]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[cleanup] Misc cleanup
[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 or text stream from where cookies should be read 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())
777 self._output_process.stdin.flush()
778 res = ''.join(self._output_channel.readline().decode()
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()
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 elif ret is not None:
1326 return ret
1327 return None
1328
1329 if self.in_download_archive(info_dict):
1330 reason = '%s has already been recorded in the archive' % video_title
1331 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1332 else:
1333 reason = check_filter()
1334 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1335 if reason is not None:
1336 if not silent:
1337 self.to_screen('[download] ' + reason)
1338 if self.params.get(break_opt, False):
1339 raise break_err()
1340 return reason
1341
1342 @staticmethod
1343 def add_extra_info(info_dict, extra_info):
1344 '''Set the keys from extra_info in info dict if they are missing'''
1345 for key, value in extra_info.items():
1346 info_dict.setdefault(key, value)
1347
1348 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1349 process=True, force_generic_extractor=False):
1350 """
1351 Return a list with a dictionary for each video extracted.
1352
1353 Arguments:
1354 url -- URL to extract
1355
1356 Keyword arguments:
1357 download -- whether to download videos during extraction
1358 ie_key -- extractor key hint
1359 extra_info -- dictionary containing the extra values to add to each result
1360 process -- whether to resolve all unresolved references (URLs, playlist items),
1361 must be True for download to work.
1362 force_generic_extractor -- force using the generic extractor
1363 """
1364
1365 if extra_info is None:
1366 extra_info = {}
1367
1368 if not ie_key and force_generic_extractor:
1369 ie_key = 'Generic'
1370
1371 if ie_key:
1372 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1373 else:
1374 ies = self._ies
1375
1376 for ie_key, ie in ies.items():
1377 if not ie.suitable(url):
1378 continue
1379
1380 if not ie.working():
1381 self.report_warning('The program functionality for this site has been marked as broken, '
1382 'and will probably not work.')
1383
1384 temp_id = ie.get_temp_id(url)
1385 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1386 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1387 if self.params.get('break_on_existing', False):
1388 raise ExistingVideoReached()
1389 break
1390 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1391 else:
1392 self.report_error('no suitable InfoExtractor for URL %s' % url)
1393
1394 def __handle_extraction_exceptions(func):
1395 @functools.wraps(func)
1396 def wrapper(self, *args, **kwargs):
1397 while True:
1398 try:
1399 return func(self, *args, **kwargs)
1400 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1401 raise
1402 except ReExtractInfo as e:
1403 if e.expected:
1404 self.to_screen(f'{e}; Re-extracting data')
1405 else:
1406 self.to_stderr('\r')
1407 self.report_warning(f'{e}; Re-extracting data')
1408 continue
1409 except GeoRestrictedError as e:
1410 msg = e.msg
1411 if e.countries:
1412 msg += '\nThis video is available in %s.' % ', '.join(
1413 map(ISO3166Utils.short2full, e.countries))
1414 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1415 self.report_error(msg)
1416 except ExtractorError as e: # An error we somewhat expected
1417 self.report_error(str(e), e.format_traceback())
1418 except Exception as e:
1419 if self.params.get('ignoreerrors'):
1420 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1421 else:
1422 raise
1423 break
1424 return wrapper
1425
1426 def _wait_for_video(self, ie_result):
1427 if (not self.params.get('wait_for_video')
1428 or ie_result.get('_type', 'video') != 'video'
1429 or ie_result.get('formats') or ie_result.get('url')):
1430 return
1431
1432 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1433 last_msg = ''
1434
1435 def progress(msg):
1436 nonlocal last_msg
1437 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1438 last_msg = msg
1439
1440 min_wait, max_wait = self.params.get('wait_for_video')
1441 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1442 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1443 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1444 self.report_warning('Release time of video is not known')
1445 elif (diff or 0) <= 0:
1446 self.report_warning('Video should already be available according to extracted info')
1447 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1448 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1449
1450 wait_till = time.time() + diff
1451 try:
1452 while True:
1453 diff = wait_till - time.time()
1454 if diff <= 0:
1455 progress('')
1456 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1457 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1458 time.sleep(1)
1459 except KeyboardInterrupt:
1460 progress('')
1461 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1462 except BaseException as e:
1463 if not isinstance(e, ReExtractInfo):
1464 self.to_screen('')
1465 raise
1466
1467 @__handle_extraction_exceptions
1468 def __extract_info(self, url, ie, download, extra_info, process):
1469 ie_result = ie.extract(url)
1470 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1471 return
1472 if isinstance(ie_result, list):
1473 # Backwards compatibility: old IE result format
1474 ie_result = {
1475 '_type': 'compat_list',
1476 'entries': ie_result,
1477 }
1478 if extra_info.get('original_url'):
1479 ie_result.setdefault('original_url', extra_info['original_url'])
1480 self.add_default_extra_info(ie_result, ie, url)
1481 if process:
1482 self._wait_for_video(ie_result)
1483 return self.process_ie_result(ie_result, download, extra_info)
1484 else:
1485 return ie_result
1486
1487 def add_default_extra_info(self, ie_result, ie, url):
1488 if url is not None:
1489 self.add_extra_info(ie_result, {
1490 'webpage_url': url,
1491 'original_url': url,
1492 })
1493 webpage_url = ie_result.get('webpage_url')
1494 if webpage_url:
1495 self.add_extra_info(ie_result, {
1496 'webpage_url_basename': url_basename(webpage_url),
1497 'webpage_url_domain': get_domain(webpage_url),
1498 })
1499 if ie is not None:
1500 self.add_extra_info(ie_result, {
1501 'extractor': ie.IE_NAME,
1502 'extractor_key': ie.ie_key(),
1503 })
1504
1505 def process_ie_result(self, ie_result, download=True, extra_info=None):
1506 """
1507 Take the result of the ie(may be modified) and resolve all unresolved
1508 references (URLs, playlist items).
1509
1510 It will also download the videos if 'download'.
1511 Returns the resolved ie_result.
1512 """
1513 if extra_info is None:
1514 extra_info = {}
1515 result_type = ie_result.get('_type', 'video')
1516
1517 if result_type in ('url', 'url_transparent'):
1518 ie_result['url'] = sanitize_url(ie_result['url'])
1519 if ie_result.get('original_url'):
1520 extra_info.setdefault('original_url', ie_result['original_url'])
1521
1522 extract_flat = self.params.get('extract_flat', False)
1523 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1524 or extract_flat is True):
1525 info_copy = ie_result.copy()
1526 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1527 if ie and not ie_result.get('id'):
1528 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1529 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1530 self.add_extra_info(info_copy, extra_info)
1531 info_copy, _ = self.pre_process(info_copy)
1532 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1533 if self.params.get('force_write_download_archive', False):
1534 self.record_download_archive(info_copy)
1535 return ie_result
1536
1537 if result_type == 'video':
1538 self.add_extra_info(ie_result, extra_info)
1539 ie_result = self.process_video_result(ie_result, download=download)
1540 additional_urls = (ie_result or {}).get('additional_urls')
1541 if additional_urls:
1542 # TODO: Improve MetadataParserPP to allow setting a list
1543 if isinstance(additional_urls, compat_str):
1544 additional_urls = [additional_urls]
1545 self.to_screen(
1546 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1547 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1548 ie_result['additional_entries'] = [
1549 self.extract_info(
1550 url, download, extra_info=extra_info,
1551 force_generic_extractor=self.params.get('force_generic_extractor'))
1552 for url in additional_urls
1553 ]
1554 return ie_result
1555 elif result_type == 'url':
1556 # We have to add extra_info to the results because it may be
1557 # contained in a playlist
1558 return self.extract_info(
1559 ie_result['url'], download,
1560 ie_key=ie_result.get('ie_key'),
1561 extra_info=extra_info)
1562 elif result_type == 'url_transparent':
1563 # Use the information from the embedding page
1564 info = self.extract_info(
1565 ie_result['url'], ie_key=ie_result.get('ie_key'),
1566 extra_info=extra_info, download=False, process=False)
1567
1568 # extract_info may return None when ignoreerrors is enabled and
1569 # extraction failed with an error, don't crash and return early
1570 # in this case
1571 if not info:
1572 return info
1573
1574 new_result = info.copy()
1575 new_result.update(filter_dict(ie_result, lambda k, v: (
1576 v is not None and k not in {'_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'})))
1577
1578 # Extracted info may not be a video result (i.e.
