]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
eaf2d9216788d137df6f1475bdae72c84768bbfb
[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 enum import Enum
27 from string import ascii_letters
28
29 from .cache import Cache
30 from .compat import (
31 compat_brotli,
32 compat_get_terminal_size,
33 compat_os_name,
34 compat_pycrypto_AES,
35 compat_shlex_quote,
36 compat_str,
37 compat_urllib_error,
38 compat_urllib_request,
39 windows_enable_vt_mode,
40 )
41 from .cookies import load_cookies
42 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
43 from .downloader.rtmp import rtmpdump_version
44 from .extractor import _LAZY_LOADER
45 from .extractor import _PLUGIN_CLASSES as plugin_extractors
46 from .extractor import gen_extractor_classes, get_info_extractor
47 from .extractor.openload import PhantomJSwrapper
48 from .minicurses import format_text
49 from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
50 from .postprocessor import (
51 EmbedThumbnailPP,
52 FFmpegFixupDuplicateMoovPP,
53 FFmpegFixupDurationPP,
54 FFmpegFixupM3u8PP,
55 FFmpegFixupM4aPP,
56 FFmpegFixupStretchedPP,
57 FFmpegFixupTimestampPP,
58 FFmpegMergerPP,
59 FFmpegPostProcessor,
60 MoveFilesAfterDownloadPP,
61 get_postprocessor,
62 )
63 from .update import detect_variant
64 from .utils import (
65 DEFAULT_OUTTMPL,
66 LINK_TEMPLATES,
67 NO_DEFAULT,
68 OUTTMPL_TYPES,
69 POSTPROCESS_WHEN,
70 STR_FORMAT_RE_TMPL,
71 STR_FORMAT_TYPES,
72 ContentTooShortError,
73 DateRange,
74 DownloadCancelled,
75 DownloadError,
76 EntryNotInPlaylist,
77 ExistingVideoReached,
78 ExtractorError,
79 GeoRestrictedError,
80 HEADRequest,
81 InAdvancePagedList,
82 ISO3166Utils,
83 LazyList,
84 MaxDownloadsReached,
85 PagedList,
86 PerRequestProxyHandler,
87 Popen,
88 PostProcessingError,
89 ReExtractInfo,
90 RejectedVideoReached,
91 SameFileError,
92 UnavailableVideoError,
93 YoutubeDLCookieProcessor,
94 YoutubeDLHandler,
95 YoutubeDLRedirectHandler,
96 age_restricted,
97 args_to_str,
98 date_from_str,
99 determine_ext,
100 determine_protocol,
101 encode_compat_str,
102 encodeFilename,
103 error_to_compat_str,
104 expand_path,
105 filter_dict,
106 float_or_none,
107 format_bytes,
108 format_decimal_suffix,
109 format_field,
110 formatSeconds,
111 get_domain,
112 has_certifi,
113 int_or_none,
114 iri_to_uri,
115 join_nonempty,
116 locked_file,
117 make_dir,
118 make_HTTPS_handler,
119 merge_headers,
120 network_exceptions,
121 number_of_digits,
122 orderedSet,
123 parse_filesize,
124 platform_name,
125 preferredencoding,
126 prepend_extension,
127 register_socks_protocols,
128 remove_terminal_sequences,
129 render_table,
130 replace_extension,
131 sanitize_filename,
132 sanitize_path,
133 sanitize_url,
134 sanitized_Request,
135 std_headers,
136 str_or_none,
137 strftime_or_none,
138 subtitles_filename,
139 supports_terminal_sequences,
140 timetuple_from_msec,
141 to_high_limit_path,
142 traverse_obj,
143 try_get,
144 url_basename,
145 variadic,
146 version_tuple,
147 write_json_file,
148 write_string,
149 )
150 from .version import RELEASE_GIT_HEAD, __version__
151
152 if compat_os_name == 'nt':
153 import ctypes
154
155
156 class YoutubeDL:
157 """YoutubeDL class.
158
159 YoutubeDL objects are the ones responsible of downloading the
160 actual video file and writing it to disk if the user has requested
161 it, among some other tasks. In most cases there should be one per
162 program. As, given a video URL, the downloader doesn't know how to
163 extract all the needed information, task that InfoExtractors do, it
164 has to pass the URL to one of them.
