_name = self.params.get('trim_file_name', False)
1246 if trim_file_name:
1247 no_ext, *ext = filename.rsplit('.', 2)
1248 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1249
1250 return filename
1251 except ValueError as err:
1252 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1253 return None
1254
1255 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1256 """Generate the output filename"""
1257 if outtmpl:
1258 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1259 dir_type = None
1260 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1261 if not filename and dir_type not in ('', 'temp'):
1262 return ''
1263
1264 if warn:
1265 if not self.params.get('paths'):
1266 pass
1267 elif filename == '-':
1268 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1269 elif os.path.isabs(filename):
1270 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1271 if filename == '-' or not filename:
1272 return filename
1273
1274 return self.get_output_path(dir_type, filename)
1275
1276 def _match_entry(self, info_dict, incomplete=False, silent=False):
1277 """ Returns None if the file should be downloaded """
1278
1279 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1280
1281 def check_filter():
1282 if 'title' in info_dict:
1283 # This can happen when we're just evaluating the playlist
1284 title = info_dict['title']
1285 matchtitle = self.params.get('matchtitle', False)
1286 if matchtitle:
1287 if not re.search(matchtitle, title, re.IGNORECASE):
1288 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1289 rejecttitle = self.params.get('rejecttitle', False)
1290 if rejecttitle:
1291 if re.search(rejecttitle, title, re.IGNORECASE):
1292 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1293 date = info_dict.get('upload_date')
1294 if date is not None:
1295 dateRange = self.params.get('daterange', DateRange())
1296 if date not in dateRange:
1297 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1298 view_count = info_dict.get('view_count')
1299 if view_count is not None:
1300 min_views = self.params.get('min_views')
1301 if min_views is not None and view_count < min_views:
1302 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1303 max_views = self.params.get('max_views')
1304 if max_views is not None and view_count > max_views:
1305 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1306 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1307 return 'Skipping "%s" because it is age restricted' % video_title
1308
1309 match_filter = self.params.get('match_filter')
1310 if match_filter is not None:
1311 try:
1312 ret = match_filter(info_dict, incomplete=incomplete)
1313 except TypeError:
1314 # For backward compatibility
1315 ret = None if incomplete else match_filter(info_dict)
1316 if ret is NO_DEFAULT:
1317 while True:
1318 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1319 reply = input(self._format_screen(
1320 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1321 if reply in {'y', ''}:
1322 return None
1323 elif reply == 'n':
1324 return f'Skipping {video_title}'
1325 return True
1326 elif ret is not None:
1327 return ret
1328 return None
1329
1330 if self.in_download_archive(info_dict):
1331 reason = '%s has already been recorded in the archive' % video_title
1332 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1333 else:
1334 reason = check_filter()
1335 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1336 if reason is not None:
1337 if not silent:
1338 self.to_screen('[download] ' + reason)
1339 if self.params.get(break_opt, False):
1340 raise break_err()
1341 return reason
1342
1343 @staticmethod
1344 def add_extra_info(info_dict, extra_info):
1345 '''Set the keys from extra_info in info dict if they are missing'''
1346 for key, value in extra_info.items():
1347 info_dict.setdefault(key, value)
1348
1349 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1350 process=True, force_generic_extractor=False):
1351 """
1352 Return a list with a dictionary for each video extracted.
1353
1354 Arguments:
1355 url -- URL to extract
1356
1357 Keyword arguments:
1358 download -- whether to download videos during extraction
1359 ie_key -- extractor key hint
1360 extra_info -- dictionary containing the extra values to add to each result
1361 process -- whether to resolve all unresolved references (URLs, playlist items),
1362 must be True for download to work.
1363 force_generic_extractor -- force using the generic extractor
1364 """
1365
1366 if extra_info is None:
1367 extra_info = {}
1368
1369 if not ie_key and force_generic_extractor:
1370 ie_key = 'Generic'
1371
1372 if ie_key:
1373 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1374 else:
1375 ies = self._ies
1376
1377 for ie_key, ie in ies.items():
1378 if not ie.suitable(url):
1379 continue
1380
1381 if not ie.working():
1382 self.report_warning('The program functionality for this site has been marked as broken, '
1383 'and will probably not work.')
1384
1385 temp_id = ie.get_temp_id(url)
1386 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1387 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1388 if self.params.get('break_on_existing', False):
1389 raise ExistingVideoReached()
1390 break
1391 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1392 else:
1393 self.report_error('no suitable InfoExtractor for URL %s' % url)
1394
1395 def __handle_extraction_exceptions(func):
1396 @functools.wraps(func)
1397 def wrapper(self, *args, **kwargs):
1398 while True:
1399 try:
1400 return func(self, *args, **kwargs)
1401 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1402 raise
1403 except ReExtractInfo as e:
1404 if e.expected:
1405 self.to_screen(f'{e}; Re-extracting data')
1406 else:
1407 self.to_stderr('\r')
1408 self.report_warning(f'{e}; Re-extracting data')
1409 continue
1410 except GeoRestrictedError as e:
1411 msg = e.msg
1412 if e.countries:
1413 msg += '\nThis video is available in %s.' % ', '.join(
1414 map(ISO3166Utils.short2full, e.countries))
1415 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1416 self.report_error(msg)
1417 except ExtractorError as e: # An error we somewhat expected
1418 self.report_error(str(e), e.format_traceback())
1419 except Exception as e:
1420 if self.params.get('ignoreerrors'):
1421 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1422 else:
1423 raise
1424 break
1425 return wrapper
1426
1427 def _wait_for_video(self, ie_result):
1428 if (not self.params.get('wait_for_video')
1429 or ie_result.get('_type', 'video') != 'video'
1430 or ie_result.get('formats') or ie_result.get('url')):
1431 return
1432
1433 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1434 last_msg = ''
1435
1436 def progress(msg):
1437 nonlocal last_msg
1438 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1439 last_msg = msg
1440
1441 min_wait, max_wait = self.params.get('wait_for_video')
1442 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1443 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1444 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1445 self.report_warning('Release time of video is not known')
1446 elif (diff or 0) <= 0:
1447 self.report_warning('Video should already be available according to extracted info')
1448 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1449 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1450
1451 wait_till = time.time() + diff
1452 try:
1453 while True:
1454 diff = wait_till - time.time()
1455 if diff <= 0:
1456 progress('')
1457 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1458 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1459 time.sleep(1)
1460 except KeyboardInterrupt:
1461 progress('')
1462 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1463 except BaseException as e:
1464 if not isinstance(e, ReExtractInfo):
1465 self.to_screen('')
1466 raise
1467
1468 @__handle_extraction_exceptions
1469 def __extract_info(self, url, ie, download, extra_info, process):
1470 ie_result = ie.extract(url)
1471 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1472 return
1473 if isinstance(ie_result, list):
1474 # Backwards compatibility: old IE result format
1475 ie_result = {
1476 '_type': 'compat_list',
1477 'entries': ie_result,
1478 }
1479 if extra_info.get('original_url'):
1480 ie_result.setdefault('original_url', extra_info['original_url'])
1481 self.add_default_extra_info(ie_result, ie, url)
1482 if process:
1483 self._wait_for_video(ie_result)
1484 return self.process_ie_result(ie_result, download, extra_info)
1485 else:
1486 return ie_result
1487
1488 def add_default_extra_info(self, ie_result, ie, url):
1489 if url is not None:
1490 self.add_extra_info(ie_result, {
1491 'webpage_url': url,
1492 'original_url': url,
1493 })
1494 webpage_url = ie_result.get('webpage_url')
1495 if webpage_url:
1496 self.add_extra_info(ie_result, {
1497 'webpage_url_basename': url_basename(webpage_url),
1498 'webpage_url_domain': get_domain(webpage_url),
1499 })
1500 if ie is not None:
1501 self.add_extra_info(ie_result, {
1502 'extractor': ie.IE_NAME,
1503 'extractor_key': ie.ie_key(),
1504 })
1505
1506 def process_ie_result(self, ie_result, download=True, extra_info=None):
1507 """
1508 Take the result of the ie(may be modified) and resolve all unresolved
1509 references (URLs, playlist items).
1510
1511 It will also download the videos if 'download'.
1512 Returns the resolved ie_result.
1513 """
1514 if extra_info is None:
1515 extra_info = {}
1516 result_type = ie_result.get('_type', 'video')
1517
1518 if result_type in ('url', 'url_transparent'):
1519 ie_result['url'] = sanitize_url(ie_result['url'])
1520 if ie_result.get('original_url'):
1521 extra_info.setdefault('original_url', ie_result['original_url'])
1522
1523 extract_flat = self.params.get('extract_flat', False)
1524 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1525 or extract_flat is True):
1526 info_copy = ie_result.copy()
1527 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1528 if ie and not ie_result.get('id'):
1529 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1530 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1531 self.add_extra_info(info_copy, extra_info)
1532 info_copy, _ = self.pre_process(info_copy)
1533 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1534 if self.params.get('force_write_download_archive', False):
1535 self.record_download_archive(info_copy)
1536 return ie_result
1537
1538 if result_type == 'video':
1539 self.add_extra_info(ie_result, extra_info)
1540 ie_result = self.process_video_result(ie_result, download=download)
1541 additional_urls = (ie_result or {}).get('additional_urls')
1542 if additional_urls:
1543 # TODO: Improve MetadataParserPP to allow setting a list
1544 if isinstance(additional_urls, compat_str):
1545 additional_urls = [additional_urls]
1546 self.to_screen(
1547 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1548 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1549 ie_result['additional_entries'] = [
1550 self.extract_info(
1551 url, download, extra_info=extra_info,
1552 force_generic_extractor=self.params.get('force_generic_extractor'))
1553 for url in additional_urls
1554 ]
1555 return ie_result
1556 elif result_type == 'url':
1557 # We have to add extra_info to the results because it may be
1558 # contained in a playlist
1559 return self.extract_info(
1560 ie_result['url'], download,
1561 ie_key=ie_result.get('ie_key'),
1562 extra_info=extra_info)
1563 elif result_type == 'url_transparent':
1564 # Use the information from the embedding page
1565 info = self.extract_info(
1566 ie_result['url'], ie_key=ie_result.get('ie_key'),
1567 extra_info=extra_info, download=False, process=False)
1568
1569 # extract_info may return None when ignoreerrors is enabled and
1570 # extraction failed with an error, don't crash and return early
1571 # in this case
1572 if not info:
1573 return info
1574
1575 new_result = info.copy()
1576 new_result.update(filter_dict(ie_result, lambda k, v: (
1577 v is not None and k not in {'_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'})))
1578
1579 # Extracted info may not be a video result (i.e.