1579 # info.get('_type', 'video') != video) but rather an url or
1580 # url_transparent. In such cases outer metadata (from ie_result)
1581 # should be propagated to inner one (info). For this to happen
1582 # _type of info should be overridden with url_transparent. This
1583 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1584 if new_result.get('_type') == 'url':
1585 new_result['_type'] = 'url_transparent'
1586
1587 return self.process_ie_result(
1588 new_result, download=download, extra_info=extra_info)
1589 elif result_type in ('playlist', 'multi_video'):
1590 # Protect from infinite recursion due to recursively nested playlists
1591 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1592 webpage_url = ie_result['webpage_url']
1593 if webpage_url in self._playlist_urls:
1594 self.to_screen(
1595 '[download] Skipping already downloaded playlist: %s'
1596 % ie_result.get('title') or ie_result.get('id'))
1597 return
1598
1599 self._playlist_level += 1
1600 self._playlist_urls.add(webpage_url)
1601 self._fill_common_fields(ie_result, False)
1602 self._sanitize_thumbnails(ie_result)
1603 try:
1604 return self.__process_playlist(ie_result, download)
1605 finally:
1606 self._playlist_level -= 1
1607 if not self._playlist_level:
1608 self._playlist_urls.clear()
1609 elif result_type == 'compat_list':
1610 self.report_warning(
1611 'Extractor %s returned a compat_list result. '
1612 'It needs to be updated.' % ie_result.get('extractor'))
1613
1614 def _fixup(r):
1615 self.add_extra_info(r, {
1616 'extractor': ie_result['extractor'],
1617 'webpage_url': ie_result['webpage_url'],
1618 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1619 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1620 'extractor_key': ie_result['extractor_key'],
1621 })
1622 return r
1623 ie_result['entries'] = [
1624 self.process_ie_result(_fixup(r), download, extra_info)
1625 for r in ie_result['entries']
1626 ]
1627 return ie_result
1628 else:
1629 raise Exception('Invalid result type: %s' % result_type)
1630
1631 def _ensure_dir_exists(self, path):
1632 return make_dir(path, self.report_error)
1633
1634 @staticmethod
1635 def _playlist_infodict(ie_result, **kwargs):
1636 return {
1637 **ie_result,
1638 'playlist': ie_result.get('title') or ie_result.get('id'),
1639 'playlist_id': ie_result.get('id'),
1640 'playlist_title': ie_result.get('title'),
1641 'playlist_uploader': ie_result.get('uploader'),
1642 'playlist_uploader_id': ie_result.get('uploader_id'),
1643 'playlist_index': 0,
1644 **kwargs,
1645 }
1646
1647 def __process_playlist(self, ie_result, download):
1648 # We process each entry in the playlist
1649 playlist = ie_result.get('title') or ie_result.get('id')
1650 self.to_screen('[download] Downloading playlist: %s' % playlist)
1651
1652 if 'entries' not in ie_result:
1653 raise EntryNotInPlaylist('There are no entries')
1654
1655 MissingEntry = object()
1656 incomplete_entries = bool(ie_result.get('requested_entries'))
1657 if incomplete_entries:
1658 def fill_missing_entries(entries, indices):
1659 ret = [MissingEntry] * max(indices)
1660 for i, entry in zip(indices, entries):
1661 ret[i - 1] = entry
1662 return ret
1663 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1664
1665 playlist_results = []
1666
1667 playliststart = self.params.get('playliststart', 1)
1668 playlistend = self.params.get('playlistend')
1669 # For backwards compatibility, interpret -1 as whole list
1670 if playlistend == -1:
1671 playlistend = None
1672
1673 playlistitems_str = self.params.get('playlist_items')
1674 playlistitems = None
1675 if playlistitems_str is not None:
1676 def iter_playlistitems(format):
1677 for string_segment in format.split(','):
1678 if '-' in string_segment:
1679 start, end = string_segment.split('-')
1680 for item in range(int(start), int(end) + 1):
1681 yield int(item)
1682 else:
1683 yield int(string_segment)
1684 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1685
1686 ie_entries = ie_result['entries']
1687 if isinstance(ie_entries, list):
1688 playlist_count = len(ie_entries)
1689 msg = f'Collected {playlist_count} videos; downloading %d of them'
1690 ie_result['playlist_count'] = ie_result.get('playlist_count') or playlist_count
1691
1692 def get_entry(i):
1693 return ie_entries[i - 1]
1694 else:
1695 msg = 'Downloading %d videos'
1696 if not isinstance(ie_entries, (PagedList, LazyList)):
1697 ie_entries = LazyList(ie_entries)
1698 elif isinstance(ie_entries, InAdvancePagedList):
1699 if ie_entries._pagesize == 1:
1700 playlist_count = ie_entries._pagecount
1701
1702 def get_entry(i):
1703 return YoutubeDL.__handle_extraction_exceptions(
1704 lambda self, i: ie_entries[i - 1]
1705 )(self, i)
1706
1707 entries, broken = [], False
1708 items = playlistitems if playlistitems is not None else itertools.count(playliststart)
1709 for i in items:
1710 if i == 0:
1711 continue
1712 if playlistitems is None and playlistend is not None and playlistend < i:
1713 break
1714 entry = None
1715 try:
1716 entry = get_entry(i)
1717 if entry is MissingEntry:
1718 raise EntryNotInPlaylist()
1719 except (IndexError, EntryNotInPlaylist):
1720 if incomplete_entries:
1721 raise EntryNotInPlaylist(f'Entry {i} cannot be found')
1722 elif not playlistitems:
1723 break
1724 entries.append(entry)
1725 try:
1726 if entry is not None:
1727 # TODO: Add auto-generated fields
1728 self._match_entry(entry, incomplete=True, silent=True)
1729 except (ExistingVideoReached, RejectedVideoReached):
1730 broken = True
1731 break
1732 ie_result['entries'] = entries
1733
1734 # Save playlist_index before re-ordering
1735 entries = [
1736 ((playlistitems[i - 1] if playlistitems else i + playliststart - 1), entry)
1737 for i, entry in enumerate(entries, 1)
1738 if entry is not None]
1739 n_entries = len(entries)
1740
1741 if not (ie_result.get('playlist_count') or broken or playlistitems or playlistend):
1742 ie_result['playlist_count'] = n_entries
1743
1744 if not playlistitems and (playliststart != 1 or playlistend):
1745 playlistitems = list(range(playliststart, playliststart + n_entries))
1746 ie_result['requested_entries'] = playlistitems
1747
1748 _infojson_written = False
1749 write_playlist_files = self.params.get('allow_playlist_files', True)
1750 if write_playlist_files and self.params.get('list_thumbnails'):
1751 self.list_thumbnails(ie_result)
1752 if write_playlist_files and not self.params.get('simulate'):
1753 ie_copy = self._playlist_infodict(ie_result, n_entries=n_entries)
1754 _infojson_written = self._write_info_json(
1755 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1756 if _infojson_written is None:
1757 return
1758 if self._write_description('playlist', ie_result,
1759 self.prepare_filename(ie_copy, 'pl_description')) is None:
1760 return
1761 # TODO: This should be passed to ThumbnailsConvertor if necessary
1762 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1763
1764 if self.params.get('playlistreverse', False):
1765 entries = entries[::-1]
1766 if self.params.get('playlistrandom', False):
1767 random.shuffle(entries)
1768
1769 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1770
1771 self.to_screen(f'[{ie_result["extractor"]}] playlist {playlist}: {msg % n_entries}')
1772 failures = 0
1773 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1774 for i, entry_tuple in enumerate(entries, 1):
1775 playlist_index, entry = entry_tuple
1776 if 'playlist-index' in self.params.get('compat_opts', []):
1777 playlist_index = playlistitems[i - 1] if playlistitems else i + playliststart - 1
1778 self.to_screen('[download] Downloading video %s of %s' % (
1779 self._format_screen(i, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1780 # This __x_forwarded_for_ip thing is a bit ugly but requires
1781 # minimal changes
1782 if x_forwarded_for:
1783 entry['__x_forwarded_for_ip'] = x_forwarded_for
1784 extra = {
1785 'n_entries': n_entries,
1786 '__last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1787 'playlist_count': ie_result.get('playlist_count'),
1788 'playlist_index': playlist_index,
1789 'playlist_autonumber': i,
1790 'playlist': playlist,
1791 'playlist_id': ie_result.get('id'),
1792 'playlist_title': ie_result.get('title'),
1793 'playlist_uploader': ie_result.get('uploader'),
1794 'playlist_uploader_id': ie_result.get('uploader_id'),
1795 'extractor': ie_result['extractor'],
1796 'webpage_url': ie_result['webpage_url'],
1797 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1798 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1799 'extractor_key': ie_result['extractor_key'],
1800 }
1801
1802 if self._match_entry(entry, incomplete=True) is not None:
1803 continue
1804
1805 entry_result = self.__process_iterable_entry(entry, download, extra)
1806 if not entry_result:
1807 failures += 1
1808 if failures >= max_failures:
1809 self.report_error(
1810 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1811 break
1812 playlist_results.append(entry_result)
1813 ie_result['entries'] = playlist_results
1814
1815 # Write the updated info to json
1816 if _infojson_written is True and self._write_info_json(
1817 'updated playlist', ie_result,
1818 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1819 return
1820
1821 ie_result = self.run_all_pps('playlist', ie_result)
1822 self.to_screen(f'[download] Finished downloading playlist: {playlist}')
1823 return ie_result
1824
1825 @__handle_extraction_exceptions
1826 def __process_iterable_entry(self, entry, download, extra_info):
1827 return self.process_ie_result(
1828 entry, download=download, extra_info=extra_info)
1829
1830 def _build_format_filter(self, filter_spec):
1831 " Returns a function to filter the formats according to the filter_spec "
1832
1833 OPERATORS = {
1834 '<': operator.lt,
1835 '<=': operator.le,
1836 '>': operator.gt,
1837 '>=': operator.ge,
1838 '=': operator.eq,
1839 '!=': operator.ne,
1840 }
1841 operator_rex = re.compile(r'''(?x)\s*
1842 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1843 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1844 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1845 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1846 m = operator_rex.fullmatch(filter_spec)
1847 if m:
1848 try:
1849 comparison_value = int(m.group('value'))
1850 except ValueError:
1851 comparison_value = parse_filesize(m.group('value'))
1852 if comparison_value is None:
1853 comparison_value = parse_filesize(m.group('value') + 'B')
1854 if comparison_value is None:
1855 raise ValueError(
1856 'Invalid value %r in format specification %r' % (
1857 m.group('value'), filter_spec))
1858 op = OPERATORS[m.group('op')]
1859
1860 if not m:
1861 STR_OPERATORS = {
1862 '=': operator.eq,
1863 '^=': lambda attr, value: attr.startswith(value),
1864 '$=': lambda attr, value: attr.endswith(value),
1865 '*=': lambda attr, value: value in attr,
1866 '~=': lambda attr, value: value.search(attr) is not None
1867 }
1868 str_operator_rex = re.compile(r'''(?x)\s*
1869 (?P<key>[a-zA-Z0-9._-]+)\s*
1870 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1871 (?P<quote>["'])?