165
166 For this, YoutubeDL objects have a method that allows
167 InfoExtractors to be registered in a given order. When it is passed
168 a URL, the YoutubeDL object handles it to the first InfoExtractor it
169 finds that reports being able to handle it. The InfoExtractor extracts
170 all the information about the video or videos the URL refers to, and
171 YoutubeDL process the extracted information, possibly using a File
172 Downloader to download the video.
173
174 YoutubeDL objects accept a lot of parameters. In order not to saturate
175 the object constructor with arguments, it receives a dictionary of
176 options instead. These options are available through the params
177 attribute for the InfoExtractors to use. The YoutubeDL also
178 registers itself as the downloader in charge for the InfoExtractors
179 that are added to it, so this is a "mutual registration".
180
181 Available options:
182
183 username: Username for authentication purposes.
184 password: Password for authentication purposes.
185 videopassword: Password for accessing a video.
186 ap_mso: Adobe Pass multiple-system operator identifier.
187 ap_username: Multiple-system operator account username.
188 ap_password: Multiple-system operator account password.
189 usenetrc: Use netrc for authentication instead.
190 verbose: Print additional info to stdout.
191 quiet: Do not print messages to stdout.
192 no_warnings: Do not print out anything for warnings.
193 forceprint: A dict with keys WHEN mapped to a list of templates to
194 print to stdout. The allowed keys are video or any of the
195 items in utils.POSTPROCESS_WHEN.
196 For compatibility, a single list is also accepted
197 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
198 a list of tuples with (template, filename)
199 forceurl: Force printing final URL. (Deprecated)
200 forcetitle: Force printing title. (Deprecated)
201 forceid: Force printing ID. (Deprecated)
202 forcethumbnail: Force printing thumbnail URL. (Deprecated)
203 forcedescription: Force printing description. (Deprecated)
204 forcefilename: Force printing final filename. (Deprecated)
205 forceduration: Force printing duration. (Deprecated)
206 forcejson: Force printing info_dict as JSON.
207 dump_single_json: Force printing the info_dict of the whole playlist
208 (or video) as a single JSON line.
209 force_write_download_archive: Force writing download archive regardless
210 of 'skip_download' or 'simulate'.
211 simulate: Do not download the video files. If unset (or None),
212 simulate only if listsubtitles, listformats or list_thumbnails is used
213 format: Video format code. see "FORMAT SELECTION" for more details.
214 You can also pass a function. The function takes 'ctx' as
215 argument and returns the formats to download.
216 See "build_format_selector" for an implementation
217 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
218 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
219 extracting metadata even if the video is not actually
220 available for download (experimental)
221 format_sort: A list of fields by which to sort the video formats.
222 See "Sorting Formats" for more details.
223 format_sort_force: Force the given format_sort. see "Sorting Formats"
224 for more details.
225 prefer_free_formats: Whether to prefer video formats with free containers
226 over non-free ones of same quality.
227 allow_multiple_video_streams: Allow multiple video streams to be merged
228 into a single file
229 allow_multiple_audio_streams: Allow multiple audio streams to be merged
230 into a single file
231 check_formats Whether to test if the formats are downloadable.
232 Can be True (check all), False (check none),
233 'selected' (check selected formats),
234 or None (check only if requested by extractor)
235 paths: Dictionary of output paths. The allowed keys are 'home'
236 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
237 outtmpl: Dictionary of templates for output names. Allowed keys
238 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
239 For compatibility with youtube-dl, a single string can also be used
240 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
241 restrictfilenames: Do not allow "&" and spaces in file names
242 trim_file_name: Limit length of filename (extension excluded)
243 windowsfilenames: Force the filenames to be windows compatible
244 ignoreerrors: Do not stop on download/postprocessing errors.
245 Can be 'only_download' to ignore only download errors.
246 Default is 'only_download' for CLI, but False for API
247 skip_playlist_after_errors: Number of allowed failures until the rest of
248 the playlist is skipped
249 force_generic_extractor: Force downloader to use the generic extractor
250 overwrites: Overwrite all video and metadata files if True,
251 overwrite only non-video files if None
252 and don't overwrite any file if False
253 For compatibility with youtube-dl,
254 "nooverwrites" may also be used instead
255 playliststart: Playlist item to start at.