1580 # info.get('_type', 'video') != video) but rather an url or
1581 # url_transparent. In such cases outer metadata (from ie_result)
1582 # should be propagated to inner one (info). For this to happen
1583 # _type of info should be overridden with url_transparent. This
1584 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1585 if new_result.get('_type') == 'url':
1586 new_result['_type'] = 'url_transparent'
1587
1588 return self.process_ie_result(
1589 new_result, download=download, extra_info=extra_info)
1590 elif result_type in ('playlist', 'multi_video'):
1591 # Protect from infinite recursion due to recursively nested playlists
1592 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1593 webpage_url = ie_result['webpage_url']
1594 if webpage_url in self._playlist_urls:
1595 self.to_screen(
1596 '[download] Skipping already downloaded playlist: %s'
1597 % ie_result.get('title') or ie_result.get('id'))
1598 return
1599
1600 self._playlist_level += 1
1601 self._playlist_urls.add(webpage_url)
1602 self._fill_common_fields(ie_result, False)
1603 self._sanitize_thumbnails(ie_result)
1604 try:
1605 return self.__process_playlist(ie_result, download)
1606 finally:
1607 self._playlist_level -= 1
1608 if not self._playlist_level:
1609 self._playlist_urls.clear()
1610 elif result_type == 'compat_list':
1611 self.report_warning(
1612 'Extractor %s returned a compat_list result. '
1613 'It needs to be updated.' % ie_result.get('extractor'))
1614
1615 def _fixup(r):
1616 self.add_extra_info(r, {
1617 'extractor': ie_result['extractor'],
1618 'webpage_url': ie_result['webpage_url'],
1619 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1620 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1621 'extractor_key': ie_result['extractor_key'],
1622 })
1623 return r
1624 ie_result['entries'] = [
1625 self.process_ie_result(_fixup(r), download, extra_info)
1626 for r in ie_result['entries']
1627 ]
1628 return ie_result
1629 else:
1630 raise Exception('Invalid result type: %s' % result_type)
1631
1632 def _ensure_dir_exists(self, path):
1633 return make_dir(path, self.report_error)
1634
1635 @staticmethod
1636 def _playlist_infodict(ie_result, **kwargs):
1637 return {
1638 **ie_result,
1639 'playlist': ie_result.get('title') or ie_result.get('id'),
1640 'playlist_id': ie_result.get('id'),
1641 'playlist_title': ie_result.get('title'),
1642 'playlist_uploader': ie_result.get('uploader'),
1643 'playlist_uploader_id': ie_result.get('uploader_id'),
1644 'playlist_index': 0,
1645 **kwargs,
1646 }
1647
1648 def __process_playlist(self, ie_result, download):
1649 # We process each entry in the playlist
1650 playlist = ie_result.get('title') or ie_result.get('id')
1651 self.to_screen('[download] Downloading playlist: %s' % playlist)
1652
1653 if 'entries' not in ie_result:
1654 raise EntryNotInPlaylist('There are no entries')
1655
1656 MissingEntry = object()
1657 incomplete_entries = bool(ie_result.get('requested_entries'))
1658 if incomplete_entries:
1659 def fill_missing_entries(entries, indices):
1660 ret = [MissingEntry] * max(indices)
1661 for i, entry in zip(indices, entries):
1662 ret[i - 1] = entry
1663 return ret
1664 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1665
1666 playlist_results = []
1667
1668 playliststart = self.params.get('playliststart', 1)
1669 playlistend = self.params.get('playlistend')
1670 # For backwards compatibility, interpret -1 as whole list
1671 if playlistend == -1:
1672 playlistend = None
1673
1674 playlistitems_str = self.params.get('playlist_items')
1675 playlistitems = None
1676 if playlistitems_str is not None:
1677 def iter_playlistitems(format):
1678 for string_segment in format.split(','):
1679 if '-' in string_segment:
1680 start, end = string_segment.split('-')
1681 for item in range(int(start), int(end) + 1):
1682 yield int(item)
1683 else:
1684 yield int(string_segment)
1685 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1686
1687 ie_entries = ie_result['entries']
1688 if isinstance(ie_entries, list):
1689 playlist_count = len(ie_entries)
1690 msg = f'Collected {playlist_count} videos; downloading %d of them'
1691 ie_result['playlist_count'] = ie_result.get('playlist_count') or playlist_count
1692
1693 def get_entry(i):
1694 return ie_entries[i - 1]
1695 else:
1696 msg = 'Downloading %d videos'
1697 if not isinstance(ie_entries, (PagedList, LazyList)):
1698 ie_entries = LazyList(ie_entries)
1699 elif isinstance(ie_entries, InAdvancePagedList):
1700 if ie_entries._pagesize == 1:
1701 playlist_count = ie_entries._pagecount
1702
1703 def get_entry(i):
1704 return YoutubeDL.__handle_extraction_exceptions(
1705 lambda self, i: ie_entries[i - 1]
1706 )(self, i)
1707
1708 entries, broken = [], False
1709 items = playlistitems if playlistitems is not None else itertools.count(playliststart)
1710 for i in items:
1711 if i == 0:
1712 continue
1713 if playlistitems is None and playlistend is not None and playlistend < i:
1714 break
1715 entry = None
1716 try:
1717 entry = get_entry(i)
1718 if entry is MissingEntry:
1719 raise EntryNotInPlaylist()
1720 except (IndexError, EntryNotInPlaylist):
1721 if incomplete_entries:
1722 raise EntryNotInPlaylist(f'Entry {i} cannot be found')
1723 elif not playlistitems:
1724 break
1725 entries.append(entry)
1726 try:
1727 if entry is not None:
1728 # TODO: Add auto-generated fields
1729 self._match_entry(entry, incomplete=True, silent=True)
1730 except (ExistingVideoReached, RejectedVideoReached):
1731 broken = True
1732 break
1733 ie_result['entries'] = entries
1734
1735 # Save playlist_index before re-ordering
1736 entries = [
1737 ((playlistitems[i - 1] if playlistitems else i + playliststart - 1), entry)
1738 for i, entry in enumerate(entries, 1)
1739 if entry is not None]
1740 n_entries = len(entries)
1741
1742 if not (ie_result.get('playlist_count') or broken or playlistitems or playlistend):
1743 ie_result['playlist_count'] = n_entries
1744
1745 if not playlistitems and (playliststart != 1 or playlistend):
1746 playlistitems = list(range(playliststart, playliststart + n_entries))
1747 ie_result['requested_entries'] = playlistitems
1748
1749 _infojson_written = False
1750 write_playlist_files = self.params.get('allow_playlist_files', True)
1751 if write_playlist_files and self.params.get('list_thumbnails'):
1752 self.list_thumbnails(ie_result)
1753 if write_playlist_files and not self.params.get('simulate'):
1754 ie_copy = self._playlist_infodict(ie_result, n_entries=n_entries)
1755 _infojson_written = self._write_info_json(
1756 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1757 if _infojson_written is None:
1758 return
1759 if self._write_description('playlist', ie_result,
1760 self.prepare_filename(ie_copy, 'pl_description')) is None:
1761 return
1762 # TODO: This should be passed to ThumbnailsConvertor if necessary
1763 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1764
1765 if self.params.get('playlistreverse', False):
1766 entries = entries[::-1]
1767 if self.params.get('playlistrandom', False):
1768 random.shuffle(entries)
1769
1770 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1771
1772 self.to_screen(f'[{ie_result["extractor"]}] playlist {playlist}: {msg % n_entries}')
1773 failures = 0
1774 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1775 for i, entry_tuple in enumerate(entries, 1):
1776 playlist_index, entry = entry_tuple
1777 if 'playlist-index' in self.params.get('compat_opts', []):
1778 playlist_index = playlistitems[i - 1] if playlistitems else i + playliststart - 1
1779 self.to_screen('[download] Downloading video %s of %s' % (
1780 self._format_screen(i, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1781 # This __x_forwarded_for_ip thing is a bit ugly but requires
1782 # minimal changes
1783 if x_forwarded_for:
1784 entry['__x_forwarded_for_ip'] = x_forwarded_for
1785 extra = {
1786 'n_entries': n_entries,
1787 '__last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1788 'playlist_count': ie_result.get('playlist_count'),
1789 'playlist_index': playlist_index,
1790 'playlist_autonumber': i,
1791 'playlist': playlist,
1792 'playlist_id': ie_result.get('id'),
1793 'playlist_title': ie_result.get('title'),
1794 'playlist_uploader': ie_result.get('uploader'),
1795 'playlist_uploader_id': ie_result.get('uploader_id'),
1796 'extractor': ie_result['extractor'],
1797 'webpage_url': ie_result['webpage_url'],
1798 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1799 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1800 'extractor_key': ie_result['extractor_key'],
1801 }
1802
1803 if self._match_entry(entry, incomplete=True) is not None:
1804 continue
1805
1806 entry_result = self.__process_iterable_entry(entry, download, extra)
1807 if not entry_result:
1808 failures += 1
1809 if failures >= max_failures:
1810 self.report_error(
1811 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1812 break
1813 playlist_results.append(entry_result)
1814 ie_result['entries'] = playlist_results
1815
1816 # Write the updated info to json
1817 if _infojson_written is True and self._write_info_json(
1818 'updated playlist', ie_result,
1819 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1820 return
1821
1822 ie_result = self.run_all_pps('playlist', ie_result)
1823 self.to_screen(f'[download] Finished downloading playlist: {playlist}')
1824 return ie_result
1825
1826 @__handle_extraction_exceptions
1827 def __process_iterable_entry(self, entry, download, extra_info):
1828 return self.process_ie_result(
1829 entry, download=download, extra_info=extra_info)
1830
1831 def _build_format_filter(self, filter_spec):
1832 " Returns a function to filter the formats according to the filter_spec "
1833
1834 OPERATORS = {
1835 '<': operator.lt,
1836 '<=': operator.le,
1837 '>': operator.gt,
1838 '>=': operator.ge,
1839 '=': operator.eq,
1840 '!=': operator.ne,
1841 }
1842 operator_rex = re.compile(r'''(?x)\s*
1843 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1844 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1845 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1846 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1847 m = operator_rex.fullmatch(filter_spec)
1848 if m:
1849 try:
1850 comparison_value = int(m.group('value'))
1851 except ValueError:
1852 comparison_value = parse_filesize(m.group('value'))
1853 if comparison_value is None:
1854 comparison_value = parse_filesize(m.group('value') + 'B')
1855 if comparison_value is None:
1856 raise ValueError(
1857 'Invalid value %r in format specification %r' % (
1858 m.group('value'), filter_spec))
1859 op = OPERATORS[m.group('op')]
1860
1861 if not m:
1862 STR_OPERATORS = {
1863 '=': operator.eq,
1864 '^=': lambda attr, value: attr.startswith(value),
1865 '$=': lambda attr, value: attr.endswith(value),
1866 '*=': lambda attr, value: value in attr,
1867 '~=': lambda attr, value: value.search(attr) is not None
1868 }
1869 str_operator_rex = re.compile(r'''(?x)\s*
1870 (?P<key>[a-zA-Z0-9._-]+)\s*
1871 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1872 (?P<quote>["'])?