1872 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1873 (?(quote)(?P=quote))\s*
1874 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1875 m = str_operator_rex.fullmatch(filter_spec)
1876 if m:
1877 if m.group('op') == '~=':
1878 comparison_value = re.compile(m.group('value'))
1879 else:
1880 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1881 str_op = STR_OPERATORS[m.group('op')]
1882 if m.group('negation'):
1883 op = lambda attr, value: not str_op(attr, value)
1884 else:
1885 op = str_op
1886
1887 if not m:
1888 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1889
1890 def _filter(f):
1891 actual_value = f.get(m.group('key'))
1892 if actual_value is None:
1893 return m.group('none_inclusive')
1894 return op(actual_value, comparison_value)
1895 return _filter
1896
1897 def _check_formats(self, formats):
1898 for f in formats:
1899 self.to_screen('[info] Testing format %s' % f['format_id'])
1900 path = self.get_output_path('temp')
1901 if not self._ensure_dir_exists(f'{path}/'):
1902 continue
1903 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1904 temp_file.close()
1905 try:
1906 success, _ = self.dl(temp_file.name, f, test=True)
1907 except (DownloadError, IOError, OSError, ValueError) + network_exceptions:
1908 success = False
1909 finally:
1910 if os.path.exists(temp_file.name):
1911 try:
1912 os.remove(temp_file.name)
1913 except OSError:
1914 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1915 if success:
1916 yield f
1917 else:
1918 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1919
1920 def _default_format_spec(self, info_dict, download=True):
1921
1922 def can_merge():
1923 merger = FFmpegMergerPP(self)
1924 return merger.available and merger.can_merge()
1925
1926 prefer_best = (
1927 not self.params.get('simulate')
1928 and download
1929 and (
1930 not can_merge()
1931 or info_dict.get('is_live', False)
1932 or self.outtmpl_dict['default'] == '-'))
1933 compat = (
1934 prefer_best
1935 or self.params.get('allow_multiple_audio_streams', False)
1936 or 'format-spec' in self.params.get('compat_opts', []))
1937
1938 return (
1939 'best/bestvideo+bestaudio' if prefer_best
1940 else 'bestvideo*+bestaudio/best' if not compat
1941 else 'bestvideo+bestaudio/best')
1942
1943 def build_format_selector(self, format_spec):
1944 def syntax_error(note, start):
1945 message = (
1946 'Invalid format specification: '
1947 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1948 return SyntaxError(message)
1949
1950 PICKFIRST = 'PICKFIRST'
1951 MERGE = 'MERGE'
1952 SINGLE = 'SINGLE'
1953 GROUP = 'GROUP'
1954 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1955
1956 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1957 'video': self.params.get('allow_multiple_video_streams', False)}
1958
1959 check_formats = self.params.get('check_formats') == 'selected'
1960
1961 def _parse_filter(tokens):
1962 filter_parts = []
1963 for type, string, start, _, _ in tokens:
1964 if type == tokenize.OP and string == ']':
1965 return ''.join(filter_parts)
1966 else:
1967 filter_parts.append(string)
1968
1969 def _remove_unused_ops(tokens):
1970 # Remove operators that we don't use and join them with the surrounding strings
1971 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1972 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1973 last_string, last_start, last_end, last_line = None, None, None, None
1974 for type, string, start, end, line in tokens:
1975 if type == tokenize.OP and string == '[':
1976 if last_string:
1977 yield tokenize.NAME, last_string, last_start, last_end, last_line
1978 last_string = None
1979 yield type, string, start, end, line
1980 # everything inside brackets will be handled by _parse_filter
1981 for type, string, start, end, line in tokens:
1982 yield type, string, start, end, line
1983 if type == tokenize.OP and string == ']':
1984 break
1985 elif type == tokenize.OP and string in ALLOWED_OPS:
1986 if last_string:
1987 yield tokenize.NAME, last_string, last_start, last_end, last_line
1988 last_string = None
1989 yield type, string, start, end, line
1990 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1991 if not last_string:
1992 last_string = string
1993 last_start = start
1994 last_end = end
1995 else:
1996 last_string += string
1997 if last_string:
1998 yield tokenize.NAME, last_string, last_start, last_end, last_line
1999
2000 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2001 selectors = []
2002 current_selector = None
2003 for type, string, start, _, _ in tokens:
2004 # ENCODING is only defined in python 3.x
2005 if type == getattr(tokenize, 'ENCODING', None):
2006 continue
2007 elif type in [tokenize.NAME, tokenize.NUMBER]:
2008 current_selector = FormatSelector(SINGLE, string, [])
2009 elif type == tokenize.OP:
2010 if string == ')':
2011 if not inside_group:
2012 # ')' will be handled by the parentheses group
2013 tokens.restore_last_token()
2014 break
2015 elif inside_merge and string in ['/', ',']:
2016 tokens.restore_last_token()
2017 break
2018 elif inside_choice and string == ',':
2019 tokens.restore_last_token()
2020 break
2021 elif string == ',':
2022 if not current_selector:
2023 raise syntax_error('"," must follow a format selector', start)
2024 selectors.append(current_selector)
2025 current_selector = None
2026 elif string == '/':
2027 if not current_selector:
2028 raise syntax_error('"/" must follow a format selector', start)
2029 first_choice = current_selector
2030 second_choice = _parse_format_selection(tokens, inside_choice=True)
2031 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2032 elif string == '[':
2033 if not current_selector:
2034 current_selector = FormatSelector(SINGLE, 'best', [])
2035 format_filter = _parse_filter(tokens)
2036 current_selector.filters.append(format_filter)
2037 elif string == '(':
2038 if current_selector:
2039 raise syntax_error('Unexpected "("', start)
2040 group = _parse_format_selection(tokens, inside_group=True)
2041 current_selector = FormatSelector(GROUP, group, [])
2042 elif string == '+':
2043 if not current_selector:
2044 raise syntax_error('Unexpected "+"', start)
2045 selector_1 = current_selector
2046 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2047 if not selector_2:
2048 raise syntax_error('Expected a selector', start)
2049 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2050 else:
2051 raise syntax_error(f'Operator not recognized: "{string}"', start)
2052 elif type == tokenize.ENDMARKER:
2053 break
2054 if current_selector:
2055 selectors.append(current_selector)
2056 return selectors
2057
2058 def _merge(formats_pair):
2059 format_1, format_2 = formats_pair
2060
2061 formats_info = []
2062 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2063 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2064
2065 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2066 get_no_more = {'video': False, 'audio': False}
2067 for (i, fmt_info) in enumerate(formats_info):
2068 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2069 formats_info.pop(i)
2070 continue
2071 for aud_vid in ['audio', 'video']:
2072 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2073 if get_no_more[aud_vid]:
2074 formats_info.pop(i)
2075 break
2076 get_no_more[aud_vid] = True
2077
2078 if len(formats_info) == 1:
2079 return formats_info[0]
2080
2081 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2082 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2083
2084 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2085 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2086
2087 output_ext = self.params.get('merge_output_format')
2088 if not output_ext:
2089 if the_only_video:
2090 output_ext = the_only_video['ext']
2091 elif the_only_audio and not video_fmts:
2092 output_ext = the_only_audio['ext']
2093 else:
2094 output_ext = 'mkv'
2095
2096 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2097
2098 new_dict = {
2099 'requested_formats': formats_info,
2100 'format': '+'.join(filtered('format')),
2101 'format_id': '+'.join(filtered('format_id')),
2102 'ext': output_ext,
2103 'protocol': '+'.join(map(determine_protocol, formats_info)),
2104 'language': '+'.join(orderedSet(filtered('language'))) or None,
2105 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2106 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2107 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2108 }
2109
2110 if the_only_video:
2111 new_dict.update({
2112 'width': the_only_video.get('width'),
2113 'height': the_only_video.get('height'),
2114 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2115 'fps': the_only_video.get('fps'),
2116 'dynamic_range': the_only_video.get('dynamic_range'),
2117 'vcodec': the_only_video.get('vcodec'),
2118 'vbr': the_only_video.get('vbr'),
2119 'stretched_ratio': the_only_video.get('stretched_ratio'),
2120 })
2121
2122 if the_only_audio:
2123 new_dict.update({
2124 'acodec': the_only_audio.get('acodec'),
2125 'abr': the_only_audio.get('abr'),
2126 'asr': the_only_audio.get('asr'),
2127 })
2128
2129 return new_dict
2130
2131 def _check_formats(formats):
2132 if not check_formats:
2133 yield from formats
2134 return
2135 yield from self._check_formats(formats)
2136
2137 def _build_selector_function(selector):
2138 if isinstance(selector, list): # ,
2139 fs = [_build_selector_function(s) for s in selector]
2140
2141 def selector_function(ctx):
2142 for f in fs:
2143 yield from f(ctx)
2144 return selector_function
2145
2146 elif selector.type == GROUP: # ()
2147 selector_function = _build_selector_function(selector.selector)
2148
2149 elif selector.type == PICKFIRST: # /
2150 fs = [_build_selector_function(s) for s in selector.selector]
2151
2152 def selector_function(ctx):
2153 for f in fs:
2154 picked_formats = list(f(ctx))
2155 if picked_formats:
2156 return picked_formats
2157 return []
2158
2159 elif selector.type == MERGE: # +
2160 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2161
2162 def selector_function(ctx):
2163 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2164 yield _merge(pair)
2165
2166 elif selector.type == SINGLE: # atom
2167 format_spec = selector.selector or 'best'
2168
2169 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2170 if format_spec == 'all':
2171 def selector_function(ctx):
2172 yield from _check_formats(ctx['formats'][::-1])
2173 elif format_spec == 'mergeall':
2174 def selector_function(ctx):
2175 formats = list(_check_formats(
2176 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2177 if not formats:
2178 return
2179 merged_format = formats[-1]
2180 for f in formats[-2::-1]:
2181 merged_format = _merge((merged_format, f))
2182 yield merged_format
2183
2184 else:
2185 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2186 mobj = re.match(
2187 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2188 format_spec)
2189 if mobj is not None:
2190 format_idx = int_or_none(mobj.group('n'), default=1)
2191 format_reverse = mobj.group('bw')[0] == 'b'
2192 format_type = (mobj.group('type') or [None])[0]
2193 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2194 format_modified = mobj.group('mod') is not None
2195
2196 format_fallback = not format_type and not format_modified # for b, w
2197 _filter_f = (
2198 (lambda f: f.get('%scodec' % format_type) != 'none')
2199 if format_type and format_modified # bv*, ba*, wv*, wa*
2200 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2201 if format_type # bv, ba, wv, wa
2202 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2203 if not format_modified # b, w
2204 else lambda f: True) # b*, w*
2205 filter_f = lambda f: _filter_f(f) and (
2206 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2207 else:
2208 if format_spec in self._format_selection_exts['audio']:
2209 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2210 elif format_spec in self._format_selection_exts['video']:
2211 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2212 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2213 elif format_spec in self._format_selection_exts['storyboards']:
2214 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2215 else:
2216 filter_f = lambda f: f.get('format_id') == format_spec # id
2217
2218 def selector_function(ctx):
2219 formats = list(ctx['formats'])
2220 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2221 if not matches:
2222 if format_fallback and ctx['incomplete_formats']:
2223 # for extractors with incomplete formats (audio only (soundcloud)
2224 # or video only (imgur)) best/worst will fallback to
2225 # best/worst {video,audio}-only format
2226 matches = formats
2227 elif seperate_fallback and not ctx['has_merged_format']:
2228 # for compatibility with youtube-dl when there is no pre-merged format
2229 matches = list(filter(seperate_fallback, formats))
2230 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2231 try:
2232 yield matches[format_idx - 1]
2233 except LazyList.IndexError:
2234 return
2235
2236 filters = [self._build_format_filter(f) for f in selector.filters]
2237
2238 def final_selector(ctx):
2239 ctx_copy = dict(ctx)
2240 for _filter in filters:
2241 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2242 return selector_function(ctx_copy)
2243 return final_selector
2244
2245 stream = io.BytesIO(format_spec.encode())
2246 try:
2247 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2248 except tokenize.TokenError:
2249 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2250
2251 class TokenIterator:
2252 def __init__(self, tokens):
2253 self.tokens = tokens
2254 self.counter = 0
2255
2256 def __iter__(self):
2257 return self
2258
2259 def __next__(self):
2260 if self.counter >= len(self.tokens):
2261 raise StopIteration()
2262 value = self.tokens[self.counter]
2263 self.counter += 1
2264 return value
2265
2266 next = __next__
2267
2268 def restore_last_token(self):
2269 self.counter -= 1
2270
2271 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2272 return _build_selector_function(parsed_selector)