256 playlistend: Playlist item to end at.
257 playlist_items: Specific indices of playlist to download.
258 playlistreverse: Download playlist items in reverse order.
259 playlistrandom: Download playlist items in random order.
260 matchtitle: Download only matching titles.
261 rejecttitle: Reject downloads for matching titles.
262 logger: Log messages to a logging.Logger instance.
263 logtostderr: Log messages to stderr instead of stdout.
264 consoletitle: Display progress in console window's titlebar.
265 writedescription: Write the video description to a .description file
266 writeinfojson: Write the video description to a .info.json file
267 clean_infojson: Remove private fields from the infojson
268 getcomments: Extract video comments. This will not be written to disk
269 unless writeinfojson is also given
270 writeannotations: Write the video annotations to a .annotations.xml file
271 writethumbnail: Write the thumbnail image to a file
272 allow_playlist_files: Whether to write playlists' description, infojson etc
273 also to disk when using the 'write*' options
274 write_all_thumbnails: Write all thumbnail formats to files
275 writelink: Write an internet shortcut file, depending on the
276 current platform (.url/.webloc/.desktop)
277 writeurllink: Write a Windows internet shortcut file (.url)
278 writewebloclink: Write a macOS internet shortcut file (.webloc)
279 writedesktoplink: Write a Linux internet shortcut file (.desktop)
280 writesubtitles: Write the video subtitles to a file
281 writeautomaticsub: Write the automatically generated subtitles to a file
282 allsubtitles: Deprecated - Use subtitleslangs = ['all']
283 Downloads all the subtitles of the video
284 (requires writesubtitles or writeautomaticsub)
285 listsubtitles: Lists all available subtitles for the video
286 subtitlesformat: The format code for subtitles
287 subtitleslangs: List of languages of the subtitles to download (can be regex).
288 The list may contain "all" to refer to all the available
289 subtitles. The language can be prefixed with a "-" to
290 exclude it from the requested languages. Eg: ['all', '-live_chat']
291 keepvideo: Keep the video file after post-processing
292 daterange: A DateRange object, download only if the upload_date is in the range.
293 skip_download: Skip the actual download of the video file
294 cachedir: Location of the cache files in the filesystem.
295 False to disable filesystem cache.
296 noplaylist: Download single video instead of a playlist if in doubt.
297 age_limit: An integer representing the user's age in years.
298 Unsuitable videos for the given age are skipped.
299 min_views: An integer representing the minimum view count the video
300 must have in order to not be skipped.
301 Videos without view count information are always
302 downloaded. None for no limit.
303 max_views: An integer representing the maximum view count.
304 Videos that are more popular than that are not
305 downloaded.
306 Videos without view count information are always
307 downloaded. None for no limit.
308 download_archive: File name of a file where all downloads are recorded.
309 Videos already present in the file are not downloaded
310 again.
311 break_on_existing: Stop the download process after attempting to download a
312 file that is in the archive.
313 break_on_reject: Stop the download process when encountering a video that
314 has been filtered out.
315 break_per_url: Whether break_on_reject and break_on_existing
316 should act on each input URL as opposed to for the entire queue
317 cookiefile: File name where cookies should be read from and dumped to
318 cookiesfrombrowser: A tuple containing the name of the browser, the profile
319 name/pathfrom where cookies are loaded, and the name of the
320 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
321 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
322 support RFC 5746 secure renegotiation
323 nocheckcertificate: Do not verify SSL certificates
324 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
325 At the moment, this is only supported by YouTube.
326 http_headers: A dictionary of custom headers to be used for all requests
327 proxy: URL of the proxy server to use
328 geo_verification_proxy: URL of the proxy to use for IP address verification
329 on geo-restricted sites.
330 socket_timeout: Time to wait for unresponsive hosts, in seconds
331 bidi_workaround: Work around buggy terminals without bidirectional text
332 support, using fridibi
333 debug_printtraffic:Print out sent and received HTTP traffic
334 include_ads: Download ads as well (deprecated)
335 default_search: Prepend this string if an input url is not valid.