1873 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1874 (?(quote)(?P=quote))\s*
1875 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1876 m = str_operator_rex.fullmatch(filter_spec)
1877 if m:
1878 if m.group('op') == '~=':
1879 comparison_value = re.compile(m.group('value'))
1880 else:
1881 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1882 str_op = STR_OPERATORS[m.group('op')]
1883 if m.group('negation'):
1884 op = lambda attr, value: not str_op(attr, value)
1885 else:
1886 op = str_op
1887
1888 if not m:
1889 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1890
1891 def _filter(f):
1892 actual_value = f.get(m.group('key'))
1893 if actual_value is None:
1894 return m.group('none_inclusive')
1895 return op(actual_value, comparison_value)
1896 return _filter
1897
1898 def _check_formats(self, formats):
1899 for f in formats:
1900 self.to_screen('[info] Testing format %s' % f['format_id'])
1901 path = self.get_output_path('temp')
1902 if not self._ensure_dir_exists(f'{path}/'):
1903 continue
1904 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1905 temp_file.close()
1906 try:
1907 success, _ = self.dl(temp_file.name, f, test=True)
1908 except (DownloadError, IOError, OSError, ValueError) + network_exceptions:
1909 success = False
1910 finally:
1911 if os.path.exists(temp_file.name):
1912 try:
1913 os.remove(temp_file.name)
1914 except OSError:
1915 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1916 if success:
1917 yield f
1918 else:
1919 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1920
1921 def _default_format_spec(self, info_dict, download=True):
1922
1923 def can_merge():
1924 merger = FFmpegMergerPP(self)
1925 return merger.available and merger.can_merge()
1926
1927 prefer_best = (
1928 not self.params.get('simulate')
1929 and download
1930 and (
1931 not can_merge()
1932 or info_dict.get('is_live', False)
1933 or self.outtmpl_dict['default'] == '-'))
1934 compat = (
1935 prefer_best
1936 or self.params.get('allow_multiple_audio_streams', False)
1937 or 'format-spec' in self.params.get('compat_opts', []))
1938
1939 return (
1940 'best/bestvideo+bestaudio' if prefer_best
1941 else 'bestvideo*+bestaudio/best' if not compat
1942 else 'bestvideo+bestaudio/best')
1943
1944 def build_format_selector(self, format_spec):
1945 def syntax_error(note, start):
1946 message = (
1947 'Invalid format specification: '
1948 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1949 return SyntaxError(message)
1950
1951 PICKFIRST = 'PICKFIRST'
1952 MERGE = 'MERGE'
1953 SINGLE = 'SINGLE'
1954 GROUP = 'GROUP'
1955 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1956
1957 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1958 'video': self.params.get('allow_multiple_video_streams', False)}
1959
1960 check_formats = self.params.get('check_formats') == 'selected'
1961
1962 def _parse_filter(tokens):
1963 filter_parts = []
1964 for type, string, start, _, _ in tokens:
1965 if type == tokenize.OP and string == ']':
1966 return ''.join(filter_parts)
1967 else:
1968 filter_parts.append(string)
1969
1970 def _remove_unused_ops(tokens):
1971 # Remove operators that we don't use and join them with the surrounding strings
1972 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1973 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1974 last_string, last_start, last_end, last_line = None, None, None, None
1975 for type, string, start, end, line in tokens:
1976 if type == tokenize.OP and string == '[':
1977 if last_string:
1978 yield tokenize.NAME, last_string, last_start, last_end, last_line
1979 last_string = None
1980 yield type, string, start, end, line
1981 # everything inside brackets will be handled by _parse_filter
1982 for type, string, start, end, line in tokens:
1983 yield type, string, start, end, line
1984 if type == tokenize.OP and string == ']':
1985 break
1986 elif type == tokenize.OP and string in ALLOWED_OPS:
1987 if last_string:
1988 yield tokenize.NAME, last_string, last_start, last_end, last_line
1989 last_string = None
1990 yield type, string, start, end, line
1991 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1992 if not last_string:
1993 last_string = string
1994 last_start = start
1995 last_end = end
1996 else:
1997 last_string += string
1998 if last_string:
1999 yield tokenize.NAME, last_string, last_start, last_end, last_line
2000
2001 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2002 selectors = []
2003 current_selector = None
2004 for type, string, start, _, _ in tokens:
2005 # ENCODING is only defined in python 3.x
2006 if type == getattr(tokenize, 'ENCODING', None):
2007 continue
2008 elif type in [tokenize.NAME, tokenize.NUMBER]:
2009 current_selector = FormatSelector(SINGLE, string, [])
2010 elif type == tokenize.OP:
2011 if string == ')':
2012 if not inside_group:
2013 # ')' will be handled by the parentheses group
2014 tokens.restore_last_token()
2015 break
2016 elif inside_merge and string in ['/', ',']:
2017 tokens.restore_last_token()
2018 break
2019 elif inside_choice and string == ',':
2020 tokens.restore_last_token()
2021 break
2022 elif string == ',':
2023 if not current_selector:
2024 raise syntax_error('"," must follow a format selector', start)
2025 selectors.append(current_selector)
2026 current_selector = None
2027 elif string == '/':
2028 if not current_selector:
2029 raise syntax_error('"/" must follow a format selector', start)
2030 first_choice = current_selector
2031 second_choice = _parse_format_selection(tokens, inside_choice=True)
2032 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2033 elif string == '[':
2034 if not current_selector:
2035 current_selector = FormatSelector(SINGLE, 'best', [])
2036 format_filter = _parse_filter(tokens)
2037 current_selector.filters.append(format_filter)
2038 elif string == '(':
2039 if current_selector:
2040 raise syntax_error('Unexpected "("', start)
2041 group = _parse_format_selection(tokens, inside_group=True)
2042 current_selector = FormatSelector(GROUP, group, [])
2043 elif string == '+':
2044 if not current_selector:
2045 raise syntax_error('Unexpected "+"', start)
2046 selector_1 = current_selector
2047 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2048 if not selector_2:
2049 raise syntax_error('Expected a selector', start)
2050 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2051 else:
2052 raise syntax_error(f'Operator not recognized: "{string}"', start)
2053 elif type == tokenize.ENDMARKER:
2054 break
2055 if current_selector:
2056 selectors.append(current_selector)
2057 return selectors
2058
2059 def _merge(formats_pair):
2060 format_1, format_2 = formats_pair
2061
2062 formats_info = []
2063 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2064 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2065
2066 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2067 get_no_more = {'video': False, 'audio': False}
2068 for (i, fmt_info) in enumerate(formats_info):
2069 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2070 formats_info.pop(i)
2071 continue
2072 for aud_vid in ['audio', 'video']:
2073 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2074 if get_no_more[aud_vid]:
2075 formats_info.pop(i)
2076 break
2077 get_no_more[aud_vid] = True
2078
2079 if len(formats_info) == 1:
2080 return formats_info[0]
2081
2082 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2083 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2084
2085 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2086 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2087
2088 output_ext = self.params.get('merge_output_format')
2089 if not output_ext:
2090 if the_only_video:
2091 output_ext = the_only_video['ext']
2092 elif the_only_audio and not video_fmts:
2093 output_ext = the_only_audio['ext']
2094 else:
2095 output_ext = 'mkv'
2096
2097 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2098
2099 new_dict = {
2100 'requested_formats': formats_info,
2101 'format': '+'.join(filtered('format')),
2102 'format_id': '+'.join(filtered('format_id')),
2103 'ext': output_ext,
2104 'protocol': '+'.join(map(determine_protocol, formats_info)),
2105 'language': '+'.join(orderedSet(filtered('language'))) or None,
2106 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2107 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2108 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2109 }
2110
2111 if the_only_video:
2112 new_dict.update({
2113 'width': the_only_video.get('width'),
2114 'height': the_only_video.get('height'),
2115 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2116 'fps': the_only_video.get('fps'),
2117 'dynamic_range': the_only_video.get('dynamic_range'),
2118 'vcodec': the_only_video.get('vcodec'),
2119 'vbr': the_only_video.get('vbr'),
2120 'stretched_ratio': the_only_video.get('stretched_ratio'),
2121 })
2122
2123 if the_only_audio:
2124 new_dict.update({
2125 'acodec': the_only_audio.get('acodec'),
2126 'abr': the_only_audio.get('abr'),
2127 'asr': the_only_audio.get('asr'),
2128 })
2129
2130 return new_dict
2131
2132 def _check_formats(formats):
2133 if not check_formats:
2134 yield from formats
2135 return
2136 yield from self._check_formats(formats)
2137
2138 def _build_selector_function(selector):
2139 if isinstance(selector, list): # ,
2140 fs = [_build_selector_function(s) for s in selector]
2141
2142 def selector_function(ctx):
2143 for f in fs:
2144 yield from f(ctx)
2145 return selector_function
2146
2147 elif selector.type == GROUP: # ()
2148 selector_function = _build_selector_function(selector.selector)
2149
2150 elif selector.type == PICKFIRST: # /
2151 fs = [_build_selector_function(s) for s in selector.selector]
2152
2153 def selector_function(ctx):
2154 for f in fs:
2155 picked_formats = list(f(ctx))
2156 if picked_formats:
2157 return picked_formats
2158 return []
2159
2160 elif selector.type == MERGE: # +
2161 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2162
2163 def selector_function(ctx):
2164 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2165 yield _merge(pair)
2166
2167 elif selector.type == SINGLE: # atom
2168 format_spec = selector.selector or 'best'
2169
2170 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2171 if format_spec == 'all':
2172 def selector_function(ctx):
2173 yield from _check_formats(ctx['formats'][::-1])
2174 elif format_spec == 'mergeall':
2175 def selector_function(ctx):
2176 formats = list(_check_formats(
2177 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2178 if not formats:
2179 return
2180 merged_format = formats[-1]
2181 for f in formats[-2::-1]:
2182 merged_format = _merge((merged_format, f))
2183 yield merged_format
2184
2185 else:
2186 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2187 mobj = re.match(
2188 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2189 format_spec)
2190 if mobj is not None:
2191 format_idx = int_or_none(mobj.group('n'), default=1)
2192 format_reverse = mobj.group('bw')[0] == 'b'
2193 format_type = (mobj.group('type') or [None])[0]
2194 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2195 format_modified = mobj.group('mod') is not None
2196
2197 format_fallback = not format_type and not format_modified # for b, w
2198 _filter_f = (
2199 (lambda f: f.get('%scodec' % format_type) != 'none')
2200 if format_type and format_modified # bv*, ba*, wv*, wa*
2201 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2202 if format_type # bv, ba, wv, wa
2203 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2204 if not format_modified # b, w
2205 else lambda f: True) # b*, w*
2206 filter_f = lambda f: _filter_f(f) and (
2207 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2208 else:
2209 if format_spec in self._format_selection_exts['audio']:
2210 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2211 elif format_spec in self._format_selection_exts['video']:
2212 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2213 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2214 elif format_spec in self._format_selection_exts['storyboards']:
2215 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2216 else:
2217 filter_f = lambda f: f.get('format_id') == format_spec # id
2218
2219 def selector_function(ctx):
2220 formats = list(ctx['formats'])
2221 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2222 if not matches:
2223 if format_fallback and ctx['incomplete_formats']:
2224 # for extractors with incomplete formats (audio only (soundcloud)
2225 # or video only (imgur)) best/worst will fallback to
2226 # best/worst {video,audio}-only format
2227 matches = formats
2228 elif seperate_fallback and not ctx['has_merged_format']:
2229 # for compatibility with youtube-dl when there is no pre-merged format
2230 matches = list(filter(seperate_fallback, formats))
2231 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2232 try:
2233 yield matches[format_idx - 1]
2234 except LazyList.IndexError:
2235 return
2236
2237 filters = [self._build_format_filter(f) for f in selector.filters]
2238
2239 def final_selector(ctx):
2240 ctx_copy = dict(ctx)
2241 for _filter in filters:
2242 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2243 return selector_function(ctx_copy)
2244 return final_selector
2245
2246 stream = io.BytesIO(format_spec.encode('utf-8'))
2247 try:
2248 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2249 except tokenize.TokenError:
2250 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2251
2252 class TokenIterator:
2253 def __init__(self, tokens):
2254 self.tokens = tokens
2255 self.counter = 0
2256
2257 def __iter__(self):
2258 return self
2259
2260 def __next__(self):
2261 if self.counter >= len(self.tokens):
2262 raise StopIteration()
2263 value = self.tokens[self.counter]
2264 self.counter += 1
2265 return value
2266
2267 next = __next__
2268
2269 def restore_last_token(self):
2270 self.counter -= 1
2271