2273
2274 def _calc_headers(self, info_dict):
2275 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2276
2277 cookies = self._calc_cookies(info_dict)
2278 if cookies:
2279 res['Cookie'] = cookies
2280
2281 if 'X-Forwarded-For' not in res:
2282 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2283 if x_forwarded_for_ip:
2284 res['X-Forwarded-For'] = x_forwarded_for_ip
2285
2286 return res
2287
2288 def _calc_cookies(self, info_dict):
2289 pr = sanitized_Request(info_dict['url'])
2290 self.cookiejar.add_cookie_header(pr)
2291 return pr.get_header('Cookie')
2292
2293 def _sort_thumbnails(self, thumbnails):
2294 thumbnails.sort(key=lambda t: (
2295 t.get('preference') if t.get('preference') is not None else -1,
2296 t.get('width') if t.get('width') is not None else -1,
2297 t.get('height') if t.get('height') is not None else -1,
2298 t.get('id') if t.get('id') is not None else '',
2299 t.get('url')))
2300
2301 def _sanitize_thumbnails(self, info_dict):
2302 thumbnails = info_dict.get('thumbnails')
2303 if thumbnails is None:
2304 thumbnail = info_dict.get('thumbnail')
2305 if thumbnail:
2306 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2307 if not thumbnails:
2308 return
2309
2310 def check_thumbnails(thumbnails):
2311 for t in thumbnails:
2312 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2313 try:
2314 self.urlopen(HEADRequest(t['url']))
2315 except network_exceptions as err:
2316 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2317 continue
2318 yield t
2319
2320 self._sort_thumbnails(thumbnails)
2321 for i, t in enumerate(thumbnails):
2322 if t.get('id') is None:
2323 t['id'] = '%d' % i
2324 if t.get('width') and t.get('height'):
2325 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2326 t['url'] = sanitize_url(t['url'])
2327
2328 if self.params.get('check_formats') is True:
2329 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2330 else:
2331 info_dict['thumbnails'] = thumbnails
2332
2333 def _fill_common_fields(self, info_dict, is_video=True):
2334 # TODO: move sanitization here
2335 if is_video:
2336 # playlists are allowed to lack "title"
2337 title = info_dict.get('title', NO_DEFAULT)
2338 if title is NO_DEFAULT:
2339 raise ExtractorError('Missing "title" field in extractor result',
2340 video_id=info_dict['id'], ie=info_dict['extractor'])
2341 info_dict['fulltitle'] = title
2342 if not title:
2343 if title == '':
2344 self.write_debug('Extractor gave empty title. Creating a generic title')
2345 else:
2346 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2347 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2348
2349 if info_dict.get('duration') is not None:
2350 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2351
2352 for ts_key, date_key in (
2353 ('timestamp', 'upload_date'),
2354 ('release_timestamp', 'release_date'),
2355 ('modified_timestamp', 'modified_date'),
2356 ):
2357 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2358 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2359 # see http://bugs.python.org/issue1646728)
2360 with contextlib.suppress(ValueError, OverflowError, OSError):
2361 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2362 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2363
2364 live_keys = ('is_live', 'was_live')
2365 live_status = info_dict.get('live_status')
2366 if live_status is None:
2367 for key in live_keys:
2368 if info_dict.get(key) is False:
2369 continue
2370 if info_dict.get(key):
2371 live_status = key
2372 break
2373 if all(info_dict.get(key) is False for key in live_keys):
2374 live_status = 'not_live'
2375 if live_status:
2376 info_dict['live_status'] = live_status
2377 for key in live_keys:
2378 if info_dict.get(key) is None:
2379 info_dict[key] = (live_status == key)
2380
2381 # Auto generate title fields corresponding to the *_number fields when missing
2382 # in order to always have clean titles. This is very common for TV series.
2383 for field in ('chapter', 'season', 'episode'):
2384 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2385 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2386
2387 def process_video_result(self, info_dict, download=True):
2388 assert info_dict.get('_type', 'video') == 'video'
2389 self._num_videos += 1
2390
2391 if 'id' not in info_dict:
2392 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2393 elif not info_dict.get('id'):
2394 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2395
2396 def report_force_conversion(field, field_not, conversion):
2397 self.report_warning(
2398 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2399 % (field, field_not, conversion))
2400
2401 def sanitize_string_field(info, string_field):
2402 field = info.get(string_field)
2403 if field is None or isinstance(field, compat_str):
2404 return
2405 report_force_conversion(string_field, 'a string', 'string')
2406 info[string_field] = compat_str(field)
2407
2408 def sanitize_numeric_fields(info):
2409 for numeric_field in self._NUMERIC_FIELDS:
2410 field = info.get(numeric_field)
2411 if field is None or isinstance(field, (int, float)):
2412 continue
2413 report_force_conversion(numeric_field, 'numeric', 'int')
2414 info[numeric_field] = int_or_none(field)
2415
2416 sanitize_string_field(info_dict, 'id')
2417 sanitize_numeric_fields(info_dict)
2418 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2419 self.report_warning('"duration" field is negative, there is an error in extractor')
2420
2421 if 'playlist' not in info_dict:
2422 # It isn't part of a playlist
2423 info_dict['playlist'] = None
2424 info_dict['playlist_index'] = None
2425
2426 self._sanitize_thumbnails(info_dict)
2427
2428 thumbnail = info_dict.get('thumbnail')
2429 thumbnails = info_dict.get('thumbnails')
2430 if thumbnail:
2431 info_dict['thumbnail'] = sanitize_url(thumbnail)
2432 elif thumbnails:
2433 info_dict['thumbnail'] = thumbnails[-1]['url']
2434
2435 if info_dict.get('display_id') is None and 'id' in info_dict:
2436 info_dict['display_id'] = info_dict['id']
2437
2438 self._fill_common_fields(info_dict)
2439
2440 for cc_kind in ('subtitles', 'automatic_captions'):
2441 cc = info_dict.get(cc_kind)
2442 if cc:
2443 for _, subtitle in cc.items():
2444 for subtitle_format in subtitle:
2445 if subtitle_format.get('url'):
2446 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2447 if subtitle_format.get('ext') is None:
2448 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2449
2450 automatic_captions = info_dict.get('automatic_captions')
2451 subtitles = info_dict.get('subtitles')
2452
2453 info_dict['requested_subtitles'] = self.process_subtitles(
2454 info_dict['id'], subtitles, automatic_captions)
2455
2456 if info_dict.get('formats') is None:
2457 # There's only one format available
2458 formats = [info_dict]
2459 else:
2460 formats = info_dict['formats']
2461
2462 # or None ensures --clean-infojson removes it
2463 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2464 if not self.params.get('allow_unplayable_formats'):
2465 formats = [f for f in formats if not f.get('has_drm')]
2466 if info_dict['_has_drm'] and all(
2467 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2468 self.report_warning(
2469 'This video is DRM protected and only images are available for download. '
2470 'Use --list-formats to see them')
2471
2472 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2473 if not get_from_start:
2474 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2475 if info_dict.get('is_live') and formats:
2476 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2477 if get_from_start and not formats:
2478 self.raise_no_formats(info_dict, msg=(
2479 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2480 'If you want to download from the current time, use --no-live-from-start'))
2481
2482 if not formats:
2483 self.raise_no_formats(info_dict)
2484
2485 def is_wellformed(f):
2486 url = f.get('url')
2487 if not url:
2488 self.report_warning(
2489 '"url" field is missing or empty - skipping format, '
2490 'there is an error in extractor')
2491 return False
2492 if isinstance(url, bytes):
2493 sanitize_string_field(f, 'url')
2494 return True
2495
2496 # Filter out malformed formats for better extraction robustness
2497 formats = list(filter(is_wellformed, formats))
2498
2499 formats_dict = {}
2500
2501 # We check that all the formats have the format and format_id fields
2502 for i, format in enumerate(formats):
2503 sanitize_string_field(format, 'format_id')
2504 sanitize_numeric_fields(format)
2505 format['url'] = sanitize_url(format['url'])
2506 if not format.get('format_id'):
2507 format['format_id'] = compat_str(i)
2508 else:
2509 # Sanitize format_id from characters used in format selector expression
2510 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2511 format_id = format['format_id']
2512 if format_id not in formats_dict:
2513 formats_dict[format_id] = []
2514 formats_dict[format_id].append(format)
2515
2516 # Make sure all formats have unique format_id
2517 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2518 for format_id, ambiguous_formats in formats_dict.items():
2519 ambigious_id = len(ambiguous_formats) > 1
2520 for i, format in enumerate(ambiguous_formats):
2521 if ambigious_id:
2522 format['format_id'] = '%s-%d' % (format_id, i)
2523 if format.get('ext') is None:
2524 format['ext'] = determine_ext(format['url']).lower()
2525 # Ensure there is no conflict between id and ext in format selection
2526 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2527 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2528 format['format_id'] = 'f%s' % format['format_id']
2529
2530 for i, format in enumerate(formats):
2531 if format.get('format') is None:
2532 format['format'] = '{id} - {res}{note}'.format(
2533 id=format['format_id'],
2534 res=self.format_resolution(format),
2535 note=format_field(format, 'format_note', ' (%s)'),
2536 )
2537 if format.get('protocol') is None:
2538 format['protocol'] = determine_protocol(format)
2539 if format.get('resolution') is None:
2540 format['resolution'] = self.format_resolution(format, default=None)
2541 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2542 format['dynamic_range'] = 'SDR'
2543 if (info_dict.get('duration') and format.get('tbr')
2544 and not format.get('filesize') and not format.get('filesize_approx')):
2545 format['filesize_approx'] = info_dict['duration'] * format['tbr'] * (1024 / 8)
2546
2547 # Add HTTP headers, so that external programs can use them from the
2548 # json output
2549 full_format_info = info_dict.copy()
2550 full_format_info.update(format)
2551 format['http_headers'] = self._calc_headers(full_format_info)
2552 # Remove private housekeeping stuff
2553 if '__x_forwarded_for_ip' in info_dict:
2554 del info_dict['__x_forwarded_for_ip']
2555
2556 if self.params.get('check_formats') is True:
2557 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2558
2559 if not formats or formats[0] is not info_dict:
2560 # only set the 'formats' fields if the original info_dict list them
2561 # otherwise we end up with a circular reference, the first (and unique)
2562 # element in the 'formats' field in info_dict is info_dict itself,
2563 # which can't be exported to json
2564 info_dict['formats'] = formats
2565
2566 info_dict, _ = self.pre_process(info_dict)
2567
2568 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2569 return info_dict
2570
2571 self.post_extract(info_dict)
2572 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2573
2574 # The pre-processors may have modified the formats
2575 formats = info_dict.get('formats', [info_dict])
2576
2577 list_only = self.params.get('simulate') is None and (
2578 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2579 interactive_format_selection = not list_only and self.format_selector == '-'
2580 if self.params.get('list_thumbnails'):
2581 self.list_thumbnails(info_dict)
2582 if self.params.get('listsubtitles'):
2583 if 'automatic_captions' in info_dict:
2584 self.list_subtitles(
2585 info_dict['id'], automatic_captions, 'automatic captions')
2586 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2587 if self.params.get('listformats') or interactive_format_selection:
2588 self.list_formats(info_dict)
2589 if list_only:
2590 # Without this printing, -F --print-json will not work
2591 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2592 return
2593
2594 format_selector = self.format_selector
2595 if format_selector is None:
2596 req_format = self._default_format_spec(info_dict, download=download)
2597 self.write_debug('Default format spec: %s' % req_format)
2598 format_selector = self.build_format_selector(req_format)
2599
2600 while True:
2601 if interactive_format_selection:
2602 req_format = input(
2603 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2604 try:
2605 format_selector = self.build_format_selector(req_format)
2606 except SyntaxError as err:
2607 self.report_error(err, tb=False, is_error=False)
2608 continue
2609
2610 formats_to_download = list(format_selector({
2611 'formats': formats,
2612 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2613 'incomplete_formats': (
2614 # All formats are video-only or
2615 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2616 # all formats are audio-only
2617 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2618 }))
2619 if interactive_format_selection and not formats_to_download:
2620 self.report_error('Requested format is not available', tb=False, is_error=False)
2621 continue
2622 break
2623
2624 if not formats_to_download:
2625 if not self.params.get('ignore_no_formats_error'):
2626 raise ExtractorError(
2627 'Requested format is not available. Use --list-formats for a list of available formats',
2628 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2629 self.report_warning('Requested format is not available')
2630 # Process what we can, even without any available formats.