336 'auto' for elaborate guessing
337 encoding: Use this encoding instead of the system-specified.
338 extract_flat: Do not resolve URLs, return the immediate result.
339 Pass in 'in_playlist' to only show this behavior for
340 playlist items.
341 wait_for_video: If given, wait for scheduled streams to become available.
342 The value should be a tuple containing the range
343 (min_secs, max_secs) to wait between retries
344 postprocessors: A list of dictionaries, each with an entry
345 * key: The name of the postprocessor. See
346 yt_dlp/postprocessor/__init__.py for a list.
347 * when: When to run the postprocessor. Allowed values are
348 the entries of utils.POSTPROCESS_WHEN
349 Assumed to be 'post_process' if not given
350 post_hooks: Deprecated - Register a custom postprocessor instead
351 A list of functions that get called as the final step
352 for each video file, after all postprocessors have been
353 called. The filename will be passed as the only argument.
354 progress_hooks: A list of functions that get called on download
355 progress, with a dictionary with the entries
356 * status: One of "downloading", "error", or "finished".
357 Check this first and ignore unknown values.
358 * info_dict: The extracted info_dict
359
360 If status is one of "downloading", or "finished", the
361 following properties may also be present:
362 * filename: The final filename (always present)
363 * tmpfilename: The filename we're currently writing to
364 * downloaded_bytes: Bytes on disk
365 * total_bytes: Size of the whole file, None if unknown
366 * total_bytes_estimate: Guess of the eventual file size,
367 None if unavailable.
368 * elapsed: The number of seconds since download started.
369 * eta: The estimated time in seconds, None if unknown
370 * speed: The download speed in bytes/second, None if
371 unknown
372 * fragment_index: The counter of the currently
373 downloaded video fragment.
374 * fragment_count: The number of fragments (= individual
375 files that will be merged)
376
377 Progress hooks are guaranteed to be called at least once
378 (with status "finished") if the download is successful.
379 postprocessor_hooks: A list of functions that get called on postprocessing
380 progress, with a dictionary with the entries
381 * status: One of "started", "processing", or "finished".
382 Check this first and ignore unknown values.
383 * postprocessor: Name of the postprocessor
384 * info_dict: The extracted info_dict
385
386 Progress hooks are guaranteed to be called at least twice
387 (with status "started" and "finished") if the processing is successful.
388 merge_output_format: Extension to use when merging formats.
389 final_ext: Expected final extension; used to detect when the file was
390 already downloaded and converted
391 fixup: Automatically correct known faults of the file.
392 One of:
393 - "never": do nothing
394 - "warn": only emit a warning
395 - "detect_or_warn": check whether we can do anything
396 about it, warn otherwise (default)
397 source_address: Client-side IP address to bind to.
398 call_home: Boolean, true iff we are allowed to contact the
399 yt-dlp servers for debugging. (BROKEN)
400 sleep_interval_requests: Number of seconds to sleep between requests
401 during extraction
402 sleep_interval: Number of seconds to sleep before each download when
403 used alone or a lower bound of a range for randomized
404 sleep before each download (minimum possible number
405 of seconds to sleep) when used along with
406 max_sleep_interval.
407 max_sleep_interval:Upper bound of a range for randomized sleep before each
408 download (maximum possible number of seconds to sleep).
409 Must only be used along with sleep_interval.
410 Actual sleep time will be a random float from range
411 [sleep_interval; max_sleep_interval].
412 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
413 listformats: Print an overview of available video formats and exit.
414 list_thumbnails: Print a table of all thumbnails and exit.
415 match_filter: A function that gets called with the info_dict of
416 every video.
417 If it returns a message, the video is ignored.
418 If it returns None, the video is downloaded.
419 match_filter_func in utils.py is one example for this.
420 no_color: Do not emit color codes in output.
421 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
422 HTTP header
423 geo_bypass_country:
424 Two-letter ISO 3166-2 country code that will be used for
425 explicit geographic restriction bypassing via faking
426 X-Forwarded-For HTTP header
427 geo_bypass_ip_block:
428 IP range in CIDR notation that will be used similarly to
429 geo_bypass_country
430
431 The following options determine which downloader is picked:
432 external_downloader: A dictionary of protocol keys and the executable of the
433 external downloader to use for it. The allowed protocols
434 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
435 Set the value to 'native' to use the native downloader
436 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
437 or {'m3u8': 'ffmpeg'} instead.