2272 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2273 return _build_selector_function(parsed_selector)
2274
2275 def _calc_headers(self, info_dict):
2276 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2277
2278 cookies = self._calc_cookies(info_dict)
2279 if cookies:
2280 res['Cookie'] = cookies
2281
2282 if 'X-Forwarded-For' not in res:
2283 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2284 if x_forwarded_for_ip:
2285 res['X-Forwarded-For'] = x_forwarded_for_ip
2286
2287 return res
2288
2289 def _calc_cookies(self, info_dict):
2290 pr = sanitized_Request(info_dict['url'])
2291 self.cookiejar.add_cookie_header(pr)
2292 return pr.get_header('Cookie')
2293
2294 def _sort_thumbnails(self, thumbnails):
2295 thumbnails.sort(key=lambda t: (
2296 t.get('preference') if t.get('preference') is not None else -1,
2297 t.get('width') if t.get('width') is not None else -1,
2298 t.get('height') if t.get('height') is not None else -1,
2299 t.get('id') if t.get('id') is not None else '',
2300 t.get('url')))
2301
2302 def _sanitize_thumbnails(self, info_dict):
2303 thumbnails = info_dict.get('thumbnails')
2304 if thumbnails is None:
2305 thumbnail = info_dict.get('thumbnail')
2306 if thumbnail:
2307 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2308 if not thumbnails:
2309 return
2310
2311 def check_thumbnails(thumbnails):
2312 for t in thumbnails:
2313 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2314 try:
2315 self.urlopen(HEADRequest(t['url']))
2316 except network_exceptions as err:
2317 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2318 continue
2319 yield t
2320
2321 self._sort_thumbnails(thumbnails)
2322 for i, t in enumerate(thumbnails):
2323 if t.get('id') is None:
2324 t['id'] = '%d' % i
2325 if t.get('width') and t.get('height'):
2326 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2327 t['url'] = sanitize_url(t['url'])
2328
2329 if self.params.get('check_formats') is True:
2330 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2331 else:
2332 info_dict['thumbnails'] = thumbnails
2333
2334 def _fill_common_fields(self, info_dict, is_video=True):
2335 # TODO: move sanitization here
2336 if is_video:
2337 # playlists are allowed to lack "title"
2338 info_dict['fulltitle'] = info_dict.get('title')
2339 if 'title' not in info_dict:
2340 raise ExtractorError('Missing "title" field in extractor result',
2341 video_id=info_dict['id'], ie=info_dict['extractor'])
2342 elif not info_dict.get('title'):
2343 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2344 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2345
2346 if info_dict.get('duration') is not None:
2347 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2348
2349 for ts_key, date_key in (
2350 ('timestamp', 'upload_date'),
2351 ('release_timestamp', 'release_date'),
2352 ('modified_timestamp', 'modified_date'),
2353 ):
2354 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2355 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2356 # see http://bugs.python.org/issue1646728)
2357 with contextlib.suppress(ValueError, OverflowError, OSError):
2358 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2359 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2360
2361 live_keys = ('is_live', 'was_live')
2362 live_status = info_dict.get('live_status')
2363 if live_status is None:
2364 for key in live_keys:
2365 if info_dict.get(key) is False:
2366 continue
2367 if info_dict.get(key):
2368 live_status = key
2369 break
2370 if all(info_dict.get(key) is False for key in live_keys):
2371 live_status = 'not_live'
2372 if live_status:
2373 info_dict['live_status'] = live_status
2374 for key in live_keys:
2375 if info_dict.get(key) is None:
2376 info_dict[key] = (live_status == key)
2377
2378 # Auto generate title fields corresponding to the *_number fields when missing
2379 # in order to always have clean titles. This is very common for TV series.
2380 for field in ('chapter', 'season', 'episode'):
2381 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2382 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2383
2384 def process_video_result(self, info_dict, download=True):
2385 assert info_dict.get('_type', 'video') == 'video'
2386 self._num_videos += 1
2387
2388 if 'id' not in info_dict:
2389 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2390 elif not info_dict.get('id'):
2391 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2392
2393 def report_force_conversion(field, field_not, conversion):
2394 self.report_warning(
2395 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2396 % (field, field_not, conversion))
2397
2398 def sanitize_string_field(info, string_field):
2399 field = info.get(string_field)
2400 if field is None or isinstance(field, compat_str):
2401 return
2402 report_force_conversion(string_field, 'a string', 'string')
2403 info[string_field] = compat_str(field)
2404
2405 def sanitize_numeric_fields(info):
2406 for numeric_field in self._NUMERIC_FIELDS:
2407 field = info.get(numeric_field)
2408 if field is None or isinstance(field, (int, float)):
2409 continue
2410 report_force_conversion(numeric_field, 'numeric', 'int')
2411 info[numeric_field] = int_or_none(field)
2412
2413 sanitize_string_field(info_dict, 'id')
2414 sanitize_numeric_fields(info_dict)
2415 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2416 self.report_warning('"duration" field is negative, there is an error in extractor')
2417
2418 if 'playlist' not in info_dict:
2419 # It isn't part of a playlist
2420 info_dict['playlist'] = None
2421 info_dict['playlist_index'] = None
2422
2423 self._sanitize_thumbnails(info_dict)
2424
2425 thumbnail = info_dict.get('thumbnail')
2426 thumbnails = info_dict.get('thumbnails')
2427 if thumbnail:
2428 info_dict['thumbnail'] = sanitize_url(thumbnail)
2429 elif thumbnails:
2430 info_dict['thumbnail'] = thumbnails[-1]['url']
2431
2432 if info_dict.get('display_id') is None and 'id' in info_dict:
2433 info_dict['display_id'] = info_dict['id']
2434
2435 self._fill_common_fields(info_dict)
2436
2437 for cc_kind in ('subtitles', 'automatic_captions'):
2438 cc = info_dict.get(cc_kind)
2439 if cc:
2440 for _, subtitle in cc.items():
2441 for subtitle_format in subtitle:
2442 if subtitle_format.get('url'):
2443 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2444 if subtitle_format.get('ext') is None:
2445 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2446
2447 automatic_captions = info_dict.get('automatic_captions')
2448 subtitles = info_dict.get('subtitles')
2449
2450 info_dict['requested_subtitles'] = self.process_subtitles(
2451 info_dict['id'], subtitles, automatic_captions)
2452
2453 if info_dict.get('formats') is None:
2454 # There's only one format available
2455 formats = [info_dict]
2456 else:
2457 formats = info_dict['formats']
2458
2459 # or None ensures --clean-infojson removes it
2460 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2461 if not self.params.get('allow_unplayable_formats'):
2462 formats = [f for f in formats if not f.get('has_drm')]
2463 if info_dict['_has_drm'] and all(
2464 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2465 self.report_warning(
2466 'This video is DRM protected and only images are available for download. '
2467 'Use --list-formats to see them')
2468
2469 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2470 if not get_from_start:
2471 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2472 if info_dict.get('is_live') and formats:
2473 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2474 if get_from_start and not formats:
2475 self.raise_no_formats(info_dict, msg=(
2476 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2477 'If you want to download from the current time, use --no-live-from-start'))
2478
2479 if not formats:
2480 self.raise_no_formats(info_dict)
2481
2482 def is_wellformed(f):
2483 url = f.get('url')
2484 if not url:
2485 self.report_warning(
2486 '"url" field is missing or empty - skipping format, '
2487 'there is an error in extractor')
2488 return False
2489 if isinstance(url, bytes):
2490 sanitize_string_field(f, 'url')
2491 return True
2492
2493 # Filter out malformed formats for better extraction robustness
2494 formats = list(filter(is_wellformed, formats))
2495
2496 formats_dict = {}
2497
2498 # We check that all the formats have the format and format_id fields
2499 for i, format in enumerate(formats):
2500 sanitize_string_field(format, 'format_id')
2501 sanitize_numeric_fields(format)
2502 format['url'] = sanitize_url(format['url'])
2503 if not format.get('format_id'):
2504 format['format_id'] = compat_str(i)
2505 else:
2506 # Sanitize format_id from characters used in format selector expression
2507 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2508 format_id = format['format_id']
2509 if format_id not in formats_dict:
2510 formats_dict[format_id] = []
2511 formats_dict[format_id].append(format)
2512
2513 # Make sure all formats have unique format_id
2514 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2515 for format_id, ambiguous_formats in formats_dict.items():
2516 ambigious_id = len(ambiguous_formats) > 1
2517 for i, format in enumerate(ambiguous_formats):
2518 if ambigious_id:
2519 format['format_id'] = '%s-%d' % (format_id, i)
2520 if format.get('ext') is None:
2521 format['ext'] = determine_ext(format['url']).lower()
2522 # Ensure there is no conflict between id and ext in format selection
2523 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2524 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2525 format['format_id'] = 'f%s' % format['format_id']
2526
2527 for i, format in enumerate(formats):
2528 if format.get('format') is None:
2529 format['format'] = '{id} - {res}{note}'.format(
2530 id=format['format_id'],
2531 res=self.format_resolution(format),
2532 note=format_field(format, 'format_note', ' (%s)'),
2533 )
2534 if format.get('protocol') is None:
2535 format['protocol'] = determine_protocol(format)
2536 if format.get('resolution') is None:
2537 format['resolution'] = self.format_resolution(format, default=None)
2538 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2539 format['dynamic_range'] = 'SDR'
2540 if (info_dict.get('duration') and format.get('tbr')
2541 and not format.get('filesize') and not format.get('filesize_approx')):
2542 format['filesize_approx'] = info_dict['duration'] * format['tbr'] * (1024 / 8)
2543
2544 # Add HTTP headers, so that external programs can use them from the
2545 # json output
2546 full_format_info = info_dict.copy()
2547 full_format_info.update(format)
2548 format['http_headers'] = self._calc_headers(full_format_info)
2549 # Remove private housekeeping stuff
2550 if '__x_forwarded_for_ip' in info_dict:
2551 del info_dict['__x_forwarded_for_ip']
2552
2553 if self.params.get('check_formats') is True:
2554 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2555
2556 if not formats or formats[0] is not info_dict:
2557 # only set the 'formats' fields if the original info_dict list them
2558 # otherwise we end up with a circular reference, the first (and unique)
2559 # element in the 'formats' field in info_dict is info_dict itself,
2560 # which can't be exported to json
2561 info_dict['formats'] = formats
2562
2563 info_dict, _ = self.pre_process(info_dict)
2564
2565 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2566 return info_dict
2567
2568 self.post_extract(info_dict)
2569 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2570
2571 # The pre-processors may have modified the formats
2572 formats = info_dict.get('formats', [info_dict])
2573
2574 list_only = self.params.get('simulate') is None and (
2575 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2576 interactive_format_selection = not list_only and self.format_selector == '-'
2577 if self.params.get('list_thumbnails'):
2578 self.list_thumbnails(info_dict)
2579 if self.params.get('listsubtitles'):
2580 if 'automatic_captions' in info_dict:
2581 self.list_subtitles(
2582 info_dict['id'], automatic_captions, 'automatic captions')
2583 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2584 if self.params.get('listformats') or interactive_format_selection:
2585 self.list_formats(info_dict)
2586 if list_only:
2587 # Without this printing, -F --print-json will not work
2588 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2589 return
2590
2591 format_selector = self.format_selector
2592 if format_selector is None:
2593 req_format = self._default_format_spec(info_dict, download=download)
2594 self.write_debug('Default format spec: %s' % req_format)
2595 format_selector = self.build_format_selector(req_format)
2596
2597 while True:
2598 if interactive_format_selection:
2599 req_format = input(
2600 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2601 try:
2602 format_selector = self.build_format_selector(req_format)
2603 except SyntaxError as err:
2604 self.report_error(err, tb=False, is_error=False)
2605 continue
2606
2607 formats_to_download = list(format_selector({
2608 'formats': formats,
2609 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2610 'incomplete_formats': (
2611 # All formats are video-only or
2612 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2613 # all formats are audio-only
2614 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2615 }))
2616 if interactive_format_selection and not formats_to_download:
2617 self.report_error('Requested format is not available', tb=False, is_error=False)
2618 continue
2619 break
2620
2621 if not formats_to_download:
2622 if not self.params.get('ignore_no_formats_error'):
2623 raise ExtractorError(
2624 'Requested format is not available. Use --list-formats for a list of available formats',
2625 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2626 self.report_warning('Requested format is not available')
2627 # Process what we can, even without any available formats.