2631 formats_to_download = [{}]
2632
2633 best_format = formats_to_download[-1]
2634 if download:
2635 if best_format:
2636 self.to_screen(
2637 f'[info] {info_dict["id"]}: Downloading {len(formats_to_download)} format(s): '
2638 + ', '.join([f['format_id'] for f in formats_to_download]))
2639 max_downloads_reached = False
2640 for i, fmt in enumerate(formats_to_download):
2641 formats_to_download[i] = new_info = self._copy_infodict(info_dict)
2642 new_info.update(fmt)
2643 try:
2644 self.process_info(new_info)
2645 except MaxDownloadsReached:
2646 max_downloads_reached = True
2647 # Remove copied info
2648 for key, val in tuple(new_info.items()):
2649 if info_dict.get(key) == val:
2650 new_info.pop(key)
2651 if max_downloads_reached:
2652 break
2653
2654 write_archive = {f.get('__write_download_archive', False) for f in formats_to_download}
2655 assert write_archive.issubset({True, False, 'ignore'})
2656 if True in write_archive and False not in write_archive:
2657 self.record_download_archive(info_dict)
2658
2659 info_dict['requested_downloads'] = formats_to_download
2660 info_dict = self.run_all_pps('after_video', info_dict)
2661 if max_downloads_reached:
2662 raise MaxDownloadsReached()
2663
2664 # We update the info dict with the selected best quality format (backwards compatibility)
2665 info_dict.update(best_format)
2666 return info_dict
2667
2668 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2669 """Select the requested subtitles and their format"""
2670 available_subs, normal_sub_langs = {}, []
2671 if normal_subtitles and self.params.get('writesubtitles'):
2672 available_subs.update(normal_subtitles)
2673 normal_sub_langs = tuple(normal_subtitles.keys())
2674 if automatic_captions and self.params.get('writeautomaticsub'):
2675 for lang, cap_info in automatic_captions.items():
2676 if lang not in available_subs:
2677 available_subs[lang] = cap_info
2678
2679 if (not self.params.get('writesubtitles') and not
2680 self.params.get('writeautomaticsub') or not
2681 available_subs):
2682 return None
2683
2684 all_sub_langs = tuple(available_subs.keys())
2685 if self.params.get('allsubtitles', False):
2686 requested_langs = all_sub_langs
2687 elif self.params.get('subtitleslangs', False):
2688 # A list is used so that the order of languages will be the same as
2689 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2690 requested_langs = []
2691 for lang_re in self.params.get('subtitleslangs'):
2692 discard = lang_re[0] == '-'
2693 if discard:
2694 lang_re = lang_re[1:]
2695 if lang_re == 'all':
2696 if discard:
2697 requested_langs = []
2698 else:
2699 requested_langs.extend(all_sub_langs)
2700 continue
2701 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2702 if discard:
2703 for lang in current_langs:
2704 while lang in requested_langs:
2705 requested_langs.remove(lang)
2706 else:
2707 requested_langs.extend(current_langs)
2708 requested_langs = orderedSet(requested_langs)
2709 elif normal_sub_langs:
2710 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2711 else:
2712 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2713 if requested_langs:
2714 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2715
2716 formats_query = self.params.get('subtitlesformat', 'best')
2717 formats_preference = formats_query.split('/') if formats_query else []
2718 subs = {}
2719 for lang in requested_langs:
2720 formats = available_subs.get(lang)
2721 if formats is None:
2722 self.report_warning(f'{lang} subtitles not available for {video_id}')
2723 continue
2724 for ext in formats_preference:
2725 if ext == 'best':
2726 f = formats[-1]
2727 break
2728 matches = list(filter(lambda f: f['ext'] == ext, formats))
2729 if matches:
2730 f = matches[-1]
2731 break
2732 else:
2733 f = formats[-1]
2734 self.report_warning(
2735 'No subtitle format found matching "%s" for language %s, '
2736 'using %s' % (formats_query, lang, f['ext']))
2737 subs[lang] = f
2738 return subs
2739
2740 def _forceprint(self, key, info_dict):
2741 if info_dict is None:
2742 return
2743 info_copy = info_dict.copy()
2744 info_copy['formats_table'] = self.render_formats_table(info_dict)
2745 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2746 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2747 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2748
2749 def format_tmpl(tmpl):
2750 mobj = re.match(r'\w+(=?)$', tmpl)
2751 if mobj and mobj.group(1):
2752 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2753 elif mobj:
2754 return f'%({tmpl})s'
2755 return tmpl
2756
2757 for tmpl in self.params['forceprint'].get(key, []):
2758 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2759
2760 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2761 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2762 tmpl = format_tmpl(tmpl)
2763 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2764 if self._ensure_dir_exists(filename):
2765 with open(filename, 'a', encoding='utf-8') as f:
2766 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2767
2768 def __forced_printings(self, info_dict, filename, incomplete):
2769 def print_mandatory(field, actual_field=None):
2770 if actual_field is None:
2771 actual_field = field
2772 if (self.params.get('force%s' % field, False)
2773 and (not incomplete or info_dict.get(actual_field) is not None)):
2774 self.to_stdout(info_dict[actual_field])
2775
2776 def print_optional(field):
2777 if (self.params.get('force%s' % field, False)
2778 and info_dict.get(field) is not None):
2779 self.to_stdout(info_dict[field])
2780
2781 info_dict = info_dict.copy()
2782 if filename is not None:
2783 info_dict['filename'] = filename
2784 if info_dict.get('requested_formats') is not None:
2785 # For RTMP URLs, also include the playpath
2786 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2787 elif info_dict.get('url'):
2788 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2789
2790 if (self.params.get('forcejson')
2791 or self.params['forceprint'].get('video')
2792 or self.params['print_to_file'].get('video')):
2793 self.post_extract(info_dict)
2794 self._forceprint('video', info_dict)
2795
2796 print_mandatory('title')
2797 print_mandatory('id')
2798 print_mandatory('url', 'urls')
2799 print_optional('thumbnail')
2800 print_optional('description')
2801 print_optional('filename')
2802 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2803 self.to_stdout(formatSeconds(info_dict['duration']))
2804 print_mandatory('format')
2805
2806 if self.params.get('forcejson'):
2807 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2808
2809 def dl(self, name, info, subtitle=False, test=False):
2810 if not info.get('url'):
2811 self.raise_no_formats(info, True)
2812
2813 if test:
2814 verbose = self.params.get('verbose')
2815 params = {
2816 'test': True,
2817 'quiet': self.params.get('quiet') or not verbose,
2818 'verbose': verbose,
2819 'noprogress': not verbose,
2820 'nopart': True,
2821 'skip_unavailable_fragments': False,
2822 'keep_fragments': False,
2823 'overwrites': True,
2824 '_no_ytdl_file': True,
2825 }
2826 else:
2827 params = self.params
2828 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2829 if not test:
2830 for ph in self._progress_hooks:
2831 fd.add_progress_hook(ph)
2832 urls = '", "'.join(
2833 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2834 for f in info.get('requested_formats', []) or [info])
2835 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2836
2837 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2838 # But it may contain objects that are not deep-copyable
2839 new_info = self._copy_infodict(info)
2840 if new_info.get('http_headers') is None:
2841 new_info['http_headers'] = self._calc_headers(new_info)
2842 return fd.download(name, new_info, subtitle)
2843
2844 def existing_file(self, filepaths, *, default_overwrite=True):
2845 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2846 if existing_files and not self.params.get('overwrites', default_overwrite):
2847 return existing_files[0]
2848
2849 for file in existing_files:
2850 self.report_file_delete(file)
2851 os.remove(file)
2852 return None
2853
2854 def process_info(self, info_dict):
2855 """Process a single resolved IE result. (Modifies it in-place)"""
2856
2857 assert info_dict.get('_type', 'video') == 'video'
2858 original_infodict = info_dict
2859
2860 if 'format' not in info_dict and 'ext' in info_dict:
2861 info_dict['format'] = info_dict['ext']
2862
2863 # This is mostly just for backward compatibility of process_info
2864 # As a side-effect, this allows for format-specific filters
2865 if self._match_entry(info_dict) is not None:
2866 info_dict['__write_download_archive'] = 'ignore'
2867 return
2868
2869 # Does nothing under normal operation - for backward compatibility of process_info
2870 self.post_extract(info_dict)
2871 self._num_downloads += 1
2872
2873 # info_dict['_filename'] needs to be set for backward compatibility
2874 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2875 temp_filename = self.prepare_filename(info_dict, 'temp')
2876 files_to_move = {}
2877
2878 # Forced printings
2879 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2880
2881 if self.params.get('simulate'):
2882 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2883 return
2884
2885 if full_filename is None:
2886 return
2887 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2888 return
2889 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2890 return
2891
2892 if self._write_description('video', info_dict,
2893 self.prepare_filename(info_dict, 'description')) is None:
2894 return
2895
2896 sub_files = self._write_subtitles(info_dict, temp_filename)
2897 if sub_files is None:
2898 return
2899 files_to_move.update(dict(sub_files))
2900
2901 thumb_files = self._write_thumbnails(
2902 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2903 if thumb_files is None:
2904 return
2905 files_to_move.update(dict(thumb_files))
2906
2907 infofn = self.prepare_filename(info_dict, 'infojson')
2908 _infojson_written = self._write_info_json('video', info_dict, infofn)
2909 if _infojson_written:
2910 info_dict['infojson_filename'] = infofn
2911 # For backward compatibility, even though it was a private field
2912 info_dict['__infojson_filename'] = infofn
2913 elif _infojson_written is None:
2914 return
2915
2916 # Note: Annotations are deprecated
2917 annofn = None
2918 if self.params.get('writeannotations', False):
2919 annofn = self.prepare_filename(info_dict, 'annotation')
2920 if annofn:
2921 if not self._ensure_dir_exists(encodeFilename(annofn)):
2922 return
2923 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2924 self.to_screen('[info] Video annotations are already present')
2925 elif not info_dict.get('annotations'):
2926 self.report_warning('There are no annotations to write.')