438 Use the native HLS downloader instead of ffmpeg/avconv
439 if True, otherwise use ffmpeg/avconv if False, otherwise
440 use downloader suggested by extractor if None.
441 compat_opts: Compatibility options. See "Differences in default behavior".
442 The following options do not work when used through the API:
443 filename, abort-on-error, multistreams, no-live-chat, format-sort
444 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
445 Refer __init__.py for their implementation
446 progress_template: Dictionary of templates for progress outputs.
447 Allowed keys are 'download', 'postprocess',
448 'download-title' (console title) and 'postprocess-title'.
449 The template is mapped on a dictionary with keys 'progress' and 'info'
450
451 The following parameters are not used by YoutubeDL itself, they are used by
452 the downloader (see yt_dlp/downloader/common.py):
453 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
454 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
455 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
456 external_downloader_args, concurrent_fragment_downloads.
457
458 The following options are used by the post processors:
459 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
460 otherwise prefer ffmpeg. (avconv support is deprecated)
461 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
462 to the binary or its containing directory.
463 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
464 and a list of additional command-line arguments for the
465 postprocessor/executable. The dict can also have "PP+EXE" keys
466 which are used when the given exe is used by the given PP.
467 Use 'default' as the name for arguments to passed to all PP
468 For compatibility with youtube-dl, a single list of args
469 can also be used
470
471 The following options are used by the extractors:
472 extractor_retries: Number of times to retry for known errors
473 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
474 hls_split_discontinuity: Split HLS playlists to different formats at
475 discontinuities such as ad breaks (default: False)
476 extractor_args: A dictionary of arguments to be passed to the extractors.
477 See "EXTRACTOR ARGUMENTS" for details.
478 Eg: {'youtube': {'skip': ['dash', 'hls']}}
479 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
480 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
481 If True (default), DASH manifests and related
482 data will be downloaded and processed by extractor.
483 You can reduce network I/O by disabling it if you don't
484 care about DASH. (only for youtube)
485 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
486 If True (default), HLS manifests and related
487 data will be downloaded and processed by extractor.
488 You can reduce network I/O by disabling it if you don't
489 care about HLS. (only for youtube)
490 """
491
492 _NUMERIC_FIELDS = {
493 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
494 'timestamp', 'release_timestamp',
495 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
496 'average_rating', 'comment_count', 'age_limit',
497 'start_time', 'end_time',
498 'chapter_number', 'season_number', 'episode_number',
499 'track_number', 'disc_number', 'release_year',
500 }
501
502 _format_fields = {
503 # NB: Keep in sync with the docstring of extractor/common.py
504 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
505 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
506 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
507 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
508 'preference', 'language', 'language_preference', 'quality', 'source_preference',
509 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
510 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
511 }
512 _format_selection_exts = {
513 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
514 'video': {'mp4', 'flv', 'webm', '3gp'},
515 'storyboards': {'mhtml'},
516 }
517
518 def __init__(self, params=None, auto_init=True):
519 """Create a FileDownloader object with the given options.
520 @param auto_init Whether to load the default extractors and print header (if verbose).