2628 formats_to_download = [{}]
2629
2630 best_format = formats_to_download[-1]
2631 if download:
2632 if best_format:
2633 self.to_screen(
2634 f'[info] {info_dict["id"]}: Downloading {len(formats_to_download)} format(s): '
2635 + ', '.join([f['format_id'] for f in formats_to_download]))
2636 max_downloads_reached = False
2637 for i, fmt in enumerate(formats_to_download):
2638 formats_to_download[i] = new_info = self._copy_infodict(info_dict)
2639 new_info.update(fmt)
2640 try:
2641 self.process_info(new_info)
2642 except MaxDownloadsReached:
2643 max_downloads_reached = True
2644 # Remove copied info
2645 for key, val in tuple(new_info.items()):
2646 if info_dict.get(key) == val:
2647 new_info.pop(key)
2648 if max_downloads_reached:
2649 break
2650
2651 write_archive = {f.get('__write_download_archive', False) for f in formats_to_download}
2652 assert write_archive.issubset({True, False, 'ignore'})
2653 if True in write_archive and False not in write_archive:
2654 self.record_download_archive(info_dict)
2655
2656 info_dict['requested_downloads'] = formats_to_download
2657 info_dict = self.run_all_pps('after_video', info_dict)
2658 if max_downloads_reached:
2659 raise MaxDownloadsReached()
2660
2661 # We update the info dict with the selected best quality format (backwards compatibility)
2662 info_dict.update(best_format)
2663 return info_dict
2664
2665 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2666 """Select the requested subtitles and their format"""
2667 available_subs, normal_sub_langs = {}, []
2668 if normal_subtitles and self.params.get('writesubtitles'):
2669 available_subs.update(normal_subtitles)
2670 normal_sub_langs = tuple(normal_subtitles.keys())
2671 if automatic_captions and self.params.get('writeautomaticsub'):
2672 for lang, cap_info in automatic_captions.items():
2673 if lang not in available_subs:
2674 available_subs[lang] = cap_info
2675
2676 if (not self.params.get('writesubtitles') and not
2677 self.params.get('writeautomaticsub') or not
2678 available_subs):
2679 return None
2680
2681 all_sub_langs = tuple(available_subs.keys())
2682 if self.params.get('allsubtitles', False):
2683 requested_langs = all_sub_langs
2684 elif self.params.get('subtitleslangs', False):
2685 # A list is used so that the order of languages will be the same as
2686 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2687 requested_langs = []
2688 for lang_re in self.params.get('subtitleslangs'):
2689 discard = lang_re[0] == '-'
2690 if discard:
2691 lang_re = lang_re[1:]
2692 if lang_re == 'all':
2693 if discard:
2694 requested_langs = []
2695 else:
2696 requested_langs.extend(all_sub_langs)
2697 continue
2698 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2699 if discard:
2700 for lang in current_langs:
2701 while lang in requested_langs:
2702 requested_langs.remove(lang)
2703 else:
2704 requested_langs.extend(current_langs)
2705 requested_langs = orderedSet(requested_langs)
2706 elif normal_sub_langs:
2707 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2708 else:
2709 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2710 if requested_langs:
2711 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2712
2713 formats_query = self.params.get('subtitlesformat', 'best')
2714 formats_preference = formats_query.split('/') if formats_query else []
2715 subs = {}
2716 for lang in requested_langs:
2717 formats = available_subs.get(lang)
2718 if formats is None:
2719 self.report_warning(f'{lang} subtitles not available for {video_id}')
2720 continue
2721 for ext in formats_preference:
2722 if ext == 'best':
2723 f = formats[-1]
2724 break
2725 matches = list(filter(lambda f: f['ext'] == ext, formats))
2726 if matches:
2727 f = matches[-1]
2728 break
2729 else:
2730 f = formats[-1]
2731 self.report_warning(
2732 'No subtitle format found matching "%s" for language %s, '
2733 'using %s' % (formats_query, lang, f['ext']))
2734 subs[lang] = f
2735 return subs
2736
2737 def _forceprint(self, key, info_dict):
2738 if info_dict is None:
2739 return
2740 info_copy = info_dict.copy()
2741 info_copy['formats_table'] = self.render_formats_table(info_dict)
2742 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2743 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2744 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2745
2746 def format_tmpl(tmpl):
2747 mobj = re.match(r'\w+(=?)$', tmpl)
2748 if mobj and mobj.group(1):
2749 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2750 elif mobj:
2751 return f'%({tmpl})s'
2752 return tmpl
2753
2754 for tmpl in self.params['forceprint'].get(key, []):
2755 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2756
2757 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2758 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2759 tmpl = format_tmpl(tmpl)
2760 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2761 if self._ensure_dir_exists(filename):
2762 with open(filename, 'a', encoding='utf-8') as f:
2763 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2764
2765 def __forced_printings(self, info_dict, filename, incomplete):
2766 def print_mandatory(field, actual_field=None):
2767 if actual_field is None:
2768 actual_field = field
2769 if (self.params.get('force%s' % field, False)
2770 and (not incomplete or info_dict.get(actual_field) is not None)):
2771 self.to_stdout(info_dict[actual_field])
2772
2773 def print_optional(field):
2774 if (self.params.get('force%s' % field, False)
2775 and info_dict.get(field) is not None):
2776 self.to_stdout(info_dict[field])
2777
2778 info_dict = info_dict.copy()
2779 if filename is not None:
2780 info_dict['filename'] = filename
2781 if info_dict.get('requested_formats') is not None:
2782 # For RTMP URLs, also include the playpath
2783 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2784 elif info_dict.get('url'):
2785 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2786
2787 if (self.params.get('forcejson')
2788 or self.params['forceprint'].get('video')
2789 or self.params['print_to_file'].get('video')):
2790 self.post_extract(info_dict)
2791 self._forceprint('video', info_dict)
2792
2793 print_mandatory('title')
2794 print_mandatory('id')
2795 print_mandatory('url', 'urls')
2796 print_optional('thumbnail')
2797 print_optional('description')
2798 print_optional('filename')
2799 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2800 self.to_stdout(formatSeconds(info_dict['duration']))
2801 print_mandatory('format')
2802
2803 if self.params.get('forcejson'):
2804 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2805
2806 def dl(self, name, info, subtitle=False, test=False):
2807 if not info.get('url'):
2808 self.raise_no_formats(info, True)
2809
2810 if test:
2811 verbose = self.params.get('verbose')
2812 params = {
2813 'test': True,
2814 'quiet': self.params.get('quiet') or not verbose,
2815 'verbose': verbose,
2816 'noprogress': not verbose,
2817 'nopart': True,
2818 'skip_unavailable_fragments': False,
2819 'keep_fragments': False,
2820 'overwrites': True,
2821 '_no_ytdl_file': True,
2822 }
2823 else:
2824 params = self.params
2825 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2826 if not test:
2827 for ph in self._progress_hooks:
2828 fd.add_progress_hook(ph)
2829 urls = '", "'.join(
2830 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2831 for f in info.get('requested_formats', []) or [info])
2832 self.write_debug('Invoking downloader on "%s"' % urls)
2833
2834 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2835 # But it may contain objects that are not deep-copyable
2836 new_info = self._copy_infodict(info)
2837 if new_info.get('http_headers') is None:
2838 new_info['http_headers'] = self._calc_headers(new_info)
2839 return fd.download(name, new_info, subtitle)
2840
2841 def existing_file(self, filepaths, *, default_overwrite=True):
2842 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2843 if existing_files and not self.params.get('overwrites', default_overwrite):
2844 return existing_files[0]
2845
2846 for file in existing_files:
2847 self.report_file_delete(file)
2848 os.remove(file)
2849 return None
2850
2851 def process_info(self, info_dict):
2852 """Process a single resolved IE result. (Modifies it in-place)"""
2853
2854 assert info_dict.get('_type', 'video') == 'video'
2855 original_infodict = info_dict
2856
2857 if 'format' not in info_dict and 'ext' in info_dict:
2858 info_dict['format'] = info_dict['ext']
2859
2860 # This is mostly just for backward compatibility of process_info
2861 # As a side-effect, this allows for format-specific filters
2862 if self._match_entry(info_dict) is not None:
2863 info_dict['__write_download_archive'] = 'ignore'
2864 return
2865
2866 # Does nothing under normal operation - for backward compatibility of process_info
2867 self.post_extract(info_dict)
2868 self._num_downloads += 1
2869
2870 # info_dict['_filename'] needs to be set for backward compatibility
2871 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2872 temp_filename = self.prepare_filename(info_dict, 'temp')
2873 files_to_move = {}
2874
2875 # Forced printings
2876 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2877
2878 if self.params.get('simulate'):
2879 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2880 return
2881
2882 if full_filename is None:
2883 return
2884 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2885 return
2886 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2887 return
2888
2889 if self._write_description('video', info_dict,
2890 self.prepare_filename(info_dict, 'description')) is None:
2891 return
2892
2893 sub_files = self._write_subtitles(info_dict, temp_filename)
2894 if sub_files is None:
2895 return
2896 files_to_move.update(dict(sub_files))
2897
2898 thumb_files = self._write_thumbnails(
2899 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2900 if thumb_files is None:
2901 return
2902 files_to_move.update(dict(thumb_files))
2903
2904 infofn = self.prepare_filename(info_dict, 'infojson')
2905 _infojson_written = self._write_info_json('video', info_dict, infofn)
2906 if _infojson_written:
2907 info_dict['infojson_filename'] = infofn
2908 # For backward compatibility, even though it was a private field
2909 info_dict['__infojson_filename'] = infofn
2910 elif _infojson_written is None:
2911 return
2912
2913 # Note: Annotations are deprecated
2914 annofn = None
2915 if self.params.get('writeannotations', False):
2916 annofn = self.prepare_filename(info_dict, 'annotation')
2917 if annofn:
2918 if not self._ensure_dir_exists(encodeFilename(annofn)):
2919 return
2920 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2921 self.to_screen('[info] Video annotations are already present')
2922 elif not info_dict.get('annotations'):
2923 self.report_warning('There are no annotations to write.')