2927 else:
2928 try:
2929 self.to_screen('[info] Writing video annotations to: ' + annofn)
2930 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2931 annofile.write(info_dict['annotations'])
2932 except (KeyError, TypeError):
2933 self.report_warning('There are no annotations to write.')
2934 except OSError:
2935 self.report_error('Cannot write annotations file: ' + annofn)
2936 return
2937
2938 # Write internet shortcut files
2939 def _write_link_file(link_type):
2940 url = try_get(info_dict['webpage_url'], iri_to_uri)
2941 if not url:
2942 self.report_warning(
2943 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2944 return True
2945 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2946 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2947 return False
2948 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2949 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2950 return True
2951 try:
2952 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2953 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2954 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2955 template_vars = {'url': url}
2956 if link_type == 'desktop':
2957 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2958 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2959 except OSError:
2960 self.report_error(f'Cannot write internet shortcut {linkfn}')
2961 return False
2962 return True
2963
2964 write_links = {
2965 'url': self.params.get('writeurllink'),
2966 'webloc': self.params.get('writewebloclink'),
2967 'desktop': self.params.get('writedesktoplink'),
2968 }
2969 if self.params.get('writelink'):
2970 link_type = ('webloc' if sys.platform == 'darwin'
2971 else 'desktop' if sys.platform.startswith('linux')
2972 else 'url')
2973 write_links[link_type] = True
2974
2975 if any(should_write and not _write_link_file(link_type)
2976 for link_type, should_write in write_links.items()):
2977 return
2978
2979 def replace_info_dict(new_info):
2980 nonlocal info_dict
2981 if new_info == info_dict:
2982 return
2983 info_dict.clear()
2984 info_dict.update(new_info)
2985
2986 try:
2987 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2988 replace_info_dict(new_info)
2989 except PostProcessingError as err:
2990 self.report_error('Preprocessing: %s' % str(err))
2991 return
2992
2993 if self.params.get('skip_download'):
2994 info_dict['filepath'] = temp_filename
2995 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2996 info_dict['__files_to_move'] = files_to_move
2997 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2998 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2999 else:
3000 # Download
3001 info_dict.setdefault('__postprocessors', [])
3002 try:
3003
3004 def existing_video_file(*filepaths):
3005 ext = info_dict.get('ext')
3006 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3007 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3008 default_overwrite=False)
3009 if file:
3010 info_dict['ext'] = os.path.splitext(file)[1][1:]
3011 return file
3012
3013 success = True
3014 if info_dict.get('requested_formats') is not None:
3015
3016 def compatible_formats(formats):
3017 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3018 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3019 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3020 if len(video_formats) > 2 or len(audio_formats) > 2:
3021 return False
3022
3023 # Check extension
3024 exts = {format.get('ext') for format in formats}
3025 COMPATIBLE_EXTS = (
3026 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3027 {'webm'},
3028 )
3029 for ext_sets in COMPATIBLE_EXTS:
3030 if ext_sets.issuperset(exts):
3031 return True
3032 # TODO: Check acodec/vcodec
3033 return False
3034
3035 requested_formats = info_dict['requested_formats']
3036 old_ext = info_dict['ext']
3037 if self.params.get('merge_output_format') is None:
3038 if not compatible_formats(requested_formats):
3039 info_dict['ext'] = 'mkv'
3040 self.report_warning(
3041 'Requested formats are incompatible for merge and will be merged into mkv')
3042 if (info_dict['ext'] == 'webm'
3043 and info_dict.get('thumbnails')
3044 # check with type instead of pp_key, __name__, or isinstance
3045 # since we dont want any custom PPs to trigger this
3046 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])):
3047 info_dict['ext'] = 'mkv'
3048 self.report_warning(
3049 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3050 new_ext = info_dict['ext']
3051
3052 def correct_ext(filename, ext=new_ext):
3053 if filename == '-':
3054 return filename
3055 filename_real_ext = os.path.splitext(filename)[1][1:]
3056 filename_wo_ext = (
3057 os.path.splitext(filename)[0]
3058 if filename_real_ext in (old_ext, new_ext)
3059 else filename)
3060 return f'{filename_wo_ext}.{ext}'
3061
3062 # Ensure filename always has a correct extension for successful merge
3063 full_filename = correct_ext(full_filename)
3064 temp_filename = correct_ext(temp_filename)
3065 dl_filename = existing_video_file(full_filename, temp_filename)
3066 info_dict['__real_download'] = False
3067
3068 downloaded = []
3069 merger = FFmpegMergerPP(self)
3070
3071 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3072 if dl_filename is not None:
3073 self.report_file_already_downloaded(dl_filename)
3074 elif fd:
3075 for f in requested_formats if fd != FFmpegFD else []:
3076 f['filepath'] = fname = prepend_extension(
3077 correct_ext(temp_filename, info_dict['ext']),
3078 'f%s' % f['format_id'], info_dict['ext'])
3079 downloaded.append(fname)
3080 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3081 success, real_download = self.dl(temp_filename, info_dict)
3082 info_dict['__real_download'] = real_download
3083 else:
3084 if self.params.get('allow_unplayable_formats'):
3085 self.report_warning(
3086 'You have requested merging of multiple formats '
3087 'while also allowing unplayable formats to be downloaded. '
3088 'The formats won\'t be merged to prevent data corruption.')
3089 elif not merger.available:
3090 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3091 if not self.params.get('ignoreerrors'):
3092 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3093 return
3094 self.report_warning(f'{msg}. The formats won\'t be merged')
3095
3096 if temp_filename == '-':
3097 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3098 else 'but the formats are incompatible for simultaneous download' if merger.available
3099 else 'but ffmpeg is not installed')
3100 self.report_warning(
3101 f'You have requested downloading multiple formats to stdout {reason}. '
3102 'The formats will be streamed one after the other')
3103 fname = temp_filename
3104 for f in requested_formats:
3105 new_info = dict(info_dict)
3106 del new_info['requested_formats']
3107 new_info.update(f)
3108 if temp_filename != '-':
3109 fname = prepend_extension(
3110 correct_ext(temp_filename, new_info['ext']),
3111 'f%s' % f['format_id'], new_info['ext'])
3112 if not self._ensure_dir_exists(fname):
3113 return
3114 f['filepath'] = fname
3115 downloaded.append(fname)
3116 partial_success, real_download = self.dl(fname, new_info)
3117 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3118 success = success and partial_success
3119
3120 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3121 info_dict['__postprocessors'].append(merger)
3122 info_dict['__files_to_merge'] = downloaded
3123 # Even if there were no downloads, it is being merged only now
3124 info_dict['__real_download'] = True
3125 else:
3126 for file in downloaded:
3127 files_to_move[file] = None
3128 else:
3129 # Just a single file
3130 dl_filename = existing_video_file(full_filename, temp_filename)
3131 if dl_filename is None or dl_filename == temp_filename:
3132 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3133 # So we should try to resume the download
3134 success, real_download = self.dl(temp_filename, info_dict)
3135 info_dict['__real_download'] = real_download
3136 else:
3137 self.report_file_already_downloaded(dl_filename)
3138
3139 dl_filename = dl_filename or temp_filename
3140 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3141
3142 except network_exceptions as err:
3143 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3144 return
3145 except OSError as err:
3146 raise UnavailableVideoError(err)
3147 except (ContentTooShortError, ) as err:
3148 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3149 return
3150
3151 if success and full_filename != '-':
3152
3153 def fixup():
3154 do_fixup = True
3155 fixup_policy = self.params.get('fixup')
3156 vid = info_dict['id']
3157
3158 if fixup_policy in ('ignore', 'never'):
3159 return
3160 elif fixup_policy == 'warn':
3161 do_fixup = 'warn'
3162 elif fixup_policy != 'force':
3163 assert fixup_policy in ('detect_or_warn', None)
3164 if not info_dict.get('__real_download'):
3165 do_fixup = False
3166
3167 def ffmpeg_fixup(cndn, msg, cls):
3168 if not (do_fixup and cndn):
3169 return
3170 elif do_fixup == 'warn':
3171 self.report_warning(f'{vid}: {msg}')
3172 return
3173 pp = cls(self)
3174 if pp.available:
3175 info_dict['__postprocessors'].append(pp)
3176 else:
3177 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3178
3179 stretched_ratio = info_dict.get('stretched_ratio')
3180 ffmpeg_fixup(
3181 stretched_ratio not in (1, None),
3182 f'Non-uniform pixel ratio {stretched_ratio}',
3183 FFmpegFixupStretchedPP)
3184
3185 ffmpeg_fixup(
3186 (info_dict.get('requested_formats') is None
3187 and info_dict.get('container') == 'm4a_dash'
3188 and info_dict.get('ext') == 'm4a'),
3189 'writing DASH m4a. Only some players support this container',
3190 FFmpegFixupM4aPP)
3191
3192 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3193 downloader = downloader.FD_NAME if downloader else None
3194
3195 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3196 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3197 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3198 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3199 FFmpegFixupM3u8PP)
3200 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3201 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3202
3203 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3204 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3205
3206 fixup()
3207 try:
3208 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3209 except PostProcessingError as err:
3210 self.report_error('Postprocessing: %s' % str(err))
3211 return
3212 try:
3213 for ph in self._post_hooks:
3214 ph(info_dict['filepath'])
3215 except Exception as err:
3216 self.report_error('post hooks: %s' % str(err))
3217 return
3218 info_dict['__write_download_archive'] = True
3219
3220 if self.params.get('force_write_download_archive'):
3221 info_dict['__write_download_archive'] = True
3222
3223 # Make sure the info_dict was modified in-place
3224 assert info_dict is original_infodict
3225
3226 max_downloads = self.params.get('max_downloads')
3227 if max_downloads is not None and self._num_downloads >= int(max_downloads):
3228 raise MaxDownloadsReached()
3229
3230 def __download_wrapper(self, func):
3231 @functools.wraps(func)
3232 def wrapper(*args, **kwargs):
3233 try:
3234 res = func(*args, **kwargs)
3235 except UnavailableVideoError as e:
3236 self.report_error(e)
3237 except MaxDownloadsReached as e:
3238 self.to_screen(f'[info] {e}')
3239 raise
3240 except DownloadCancelled as e:
3241 self.to_screen(f'[info] {e}')
3242 if not self.params.get('break_per_url'):
3243 raise
3244 else:
3245 if self.params.get('dump_single_json', False):
3246 self.post_extract(res)
3247 self.to_stdout(json.dumps(self.sanitize_info(res)))
3248 return wrapper
3249
3250 def download(self, url_list):
3251 """Download a given list of URLs."""