521 Set to 'no_verbose_header' to not print the header
522 """
523 if params is None:
524 params = {}
525 self.params = params
526 self._ies = {}
527 self._ies_instances = {}
528 self._pps = {k: [] for k in POSTPROCESS_WHEN}
529 self._printed_messages = set()
530 self._first_webpage_request = True
531 self._post_hooks = []
532 self._progress_hooks = []
533 self._postprocessor_hooks = []
534 self._download_retcode = 0
535 self._num_downloads = 0
536 self._num_videos = 0
537 self._playlist_level = 0
538 self._playlist_urls = set()
539 self.cache = Cache(self)
540
541 windows_enable_vt_mode()
542 self._out_files = {
543 'error': sys.stderr,
544 'print': sys.stderr if self.params.get('logtostderr') else sys.stdout,
545 'console': None if compat_os_name == 'nt' else next(
546 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
547 }
548 self._out_files['screen'] = sys.stderr if self.params.get('quiet') else self._out_files['print']
549 self._allow_colors = {
550 type_: not self.params.get('no_color') and supports_terminal_sequences(self._out_files[type_])
551 for type_ in ('screen', 'error')
552 }
553
554 if sys.version_info < (3, 6):
555 self.report_warning(
556 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
557
558 if self.params.get('allow_unplayable_formats'):
559 self.report_warning(
560 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
561 'This is a developer option intended for debugging. \n'
562 ' If you experience any issues while using this option, '
563 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
564
565 def check_deprecated(param, option, suggestion):
566 if self.params.get(param) is not None:
567 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
568 return True
569 return False
570
571 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
572 if self.params.get('geo_verification_proxy') is None:
573 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
574
575 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
576 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
577 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
578
579 for msg in self.params.get('_warnings', []):
580 self.report_warning(msg)
581 for msg in self.params.get('_deprecation_warnings', []):
582 self.deprecation_warning(msg)
583
584 if 'list-formats' in self.params.get('compat_opts', []):
585 self.params['listformats_table'] = False
586
587 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
588 # nooverwrites was unnecessarily changed to overwrites
589 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
590 # This ensures compatibility with both keys
591 self.params['overwrites'] = not self.params['nooverwrites']
592 elif self.params.get('overwrites') is None:
593 self.params.pop('overwrites', None)
594 else:
595 self.params['nooverwrites'] = not self.params['overwrites']
596
597 self.params.setdefault('forceprint', {})
598 self.params.setdefault('print_to_file', {})
599
600 # Compatibility with older syntax
601 if not isinstance(params['forceprint'], dict):
602 self.params['forceprint'] = {'video': params['forceprint']}
603
604 if self.params.get('bidi_workaround', False):
605 try:
606 import pty
607 master, slave = pty.openpty()
608 width = compat_get_terminal_size().columns
609 if width is None:
610 width_args = []
611 else:
612 width_args = ['-w', str(width)]
613 sp_kwargs = dict(
614 stdin=subprocess.PIPE,
615 stdout=slave,
616 stderr=self._out_files['error'])
617 try:
618 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
619 except OSError:
620 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
621 self._output_channel = os.fdopen(master, 'rb')
622 except OSError as ose:
623 if ose.errno == errno.ENOENT:
624 self.report_warning(
625 'Could not find fribidi executable, ignoring --bidi-workaround. '
626 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
627 else:
628 raise
629
630 if auto_init:
631 if auto_init != 'no_verbose_header':
632 self.print_debug_header()
633 self.add_default_info_extractors()
634
635 if (sys.platform != 'win32'
636 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
637 and not self.params.get('restrictfilenames', False)):
638 # Unicode filesystem API will throw errors (#1474, #13027)
639 self.report_warning(
640 'Assuming --restrict-filenames since file system encoding '
641 'cannot encode all characters. '
642 'Set the LC_ALL environment variable to fix this.')
643 self.params['restrictfilenames'] = True
644
645 self.outtmpl_dict = self.parse_outtmpl()
646
647 # Creating format selector here allows us to catch syntax errors before the extraction
648 self.format_selector = (
649 self.params.get('format') if self.params.get('format') in (None, '-')
650 else self.params['format'] if callable(self.params['format'])
651 else self.build_format_selector(self.params['format']))
652
653 # Set http_headers defaults according to std_headers
654 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
655
656 hooks = {
657 'post_hooks': self.add_post_hook,
658 'progress_hooks': self.add_progress_hook,
659 'postprocessor_hooks': self.add_postprocessor_hook,
660 }
661 for opt, fn in hooks.items():
662 for ph in self.params.get(opt, []):
663 fn(ph)
664
665 for pp_def_raw in self.params.get('postprocessors', []):
666 pp_def = dict(pp_def_raw)
667 when = pp_def.pop('when', 'post_process')
668 self.add_post_processor(
669 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
670 when=when)
671
672 self._setup_opener()
673 register_socks_protocols()
674
675 def preload_download_archive(fn):
676 """Preload the archive, if any is specified"""
677 if fn is None:
678 return False
679 self.write_debug(f'Loading archive file {fn!r}')
680 try:
681 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
682 for line in archive_file:
683 self.archive.add(line.strip())
684 except OSError as ioe:
685 if ioe.errno != errno.ENOENT:
686 raise
687 return False
688 return True
689
690 self.archive = set()
691 preload_download_archive(self.params.get('download_archive'))
692
693 def warn_if_short_id(self, argv):
694 # short YouTube ID starting with dash?