2924 else:
2925 try:
2926 self.to_screen('[info] Writing video annotations to: ' + annofn)
2927 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2928 annofile.write(info_dict['annotations'])
2929 except (KeyError, TypeError):
2930 self.report_warning('There are no annotations to write.')
2931 except OSError:
2932 self.report_error('Cannot write annotations file: ' + annofn)
2933 return
2934
2935 # Write internet shortcut files
2936 def _write_link_file(link_type):
2937 url = try_get(info_dict['webpage_url'], iri_to_uri)
2938 if not url:
2939 self.report_warning(
2940 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2941 return True
2942 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2943 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2944 return False
2945 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2946 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2947 return True
2948 try:
2949 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2950 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2951 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2952 template_vars = {'url': url}
2953 if link_type == 'desktop':
2954 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2955 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2956 except OSError:
2957 self.report_error(f'Cannot write internet shortcut {linkfn}')
2958 return False
2959 return True
2960
2961 write_links = {
2962 'url': self.params.get('writeurllink'),
2963 'webloc': self.params.get('writewebloclink'),
2964 'desktop': self.params.get('writedesktoplink'),
2965 }
2966 if self.params.get('writelink'):
2967 link_type = ('webloc' if sys.platform == 'darwin'
2968 else 'desktop' if sys.platform.startswith('linux')
2969 else 'url')
2970 write_links[link_type] = True
2971
2972 if any(should_write and not _write_link_file(link_type)
2973 for link_type, should_write in write_links.items()):
2974 return
2975
2976 def replace_info_dict(new_info):
2977 nonlocal info_dict
2978 if new_info == info_dict:
2979 return
2980 info_dict.clear()
2981 info_dict.update(new_info)
2982
2983 try:
2984 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2985 replace_info_dict(new_info)
2986 except PostProcessingError as err:
2987 self.report_error('Preprocessing: %s' % str(err))
2988 return
2989
2990 if self.params.get('skip_download'):
2991 info_dict['filepath'] = temp_filename
2992 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2993 info_dict['__files_to_move'] = files_to_move
2994 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2995 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2996 else:
2997 # Download
2998 info_dict.setdefault('__postprocessors', [])
2999 try:
3000
3001 def existing_video_file(*filepaths):
3002 ext = info_dict.get('ext')
3003 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3004 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3005 default_overwrite=False)
3006 if file:
3007 info_dict['ext'] = os.path.splitext(file)[1][1:]
3008 return file
3009
3010 success = True
3011 if info_dict.get('requested_formats') is not None:
3012
3013 def compatible_formats(formats):
3014 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3015 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3016 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3017 if len(video_formats) > 2 or len(audio_formats) > 2:
3018 return False
3019
3020 # Check extension
3021 exts = {format.get('ext') for format in formats}
3022 COMPATIBLE_EXTS = (
3023 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3024 {'webm'},
3025 )
3026 for ext_sets in COMPATIBLE_EXTS:
3027 if ext_sets.issuperset(exts):
3028 return True
3029 # TODO: Check acodec/vcodec
3030 return False
3031
3032 requested_formats = info_dict['requested_formats']
3033 old_ext = info_dict['ext']
3034 if self.params.get('merge_output_format') is None:
3035 if not compatible_formats(requested_formats):
3036 info_dict['ext'] = 'mkv'
3037 self.report_warning(
3038 'Requested formats are incompatible for merge and will be merged into mkv')
3039 if (info_dict['ext'] == 'webm'
3040 and info_dict.get('thumbnails')
3041 # check with type instead of pp_key, __name__, or isinstance
3042 # since we dont want any custom PPs to trigger this
3043 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])):
3044 info_dict['ext'] = 'mkv'
3045 self.report_warning(
3046 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3047 new_ext = info_dict['ext']
3048
3049 def correct_ext(filename, ext=new_ext):
3050 if filename == '-':
3051 return filename
3052 filename_real_ext = os.path.splitext(filename)[1][1:]
3053 filename_wo_ext = (
3054 os.path.splitext(filename)[0]
3055 if filename_real_ext in (old_ext, new_ext)
3056 else filename)
3057 return f'{filename_wo_ext}.{ext}'
3058
3059 # Ensure filename always has a correct extension for successful merge
3060 full_filename = correct_ext(full_filename)
3061 temp_filename = correct_ext(temp_filename)
3062 dl_filename = existing_video_file(full_filename, temp_filename)
3063 info_dict['__real_download'] = False
3064
3065 downloaded = []
3066 merger = FFmpegMergerPP(self)
3067
3068 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3069 if dl_filename is not None:
3070 self.report_file_already_downloaded(dl_filename)
3071 elif fd:
3072 for f in requested_formats if fd != FFmpegFD else []:
3073 f['filepath'] = fname = prepend_extension(
3074 correct_ext(temp_filename, info_dict['ext']),
3075 'f%s' % f['format_id'], info_dict['ext'])
3076 downloaded.append(fname)
3077 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3078 success, real_download = self.dl(temp_filename, info_dict)
3079 info_dict['__real_download'] = real_download
3080 else:
3081 if self.params.get('allow_unplayable_formats'):
3082 self.report_warning(
3083 'You have requested merging of multiple formats '
3084 'while also allowing unplayable formats to be downloaded. '
3085 'The formats won\'t be merged to prevent data corruption.')
3086 elif not merger.available:
3087 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3088 if not self.params.get('ignoreerrors'):
3089 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3090 return
3091 self.report_warning(f'{msg}. The formats won\'t be merged')
3092
3093 if temp_filename == '-':
3094 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3095 else 'but the formats are incompatible for simultaneous download' if merger.available
3096 else 'but ffmpeg is not installed')
3097 self.report_warning(
3098 f'You have requested downloading multiple formats to stdout {reason}. '
3099 'The formats will be streamed one after the other')
3100 fname = temp_filename
3101 for f in requested_formats:
3102 new_info = dict(info_dict)
3103 del new_info['requested_formats']
3104 new_info.update(f)
3105 if temp_filename != '-':
3106 fname = prepend_extension(
3107 correct_ext(temp_filename, new_info['ext']),
3108 'f%s' % f['format_id'], new_info['ext'])
3109 if not self._ensure_dir_exists(fname):
3110 return
3111 f['filepath'] = fname
3112 downloaded.append(fname)
3113 partial_success, real_download = self.dl(fname, new_info)
3114 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3115 success = success and partial_success
3116
3117 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3118 info_dict['__postprocessors'].append(merger)
3119 info_dict['__files_to_merge'] = downloaded
3120 # Even if there were no downloads, it is being merged only now
3121 info_dict['__real_download'] = True
3122 else:
3123 for file in downloaded:
3124 files_to_move[file] = None
3125 else:
3126 # Just a single file
3127 dl_filename = existing_video_file(full_filename, temp_filename)
3128 if dl_filename is None or dl_filename == temp_filename:
3129 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3130 # So we should try to resume the download
3131 success, real_download = self.dl(temp_filename, info_dict)
3132 info_dict['__real_download'] = real_download
3133 else:
3134 self.report_file_already_downloaded(dl_filename)
3135
3136 dl_filename = dl_filename or temp_filename
3137 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3138
3139 except network_exceptions as err:
3140 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3141 return
3142 except OSError as err:
3143 raise UnavailableVideoError(err)
3144 except (ContentTooShortError, ) as err:
3145 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3146 return
3147
3148 if success and full_filename != '-':
3149
3150 def fixup():
3151 do_fixup = True
3152 fixup_policy = self.params.get('fixup')
3153 vid = info_dict['id']
3154
3155 if fixup_policy in ('ignore', 'never'):
3156 return
3157 elif fixup_policy == 'warn':
3158 do_fixup = 'warn'
3159 elif fixup_policy != 'force':
3160 assert fixup_policy in ('detect_or_warn', None)
3161 if not info_dict.get('__real_download'):
3162 do_fixup = False
3163
3164 def ffmpeg_fixup(cndn, msg, cls):
3165 if not (do_fixup and cndn):
3166 return
3167 elif do_fixup == 'warn':
3168 self.report_warning(f'{vid}: {msg}')
3169 return
3170 pp = cls(self)
3171 if pp.available:
3172 info_dict['__postprocessors'].append(pp)
3173 else:
3174 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3175
3176 stretched_ratio = info_dict.get('stretched_ratio')
3177 ffmpeg_fixup(
3178 stretched_ratio not in (1, None),
3179 f'Non-uniform pixel ratio {stretched_ratio}',
3180 FFmpegFixupStretchedPP)
3181
3182 ffmpeg_fixup(
3183 (info_dict.get('requested_formats') is None
3184 and info_dict.get('container') == 'm4a_dash'
3185 and info_dict.get('ext') == 'm4a'),
3186 'writing DASH m4a. Only some players support this container',
3187 FFmpegFixupM4aPP)
3188
3189 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3190 downloader = downloader.__name__ if downloader else None
3191
3192 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3193 ffmpeg_fixup(downloader == 'HlsFD',
3194 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3195 FFmpegFixupM3u8PP)
3196 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3197 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3198
3199 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3200 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed duration detected', FFmpegFixupDurationPP)
3201
3202 fixup()
3203 try:
3204 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3205 except PostProcessingError as err:
3206 self.report_error('Postprocessing: %s' % str(err))
3207 return
3208 try:
3209 for ph in self._post_hooks:
3210 ph(info_dict['filepath'])
3211 except Exception as err:
3212 self.report_error('post hooks: %s' % str(err))
3213 return
3214 info_dict['__write_download_archive'] = True
3215
3216 if self.params.get('force_write_download_archive'):
3217 info_dict['__write_download_archive'] = True
3218
3219 # Make sure the info_dict was modified in-place
3220 assert info_dict is original_infodict
3221
3222 max_downloads = self.params.get('max_downloads')
3223 if max_downloads is not None and self._num_downloads >= int(max_downloads):
3224 raise MaxDownloadsReached()
3225
3226 def __download_wrapper(self, func):
3227 @functools.wraps(func)
3228 def wrapper(*args, **kwargs):
3229 try:
3230 res = func(*args, **kwargs)
3231 except UnavailableVideoError as e:
3232 self.report_error(e)
3233 except MaxDownloadsReached as e:
3234 self.to_screen(f'[info] {e}')
3235 raise
3236 except DownloadCancelled as e:
3237 self.to_screen(f'[info] {e}')
3238 if not self.params.get('break_per_url'):
3239 raise
3240 else:
3241 if self.params.get('dump_single_json', False):
3242 self.post_extract(res)
3243 self.to_stdout(json.dumps(self.sanitize_info(res)))
3244 return wrapper
3245
3246 def download(self, url_list):
3247 """Download a given list of URLs."""