3252 url_list = variadic(url_list) # Passing a single URL is a common mistake
3253 outtmpl = self.outtmpl_dict['default']
3254 if (len(url_list) > 1
3255 and outtmpl != '-'
3256 and '%' not in outtmpl
3257 and self.params.get('max_downloads') != 1):
3258 raise SameFileError(outtmpl)
3259
3260 for url in url_list:
3261 self.__download_wrapper(self.extract_info)(
3262 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3263
3264 return self._download_retcode
3265
3266 def download_with_info_file(self, info_filename):
3267 with contextlib.closing(fileinput.FileInput(
3268 [info_filename], mode='r',
3269 openhook=fileinput.hook_encoded('utf-8'))) as f:
3270 # FileInput doesn't have a read method, we can't call json.load
3271 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3272 try:
3273 self.__download_wrapper(self.process_ie_result)(info, download=True)
3274 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3275 if not isinstance(e, EntryNotInPlaylist):
3276 self.to_stderr('\r')
3277 webpage_url = info.get('webpage_url')
3278 if webpage_url is not None:
3279 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3280 return self.download([webpage_url])
3281 else:
3282 raise
3283 return self._download_retcode
3284
3285 @staticmethod
3286 def sanitize_info(info_dict, remove_private_keys=False):
3287 ''' Sanitize the infodict for converting to json '''
3288 if info_dict is None:
3289 return info_dict
3290 info_dict.setdefault('epoch', int(time.time()))
3291 info_dict.setdefault('_type', 'video')
3292
3293 if remove_private_keys:
3294 reject = lambda k, v: v is None or k.startswith('__') or k in {
3295 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3296 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3297 }
3298 else:
3299 reject = lambda k, v: False
3300
3301 def filter_fn(obj):
3302 if isinstance(obj, dict):
3303 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3304 elif isinstance(obj, (list, tuple, set, LazyList)):
3305 return list(map(filter_fn, obj))
3306 elif obj is None or isinstance(obj, (str, int, float, bool)):
3307 return obj
3308 else:
3309 return repr(obj)
3310
3311 return filter_fn(info_dict)
3312
3313 @staticmethod
3314 def filter_requested_info(info_dict, actually_filter=True):
3315 ''' Alias of sanitize_info for backward compatibility '''
3316 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3317
3318 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3319 for filename in set(filter(None, files_to_delete)):
3320 if msg:
3321 self.to_screen(msg % filename)
3322 try:
3323 os.remove(filename)
3324 except OSError:
3325 self.report_warning(f'Unable to delete file {filename}')
3326 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3327 del info['__files_to_move'][filename]
3328
3329 @staticmethod
3330 def post_extract(info_dict):
3331 def actual_post_extract(info_dict):
3332 if info_dict.get('_type') in ('playlist', 'multi_video'):
3333 for video_dict in info_dict.get('entries', {}):
3334 actual_post_extract(video_dict or {})
3335 return
3336
3337 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3338 info_dict.update(post_extractor())
3339
3340 actual_post_extract(info_dict or {})
3341
3342 def run_pp(self, pp, infodict):
3343 files_to_delete = []
3344 if '__files_to_move' not in infodict:
3345 infodict['__files_to_move'] = {}
3346 try:
3347 files_to_delete, infodict = pp.run(infodict)
3348 except PostProcessingError as e:
3349 # Must be True and not 'only_download'
3350 if self.params.get('ignoreerrors') is True:
3351 self.report_error(e)
3352 return infodict
3353 raise
3354
3355 if not files_to_delete:
3356 return infodict
3357 if self.params.get('keepvideo', False):
3358 for f in files_to_delete:
3359 infodict['__files_to_move'].setdefault(f, '')
3360 else:
3361 self._delete_downloaded_files(
3362 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3363 return infodict
3364
3365 def run_all_pps(self, key, info, *, additional_pps=None):
3366 self._forceprint(key, info)
3367 for pp in (additional_pps or []) + self._pps[key]:
3368 info = self.run_pp(pp, info)
3369 return info
3370
3371 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3372 info = dict(ie_info)
3373 info['__files_to_move'] = files_to_move or {}
3374 info = self.run_all_pps(key, info)
3375 return info, info.pop('__files_to_move', None)
3376
3377 def post_process(self, filename, info, files_to_move=None):
3378 """Run all the postprocessors on the given file."""
3379 info['filepath'] = filename
3380 info['__files_to_move'] = files_to_move or {}
3381 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3382 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3383 del info['__files_to_move']
3384 return self.run_all_pps('after_move', info)
3385
3386 def _make_archive_id(self, info_dict):
3387 video_id = info_dict.get('id')
3388 if not video_id:
3389 return
3390 # Future-proof against any change in case
3391 # and backwards compatibility with prior versions
3392 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3393 if extractor is None:
3394 url = str_or_none(info_dict.get('url'))
3395 if not url:
3396 return
3397 # Try to find matching extractor for the URL and take its ie_key
3398 for ie_key, ie in self._ies.items():
3399 if ie.suitable(url):
3400 extractor = ie_key
3401 break
3402 else:
3403 return
3404 return f'{extractor.lower()} {video_id}'
3405
3406 def in_download_archive(self, info_dict):
3407 fn = self.params.get('download_archive')
3408 if fn is None:
3409 return False
3410
3411 vid_id = self._make_archive_id(info_dict)
3412 if not vid_id:
3413 return False # Incomplete video information
3414
3415 return vid_id in self.archive
3416
3417 def record_download_archive(self, info_dict):
3418 fn = self.params.get('download_archive')
3419 if fn is None:
3420 return
3421 vid_id = self._make_archive_id(info_dict)
3422 assert vid_id
3423 self.write_debug(f'Adding to archive: {vid_id}')
3424 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3425 archive_file.write(vid_id + '\n')
3426 self.archive.add(vid_id)
3427
3428 @staticmethod
3429 def format_resolution(format, default='unknown'):
3430 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3431 return 'audio only'
3432 if format.get('resolution') is not None:
3433 return format['resolution']
3434 if format.get('width') and format.get('height'):
3435 return '%dx%d' % (format['width'], format['height'])
3436 elif format.get('height'):
3437 return '%sp' % format['height']
3438 elif format.get('width'):
3439 return '%dx?' % format['width']
3440 return default
3441
3442 def _list_format_headers(self, *headers):
3443 if self.params.get('listformats_table', True) is not False:
3444 return [self._format_screen(header, self.Styles.HEADERS) for header in headers]
3445 return headers
3446
3447 def _format_note(self, fdict):
3448 res = ''
3449 if fdict.get('ext') in ['f4f', 'f4m']:
3450 res += '(unsupported)'
3451 if fdict.get('language'):
3452 if res:
3453 res += ' '
3454 res += '[%s]' % fdict['language']
3455 if fdict.get('format_note') is not None:
3456 if res:
3457 res += ' '
3458 res += fdict['format_note']
3459 if fdict.get('tbr') is not None:
3460 if res:
3461 res += ', '
3462 res += '%4dk' % fdict['tbr']
3463 if fdict.get('container') is not None:
3464 if res:
3465 res += ', '
3466 res += '%s container' % fdict['container']
3467 if (fdict.get('vcodec') is not None
3468 and fdict.get('vcodec') != 'none'):
3469 if res:
3470 res += ', '
3471 res += fdict['vcodec']
3472 if fdict.get('vbr') is not None:
3473 res += '@'
3474 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3475 res += 'video@'
3476 if fdict.get('vbr') is not None:
3477 res += '%4dk' % fdict['vbr']
3478 if fdict.get('fps') is not None:
3479 if res:
3480 res += ', '
3481 res += '%sfps' % fdict['fps']
3482 if fdict.get('acodec') is not None:
3483 if res:
3484 res += ', '
3485 if fdict['acodec'] == 'none':
3486 res += 'video only'
3487 else:
3488 res += '%-5s' % fdict['acodec']
3489 elif fdict.get('abr') is not None:
3490 if res:
3491 res += ', '
3492 res += 'audio'
3493 if fdict.get('abr') is not None:
3494 res += '@%3dk' % fdict['abr']
3495 if fdict.get('asr') is not None:
3496 res += ' (%5dHz)' % fdict['asr']
3497 if fdict.get('filesize') is not None:
3498 if res:
3499 res += ', '
3500 res += format_bytes(fdict['filesize'])
3501 elif fdict.get('filesize_approx') is not None:
3502 if res:
3503 res += ', '
3504 res += '~' + format_bytes(fdict['filesize_approx'])
3505 return res
3506
3507 def render_formats_table(self, info_dict):
3508 if not info_dict.get('formats') and not info_dict.get('url'):
3509 return None
3510
3511 formats = info_dict.get('formats', [info_dict])
3512 if not self.params.get('listformats_table', True) is not False:
3513 table = [
3514 [
3515 format_field(f, 'format_id'),
3516 format_field(f, 'ext'),
3517 self.format_resolution(f),
3518 self._format_note(f)
3519 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3520 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3521
3522 delim = self._format_screen('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3523 table = [
3524 [
3525 self._format_screen(format_field(f, 'format_id'), self.Styles.ID),
3526 format_field(f, 'ext'),
3527 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3528 format_field(f, 'fps', '\t%d'),
3529 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3530 delim,
3531 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3532 format_field(f, 'tbr', '\t%dk'),
3533 shorten_protocol_name(f.get('protocol', '')),
3534 delim,
3535 format_field(f, 'vcodec', default='unknown').replace(
3536 'none', 'images' if f.get('acodec') == 'none'
3537 else self._format_screen('audio only', self.Styles.SUPPRESS)),
3538 format_field(f, 'vbr', '\t%dk'),
3539 format_field(f, 'acodec', default='unknown').replace(
3540 'none', '' if f.get('vcodec') == 'none'
3541 else self._format_screen('video only', self.Styles.SUPPRESS)),
3542 format_field(f, 'abr', '\t%dk'),
3543 format_field(f, 'asr', '\t%dHz'),
3544 join_nonempty(
3545 self._format_screen('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3546 format_field(f, 'language', '[%s]'),
3547 join_nonempty(format_field(f, 'format_note'),
3548 format_field(f, 'container', ignore=(None, f.get('ext'))),
3549 delim=', '),
3550 delim=' '),
3551 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3552 header_line = self._list_format_headers(
3553 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3554 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3555
3556 return render_table(
3557 header_line, table, hide_empty=True,
3558 delim=self._format_screen('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3559
3560 def render_thumbnails_table(self, info_dict):
3561 thumbnails = list(info_dict.get('thumbnails') or [])
3562 if not thumbnails:
3563 return None
3564 return render_table(
3565 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3566 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3567
3568 def render_subtitles_table(self, video_id, subtitles):
3569 def _row(lang, formats):
3570 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3571 if len(set(names)) == 1:
3572 names = [] if names[0] == 'unknown' else names[:1]
3573 return [lang, ', '.join(names), ', '.join(exts)]
3574
3575 if not subtitles:
3576 return None
3577 return render_table(
3578 self._list_format_headers('Language', 'Name', 'Formats'),
3579 [_row(lang, formats) for lang, formats in subtitles.items()],
3580 hide_empty=True)
3581
3582 def __list_table(self, video_id, name, func, *args):
3583 table = func(*args)
3584 if not table:
3585 self.to_screen(f'{video_id} has no {name}')
3586 return
3587 self.to_screen(f'[info] Available {name} for {video_id}:')
3588 self.to_stdout(table)
3589
3590 def list_formats(self, info_dict):
3591 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3592
3593 def list_thumbnails(self, info_dict):
3594 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3595
3596 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3597 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3598
3599 def urlopen(self, req):
3600 """ Start an HTTP download """
3601 if isinstance(req, str):