3248 url_list = variadic(url_list) # Passing a single URL is a common mistake
3249 outtmpl = self.outtmpl_dict['default']
3250 if (len(url_list) > 1
3251 and outtmpl != '-'
3252 and '%' not in outtmpl
3253 and self.params.get('max_downloads') != 1):
3254 raise SameFileError(outtmpl)
3255
3256 for url in url_list:
3257 self.__download_wrapper(self.extract_info)(
3258 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3259
3260 return self._download_retcode
3261
3262 def download_with_info_file(self, info_filename):
3263 with contextlib.closing(fileinput.FileInput(
3264 [info_filename], mode='r',
3265 openhook=fileinput.hook_encoded('utf-8'))) as f:
3266 # FileInput doesn't have a read method, we can't call json.load
3267 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3268 try:
3269 self.__download_wrapper(self.process_ie_result)(info, download=True)
3270 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3271 if not isinstance(e, EntryNotInPlaylist):
3272 self.to_stderr('\r')
3273 webpage_url = info.get('webpage_url')
3274 if webpage_url is not None:
3275 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3276 return self.download([webpage_url])
3277 else:
3278 raise
3279 return self._download_retcode
3280
3281 @staticmethod
3282 def sanitize_info(info_dict, remove_private_keys=False):
3283 ''' Sanitize the infodict for converting to json '''
3284 if info_dict is None:
3285 return info_dict
3286 info_dict.setdefault('epoch', int(time.time()))
3287 info_dict.setdefault('_type', 'video')
3288
3289 if remove_private_keys:
3290 reject = lambda k, v: v is None or k.startswith('__') or k in {
3291 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3292 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3293 }
3294 else:
3295 reject = lambda k, v: False
3296
3297 def filter_fn(obj):
3298 if isinstance(obj, dict):
3299 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3300 elif isinstance(obj, (list, tuple, set, LazyList)):
3301 return list(map(filter_fn, obj))
3302 elif obj is None or isinstance(obj, (str, int, float, bool)):
3303 return obj
3304 else:
3305 return repr(obj)
3306
3307 return filter_fn(info_dict)
3308
3309 @staticmethod
3310 def filter_requested_info(info_dict, actually_filter=True):
3311 ''' Alias of sanitize_info for backward compatibility '''
3312 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3313
3314 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3315 for filename in set(filter(None, files_to_delete)):
3316 if msg:
3317 self.to_screen(msg % filename)
3318 try:
3319 os.remove(filename)
3320 except OSError:
3321 self.report_warning(f'Unable to delete file {filename}')
3322 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3323 del info['__files_to_move'][filename]
3324
3325 @staticmethod
3326 def post_extract(info_dict):
3327 def actual_post_extract(info_dict):
3328 if info_dict.get('_type') in ('playlist', 'multi_video'):
3329 for video_dict in info_dict.get('entries', {}):
3330 actual_post_extract(video_dict or {})
3331 return
3332
3333 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3334 info_dict.update(post_extractor())
3335
3336 actual_post_extract(info_dict or {})
3337
3338 def run_pp(self, pp, infodict):
3339 files_to_delete = []
3340 if '__files_to_move' not in infodict:
3341 infodict['__files_to_move'] = {}
3342 try:
3343 files_to_delete, infodict = pp.run(infodict)
3344 except PostProcessingError as e:
3345 # Must be True and not 'only_download'
3346 if self.params.get('ignoreerrors') is True:
3347 self.report_error(e)
3348 return infodict
3349 raise
3350
3351 if not files_to_delete:
3352 return infodict
3353 if self.params.get('keepvideo', False):
3354 for f in files_to_delete:
3355 infodict['__files_to_move'].setdefault(f, '')
3356 else:
3357 self._delete_downloaded_files(
3358 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3359 return infodict
3360
3361 def run_all_pps(self, key, info, *, additional_pps=None):
3362 self._forceprint(key, info)
3363 for pp in (additional_pps or []) + self._pps[key]:
3364 info = self.run_pp(pp, info)
3365 return info
3366
3367 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3368 info = dict(ie_info)
3369 info['__files_to_move'] = files_to_move or {}
3370 info = self.run_all_pps(key, info)
3371 return info, info.pop('__files_to_move', None)
3372
3373 def post_process(self, filename, info, files_to_move=None):
3374 """Run all the postprocessors on the given file."""
3375 info['filepath'] = filename
3376 info['__files_to_move'] = files_to_move or {}
3377 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3378 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3379 del info['__files_to_move']
3380 return self.run_all_pps('after_move', info)
3381
3382 def _make_archive_id(self, info_dict):
3383 video_id = info_dict.get('id')
3384 if not video_id:
3385 return
3386 # Future-proof against any change in case
3387 # and backwards compatibility with prior versions
3388 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3389 if extractor is None:
3390 url = str_or_none(info_dict.get('url'))
3391 if not url:
3392 return
3393 # Try to find matching extractor for the URL and take its ie_key
3394 for ie_key, ie in self._ies.items():
3395 if ie.suitable(url):
3396 extractor = ie_key
3397 break
3398 else:
3399 return
3400 return f'{extractor.lower()} {video_id}'
3401
3402 def in_download_archive(self, info_dict):
3403 fn = self.params.get('download_archive')
3404 if fn is None:
3405 return False
3406
3407 vid_id = self._make_archive_id(info_dict)
3408 if not vid_id:
3409 return False # Incomplete video information
3410
3411 return vid_id in self.archive
3412
3413 def record_download_archive(self, info_dict):
3414 fn = self.params.get('download_archive')
3415 if fn is None:
3416 return
3417 vid_id = self._make_archive_id(info_dict)
3418 assert vid_id
3419 self.write_debug(f'Adding to archive: {vid_id}')
3420 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3421 archive_file.write(vid_id + '\n')
3422 self.archive.add(vid_id)
3423
3424 @staticmethod
3425 def format_resolution(format, default='unknown'):
3426 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3427 return 'audio only'
3428 if format.get('resolution') is not None:
3429 return format['resolution']
3430 if format.get('width') and format.get('height'):
3431 return '%dx%d' % (format['width'], format['height'])
3432 elif format.get('height'):
3433 return '%sp' % format['height']
3434 elif format.get('width'):
3435 return '%dx?' % format['width']
3436 return default
3437
3438 def _list_format_headers(self, *headers):
3439 if self.params.get('listformats_table', True) is not False:
3440 return [self._format_screen(header, self.Styles.HEADERS) for header in headers]
3441 return headers
3442
3443 def _format_note(self, fdict):
3444 res = ''
3445 if fdict.get('ext') in ['f4f', 'f4m']:
3446 res += '(unsupported)'
3447 if fdict.get('language'):
3448 if res:
3449 res += ' '
3450 res += '[%s]' % fdict['language']
3451 if fdict.get('format_note') is not None:
3452 if res:
3453 res += ' '
3454 res += fdict['format_note']
3455 if fdict.get('tbr') is not None:
3456 if res:
3457 res += ', '
3458 res += '%4dk' % fdict['tbr']
3459 if fdict.get('container') is not None:
3460 if res:
3461 res += ', '
3462 res += '%s container' % fdict['container']
3463 if (fdict.get('vcodec') is not None
3464 and fdict.get('vcodec') != 'none'):
3465 if res:
3466 res += ', '
3467 res += fdict['vcodec']
3468 if fdict.get('vbr') is not None:
3469 res += '@'
3470 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3471 res += 'video@'
3472 if fdict.get('vbr') is not None:
3473 res += '%4dk' % fdict['vbr']
3474 if fdict.get('fps') is not None:
3475 if res:
3476 res += ', '
3477 res += '%sfps' % fdict['fps']
3478 if fdict.get('acodec') is not None:
3479 if res:
3480 res += ', '
3481 if fdict['acodec'] == 'none':
3482 res += 'video only'
3483 else:
3484 res += '%-5s' % fdict['acodec']
3485 elif fdict.get('abr') is not None:
3486 if res:
3487 res += ', '
3488 res += 'audio'
3489 if fdict.get('abr') is not None:
3490 res += '@%3dk' % fdict['abr']
3491 if fdict.get('asr') is not None:
3492 res += ' (%5dHz)' % fdict['asr']
3493 if fdict.get('filesize') is not None:
3494 if res:
3495 res += ', '
3496 res += format_bytes(fdict['filesize'])
3497 elif fdict.get('filesize_approx') is not None:
3498 if res:
3499 res += ', '
3500 res += '~' + format_bytes(fdict['filesize_approx'])
3501 return res
3502
3503 def render_formats_table(self, info_dict):
3504 if not info_dict.get('formats') and not info_dict.get('url'):
3505 return None
3506
3507 formats = info_dict.get('formats', [info_dict])
3508 if not self.params.get('listformats_table', True) is not False:
3509 table = [
3510 [
3511 format_field(f, 'format_id'),
3512 format_field(f, 'ext'),
3513 self.format_resolution(f),
3514 self._format_note(f)
3515 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3516 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3517
3518 delim = self._format_screen('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3519 table = [
3520 [
3521 self._format_screen(format_field(f, 'format_id'), self.Styles.ID),
3522 format_field(f, 'ext'),
3523 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3524 format_field(f, 'fps', '\t%d'),
3525 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3526 delim,
3527 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3528 format_field(f, 'tbr', '\t%dk'),
3529 shorten_protocol_name(f.get('protocol', '')),
3530 delim,
3531 format_field(f, 'vcodec', default='unknown').replace(
3532 'none', 'images' if f.get('acodec') == 'none'
3533 else self._format_screen('audio only', self.Styles.SUPPRESS)),
3534 format_field(f, 'vbr', '\t%dk'),
3535 format_field(f, 'acodec', default='unknown').replace(
3536 'none', '' if f.get('vcodec') == 'none'
3537 else self._format_screen('video only', self.Styles.SUPPRESS)),
3538 format_field(f, 'abr', '\t%dk'),
3539 format_field(f, 'asr', '\t%dHz'),
3540 join_nonempty(
3541 self._format_screen('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3542 format_field(f, 'language', '[%s]'),
3543 join_nonempty(format_field(f, 'format_note'),
3544 format_field(f, 'container', ignore=(None, f.get('ext'))),
3545 delim=', '),
3546 delim=' '),
3547 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3548 header_line = self._list_format_headers(
3549 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3550 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3551
3552 return render_table(
3553 header_line, table, hide_empty=True,
3554 delim=self._format_screen('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3555
3556 def render_thumbnails_table(self, info_dict):
3557 thumbnails = list(info_dict.get('thumbnails') or [])
3558 if not thumbnails:
3559 return None
3560 return render_table(
3561 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3562 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3563
3564 def render_subtitles_table(self, video_id, subtitles):
3565 def _row(lang, formats):
3566 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3567 if len(set(names)) == 1:
3568 names = [] if names[0] == 'unknown' else names[:1]
3569 return [lang, ', '.join(names), ', '.join(exts)]
3570
3571 if not subtitles:
3572 return None
3573 return render_table(
3574 self._list_format_headers('Language', 'Name', 'Formats'),
3575 [_row(lang, formats) for lang, formats in subtitles.items()],
3576 hide_empty=True)
3577
3578 def __list_table(self, video_id, name, func, *args):
3579 table = func(*args)
3580 if not table:
3581 self.to_screen(f'{video_id} has no {name}')
3582 return
3583 self.to_screen(f'[info] Available {name} for {video_id}:')
3584 self.to_stdout(table)
3585
3586 def list_formats(self, info_dict):
3587 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3588
3589 def list_thumbnails(self, info_dict):
3590 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3591
3592 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3593 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3594
3595 def urlopen(self, req):
3596 """ Start an HTTP download """
3597 if isinstance(req, str):
3598 req = sanitized_Request(req)
3599 return self._opener.open(req, timeout=self._socket_timeout)
3600
3601 def print_debug_header(self):
3602 if not self.params.get('verbose'):
3603 return
3604
3605 def get_encoding(stream):
3606 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3607 if not supports_terminal_sequences(stream):
3608 from .compat import WINDOWS_VT_MODE # Must be imported locally
3609 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3610 return ret
3611
3612 encoding_str = 'Encodings: locale %s, fs %s, out %s, err %s, pref %s' % (
3613 locale.getpreferredencoding(),
3614 sys.getfilesystemencoding(),
3615 get_encoding(self._out_files['screen']), get_encoding(self._out_files['error']),
3616 self.get_encoding())
3617
3618 logger = self.params.get('logger')
3619 if logger:
3620 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3621 write_debug(encoding_str)
3622 else:
3623 write_string(f'[debug] {encoding_str}\n', encoding=None)
3624 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3625
3626 source = detect_variant()
3627 write_debug(join_nonempty(
3628 'yt-dlp version', __version__,
3629 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3630 '' if source == 'unknown' else f'({source})',
3631 delim=' '))
3632 if not _LAZY_LOADER:
3633 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3634 write_debug('Lazy loading extractors is forcibly disabled')
3635 else:
3636 write_debug('Lazy loading extractors is disabled')
3637 if plugin_extractors or plugin_postprocessors:
3638 write_debug('Plugins: %s' % [
3639 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3640 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3641 if self.params.get('compat_opts'):
3642 write_debug('Compatibility options: %s' % ', '.join(self.params.get('compat_opts')))
3643
3644 if source == 'source':
3645 try:
3646 sp = Popen(
3647 ['git', 'rev-parse', '--short', 'HEAD'],
3648 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3649 cwd=os.path.dirname(os.path.abspath(__file__)))
3650 out, err = sp.communicate_or_kill()
3651 out = out.decode().strip()
3652 if re.match('[0-9a-f]+', out):
3653 write_debug('Git HEAD: %s' % out)
3654 except Exception:
3655 with contextlib.suppress(Exception):
3656 sys.exc_clear()
3657
3658 def python_implementation():
3659 impl_name = platform.python_implementation()
3660 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3661 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3662 return impl_name
3663
3664 write_debug('Python version %s (%s %s) - %s' % (
3665 platform.python_version(),
3666 python_implementation(),
3667 platform.architecture()[0],
3668 platform_name()))
3669
3670 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3671 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3672 if ffmpeg_features:
3673 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3674
3675 exe_versions['rtmpdump'] = rtmpdump_version()
3676 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3677 exe_str = ', '.join(
3678 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3679 ) or 'none'
3680 write_debug('exe versions: %s' % exe_str)
3681
3682 from .compat.compat_utils import get_package_info
3683 from .dependencies import available_dependencies
3684
3685 write_debug('Optional libraries: %s' % (', '.join(sorted({
3686 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3687 })) or 'none'))
3688
3689 self._setup_opener()
3690 proxy_map = {}
3691 for handler in self._opener.handlers:
3692 if hasattr(handler, 'proxies'):
3693 proxy_map.update(handler.proxies)
3694 write_debug(f'Proxy map: {proxy_map}')
3695
3696 # Not implemented
3697 if False and self.params.get('call_home'):
3698 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3699 write_debug('Public IP address: %s' % ipaddr)
3700 latest_version = self.urlopen(
3701 'https://yt-dl.org/latest/version').read().decode('utf-8')
3702 if version_tuple(latest_version) > version_tuple(__version__):
3703 self.report_warning(
3704 'You are using an outdated version (newest version: %s)! '
3705 'See https://yt-dl.org/update if you need help updating.' %
3706 latest_version)
3707
3708 def _setup_opener(self):
3709 if hasattr(self, '_opener'):
3710 return
3711 timeout_val = self.params.get('socket_timeout')
3712 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3713
3714 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3715 opts_cookiefile = self.params.get('cookiefile')
3716 opts_proxy = self.params.get('proxy')
3717
3718 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3719
3720 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3721 if opts_proxy is not None:
3722 if opts_proxy == '':
3723 proxies = {}
3724 else:
3725 proxies = {'http': opts_proxy, 'https': opts_proxy}
3726 else:
3727 proxies = compat_urllib_request.getproxies()
3728 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3729 if 'http' in proxies and 'https' not in proxies:
3730 proxies['https'] = proxies['http']
3731 proxy_handler = PerRequestProxyHandler(proxies)
3732
3733 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3734 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3735 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3736 redirect_handler = YoutubeDLRedirectHandler()
3737 data_handler = urllib.request.DataHandler()
3738
3739 # When passing our own FileHandler instance, build_opener won't add the
3740 # default FileHandler and allows us to disable the file protocol, which
3741 # can be used for malicious purposes (see
3742 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3743 file_handler = compat_urllib_request.FileHandler()
3744
3745 def file_open(*args, **kwargs):
3746 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3747 file_handler.file_open = file_open
3748
3749 opener = compat_urllib_request.build_opener(
3750 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3751
3752 # Delete the default user-agent header, which would otherwise apply in
3753 # cases where our custom HTTP handler doesn't come into play
3754 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3755 opener.addheaders = []
3756 self._opener = opener
3757
3758 def encode(self, s):
3759 if isinstance(s, bytes):
3760 return s # Already encoded
3761
3762 try:
3763 return s.encode(self.get_encoding())
3764 except UnicodeEncodeError as err:
3765 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3766 raise
3767
3768 def get_encoding(self):
3769 encoding = self.params.get('encoding')
3770 if encoding is None:
3771 encoding = preferredencoding()
3772 return encoding
3773
3774 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3775 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3776 if overwrite is None:
3777 overwrite = self.params.get('overwrites', True)
3778 if not self.params.get('writeinfojson'):
3779 return False
3780 elif not infofn:
3781 self.write_debug(f'Skipping writing {label} infojson')
3782 return False
3783 elif not self._ensure_dir_exists(infofn):
3784 return None
3785 elif not overwrite and os.path.exists(infofn):
3786 self.to_screen(f'[info] {label.title()} metadata is already present')
3787 return 'exists'
3788
3789 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3790 try:
3791 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3792 return True
3793 except OSError:
3794 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3795 return None
3796
3797 def _write_description(self, label, ie_result, descfn):
3798 ''' Write description and returns True = written, False = skip, None = error '''
3799 if not self.params.get('writedescription'):
3800 return False
3801 elif not descfn:
3802 self.write_debug(f'Skipping writing {label} description')
3803 return False
3804 elif not self._ensure_dir_exists(descfn):
3805 return None
3806 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3807 self.to_screen(f'[info] {label.title()} description is already present')
3808 elif ie_result.get('description') is None:
3809 self.report_warning(f'There\'s no {label} description to write')
3810 return False
3811 else:
3812 try:
3813 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3814 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3815 descfile.write(ie_result['description'])
3816 except OSError:
3817 self.report_error(f'Cannot write {label} description file {descfn}')
3818 return None
3819 return True
3820
3821 def _write_subtitles(self, info_dict, filename):
3822 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3823 ret = []
3824 subtitles = info_dict.get('requested_subtitles')
3825 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3826 # subtitles download errors are already managed as troubles in relevant IE
3827 # that way it will silently go on when used with unsupporting IE
3828 return ret
3829
3830 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3831 if not sub_filename_base:
3832 self.to_screen('[info] Skipping writing video subtitles')
3833 return ret
3834 for sub_lang, sub_info in subtitles.items():
3835 sub_format = sub_info['ext']
3836 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3837 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3838 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3839 if existing_sub:
3840 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3841 sub_info['filepath'] = existing_sub
3842 ret.append((existing_sub, sub_filename_final))
3843 continue
3844
3845 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3846 if sub_info.get('data') is not None:
3847 try:
3848 # Use newline='' to prevent conversion of newline characters
3849 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3850 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3851 subfile.write(sub_info['data'])
3852 sub_info['filepath'] = sub_filename
3853 ret.append((sub_filename, sub_filename_final))
3854 continue
3855 except OSError:
3856 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3857 return None
3858
3859 try:
3860 sub_copy = sub_info.copy()
3861 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3862 self.dl(sub_filename, sub_copy, subtitle=True)
3863 sub_info['filepath'] = sub_filename
3864 ret.append((sub_filename, sub_filename_final))
3865 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3866 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3867 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3868 if not self.params.get('ignoreerrors'):
3869 self.report_error(msg)
3870 raise DownloadError(msg)
3871 self.report_warning(msg)
3872 return ret
3873
3874 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3875 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3876 write_all = self.params.get('write_all_thumbnails', False)
3877 thumbnails, ret = [], []
3878 if write_all or self.params.get('writethumbnail', False):
3879 thumbnails = info_dict.get('thumbnails') or []
3880 multiple = write_all and len(thumbnails) > 1
3881
3882 if thumb_filename_base is None:
3883 thumb_filename_base = filename
3884 if thumbnails and not thumb_filename_base:
3885 self.write_debug(f'Skipping writing {label} thumbnail')
3886 return ret
3887
3888 for idx, t in list(enumerate(thumbnails))[::-1]:
3889 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3890 thumb_display_id = f'{label} thumbnail {t["id"]}'
3891 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3892 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3893
3894 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3895 if existing_thumb:
3896 self.to_screen('[info] %s is already present' % (
3897 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3898 t['filepath'] = existing_thumb
3899 ret.append((existing_thumb, thumb_filename_final))
3900 else:
3901 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3902 try:
3903 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3904 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3905 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3906 shutil.copyfileobj(uf, thumbf)
3907 ret.append((thumb_filename, thumb_filename_final))
3908 t['filepath'] = thumb_filename
3909 except network_exceptions as err:
3910 thumbnails.pop(idx)
3911 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3912 if ret and not write_all:
3913 break
3914 return ret