3602 req = sanitized_Request(req)
3603 return self._opener.open(req, timeout=self._socket_timeout)
3604
3605 def print_debug_header(self):
3606 if not self.params.get('verbose'):
3607 return
3608
3609 def get_encoding(stream):
3610 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3611 if not supports_terminal_sequences(stream):
3612 from .compat import WINDOWS_VT_MODE # Must be imported locally
3613 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3614 return ret
3615
3616 encoding_str = 'Encodings: locale %s, fs %s, out %s, err %s, pref %s' % (
3617 locale.getpreferredencoding(),
3618 sys.getfilesystemencoding(),
3619 get_encoding(self._out_files['screen']), get_encoding(self._out_files['error']),
3620 self.get_encoding())
3621
3622 logger = self.params.get('logger')
3623 if logger:
3624 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3625 write_debug(encoding_str)
3626 else:
3627 write_string(f'[debug] {encoding_str}\n', encoding=None)
3628 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3629
3630 source = detect_variant()
3631 write_debug(join_nonempty(
3632 'yt-dlp version', __version__,
3633 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3634 '' if source == 'unknown' else f'({source})',
3635 delim=' '))
3636 if not _LAZY_LOADER:
3637 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3638 write_debug('Lazy loading extractors is forcibly disabled')
3639 else:
3640 write_debug('Lazy loading extractors is disabled')
3641 if plugin_extractors or plugin_postprocessors:
3642 write_debug('Plugins: %s' % [
3643 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3644 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3645 if self.params.get('compat_opts'):
3646 write_debug('Compatibility options: %s' % ', '.join(self.params.get('compat_opts')))
3647
3648 if source == 'source':
3649 try:
3650 sp = Popen(
3651 ['git', 'rev-parse', '--short', 'HEAD'],
3652 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3653 cwd=os.path.dirname(os.path.abspath(__file__)))
3654 out, err = sp.communicate_or_kill()
3655 out = out.decode().strip()
3656 if re.match('[0-9a-f]+', out):
3657 write_debug('Git HEAD: %s' % out)
3658 except Exception:
3659 with contextlib.suppress(Exception):
3660 sys.exc_clear()
3661
3662 def python_implementation():
3663 impl_name = platform.python_implementation()
3664 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3665 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3666 return impl_name
3667
3668 write_debug('Python version %s (%s %s) - %s' % (
3669 platform.python_version(),
3670 python_implementation(),
3671 platform.architecture()[0],
3672 platform_name()))
3673
3674 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3675 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3676 if ffmpeg_features:
3677 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3678
3679 exe_versions['rtmpdump'] = rtmpdump_version()
3680 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3681 exe_str = ', '.join(
3682 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3683 ) or 'none'
3684 write_debug('exe versions: %s' % exe_str)
3685
3686 from .compat.compat_utils import get_package_info
3687 from .dependencies import available_dependencies
3688
3689 write_debug('Optional libraries: %s' % (', '.join(sorted({
3690 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3691 })) or 'none'))
3692
3693 self._setup_opener()
3694 proxy_map = {}
3695 for handler in self._opener.handlers:
3696 if hasattr(handler, 'proxies'):
3697 proxy_map.update(handler.proxies)
3698 write_debug(f'Proxy map: {proxy_map}')
3699
3700 # Not implemented
3701 if False and self.params.get('call_home'):
3702 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3703 write_debug('Public IP address: %s' % ipaddr)
3704 latest_version = self.urlopen(
3705 'https://yt-dl.org/latest/version').read().decode()
3706 if version_tuple(latest_version) > version_tuple(__version__):
3707 self.report_warning(
3708 'You are using an outdated version (newest version: %s)! '
3709 'See https://yt-dl.org/update if you need help updating.' %
3710 latest_version)
3711
3712 def _setup_opener(self):
3713 if hasattr(self, '_opener'):
3714 return
3715 timeout_val = self.params.get('socket_timeout')
3716 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3717
3718 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3719 opts_cookiefile = self.params.get('cookiefile')
3720 opts_proxy = self.params.get('proxy')
3721
3722 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3723
3724 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3725 if opts_proxy is not None:
3726 if opts_proxy == '':
3727 proxies = {}
3728 else:
3729 proxies = {'http': opts_proxy, 'https': opts_proxy}
3730 else:
3731 proxies = compat_urllib_request.getproxies()
3732 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3733 if 'http' in proxies and 'https' not in proxies:
3734 proxies['https'] = proxies['http']
3735 proxy_handler = PerRequestProxyHandler(proxies)
3736
3737 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3738 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3739 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3740 redirect_handler = YoutubeDLRedirectHandler()
3741 data_handler = urllib.request.DataHandler()
3742
3743 # When passing our own FileHandler instance, build_opener won't add the
3744 # default FileHandler and allows us to disable the file protocol, which
3745 # can be used for malicious purposes (see
3746 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3747 file_handler = compat_urllib_request.FileHandler()
3748
3749 def file_open(*args, **kwargs):
3750 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3751 file_handler.file_open = file_open
3752
3753 opener = compat_urllib_request.build_opener(
3754 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3755
3756 # Delete the default user-agent header, which would otherwise apply in
3757 # cases where our custom HTTP handler doesn't come into play
3758 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3759 opener.addheaders = []
3760 self._opener = opener
3761
3762 def encode(self, s):
3763 if isinstance(s, bytes):
3764 return s # Already encoded
3765
3766 try:
3767 return s.encode(self.get_encoding())
3768 except UnicodeEncodeError as err:
3769 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3770 raise
3771
3772 def get_encoding(self):
3773 encoding = self.params.get('encoding')
3774 if encoding is None:
3775 encoding = preferredencoding()
3776 return encoding
3777
3778 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3779 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3780 if overwrite is None:
3781 overwrite = self.params.get('overwrites', True)
3782 if not self.params.get('writeinfojson'):
3783 return False
3784 elif not infofn:
3785 self.write_debug(f'Skipping writing {label} infojson')
3786 return False
3787 elif not self._ensure_dir_exists(infofn):
3788 return None
3789 elif not overwrite and os.path.exists(infofn):
3790 self.to_screen(f'[info] {label.title()} metadata is already present')
3791 return 'exists'
3792
3793 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3794 try:
3795 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3796 return True
3797 except OSError:
3798 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3799 return None
3800
3801 def _write_description(self, label, ie_result, descfn):
3802 ''' Write description and returns True = written, False = skip, None = error '''
3803 if not self.params.get('writedescription'):
3804 return False
3805 elif not descfn:
3806 self.write_debug(f'Skipping writing {label} description')
3807 return False
3808 elif not self._ensure_dir_exists(descfn):
3809 return None
3810 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3811 self.to_screen(f'[info] {label.title()} description is already present')
3812 elif ie_result.get('description') is None:
3813 self.report_warning(f'There\'s no {label} description to write')
3814 return False
3815 else:
3816 try:
3817 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3818 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3819 descfile.write(ie_result['description'])
3820 except OSError:
3821 self.report_error(f'Cannot write {label} description file {descfn}')
3822 return None
3823 return True
3824
3825 def _write_subtitles(self, info_dict, filename):
3826 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3827 ret = []
3828 subtitles = info_dict.get('requested_subtitles')
3829 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3830 # subtitles download errors are already managed as troubles in relevant IE
3831 # that way it will silently go on when used with unsupporting IE
3832 return ret
3833
3834 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3835 if not sub_filename_base:
3836 self.to_screen('[info] Skipping writing video subtitles')
3837 return ret
3838 for sub_lang, sub_info in subtitles.items():
3839 sub_format = sub_info['ext']
3840 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3841 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3842 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3843 if existing_sub:
3844 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3845 sub_info['filepath'] = existing_sub
3846 ret.append((existing_sub, sub_filename_final))
3847 continue
3848
3849 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3850 if sub_info.get('data') is not None:
3851 try:
3852 # Use newline='' to prevent conversion of newline characters
3853 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3854 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3855 subfile.write(sub_info['data'])
3856 sub_info['filepath'] = sub_filename
3857 ret.append((sub_filename, sub_filename_final))
3858 continue
3859 except OSError:
3860 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3861 return None
3862
3863 try:
3864 sub_copy = sub_info.copy()
3865 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3866 self.dl(sub_filename, sub_copy, subtitle=True)
3867 sub_info['filepath'] = sub_filename
3868 ret.append((sub_filename, sub_filename_final))
3869 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3870 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3871 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3872 if not self.params.get('ignoreerrors'):
3873 self.report_error(msg)
3874 raise DownloadError(msg)
3875 self.report_warning(msg)
3876 return ret
3877
3878 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3879 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3880 write_all = self.params.get('write_all_thumbnails', False)
3881 thumbnails, ret = [], []
3882 if write_all or self.params.get('writethumbnail', False):
3883 thumbnails = info_dict.get('thumbnails') or []
3884 multiple = write_all and len(thumbnails) > 1
3885
3886 if thumb_filename_base is None:
3887 thumb_filename_base = filename
3888 if thumbnails and not thumb_filename_base:
3889 self.write_debug(f'Skipping writing {label} thumbnail')
3890 return ret
3891
3892 for idx, t in list(enumerate(thumbnails))[::-1]:
3893 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3894 thumb_display_id = f'{label} thumbnail {t["id"]}'
3895 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3896 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3897
3898 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3899 if existing_thumb:
3900 self.to_screen('[info] %s is already present' % (
3901 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3902 t['filepath'] = existing_thumb
3903 ret.append((existing_thumb, thumb_filename_final))
3904 else:
3905 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3906 try:
3907 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3908 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3909 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3910 shutil.copyfileobj(uf, thumbf)
3911 ret.append((thumb_filename, thumb_filename_final))
3912 t['filepath'] = thumb_filename
3913 except network_exceptions as err:
3914 thumbnails.pop(idx)
3915 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3916 if ret and not write_all:
3917 break
3918 return ret