695 idxs = [
696 i for i, a in enumerate(argv)
697 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
698 if idxs:
699 correct_argv = (
700 ['yt-dlp']
701 + [a for i, a in enumerate(argv) if i not in idxs]
702 + ['--'] + [argv[i] for i in idxs]
703 )
704 self.report_warning(
705 'Long argument string detected. '
706 'Use -- to separate parameters and URLs, like this:\n%s' %
707 args_to_str(correct_argv))
708
709 def add_info_extractor(self, ie):
710 """Add an InfoExtractor object to the end of the list."""
711 ie_key = ie.ie_key()
712 self._ies[ie_key] = ie
713 if not isinstance(ie, type):
714 self._ies_instances[ie_key] = ie
715 ie.set_downloader(self)
716
717 def _get_info_extractor_class(self, ie_key):
718 ie = self._ies.get(ie_key)
719 if ie is None:
720 ie = get_info_extractor(ie_key)
721 self.add_info_extractor(ie)
722 return ie
723
724 def get_info_extractor(self, ie_key):
725 """
726 Get an instance of an IE with name ie_key, it will try to get one from
727 the _ies list, if there's no instance it will create a new one and add
728 it to the extractor list.
729 """
730 ie = self._ies_instances.get(ie_key)
731 if ie is None:
732 ie = get_info_extractor(ie_key)()
733 self.add_info_extractor(ie)
734 return ie
735
736 def add_default_info_extractors(self):
737 """
738 Add the InfoExtractors returned by gen_extractors to the end of the list
739 """
740 for ie in gen_extractor_classes():
741 self.add_info_extractor(ie)
742
743 def add_post_processor(self, pp, when='post_process'):
744 """Add a PostProcessor object to the end of the chain."""
745 self._pps[when].append(pp)
746 pp.set_downloader(self)
747
748 def add_post_hook(self, ph):
749 """Add the post hook"""
750 self._post_hooks.append(ph)
751
752 def add_progress_hook(self, ph):
753 """Add the download progress hook"""
754 self._progress_hooks.append(ph)
755
756 def add_postprocessor_hook(self, ph):
757 """Add the postprocessing progress hook"""
758 self._postprocessor_hooks.append(ph)
759 for pps in self._pps.values():
760 for pp in pps:
761 pp.add_progress_hook(ph)
762
763 def _bidi_workaround(self, message):
764 if not hasattr(self, '_output_channel'):
765 return message
766
767 assert hasattr(self, '_output_process')
768 assert isinstance(message, compat_str)
769 line_count = message.count('\n') + 1
770 self._output_process.stdin.write((message + '\n').encode('utf-8'))
771 self._output_process.stdin.flush()
772 res = ''.join(self._output_channel.readline().decode('utf-8')
773 for _ in range(line_count))
774 return res[:-len('\n')]
775
776 def _write_string(self, message, out=None, only_once=False):
777 if only_once:
778 if message in self._printed_messages:
779 return
780 self._printed_messages.add(message)
781 write_string(message, out=out, encoding=self.params.get('encoding'))
782
783 def to_stdout(self, message, skip_eol=False, quiet=None):
784 """Print message to stdout"""
785 if quiet is not None:
786 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
787 self._write_string(
788 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
789 self._out_files['print'])
790
791 def to_screen(self, message, skip_eol=False, quiet=None):
792 """Print message to screen if not in quiet mode"""
793 if self.params.get('logger'):
794 self.params['logger'].debug(message)
795 return
796 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
797 return
798 self._write_string(
799 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
800 self._out_files['screen'])
801
802 def to_stderr(self, message, only_once=False):
803 """Print message to stderr"""
804 assert isinstance(message, compat_str)
805 if self.params.get('logger'):
806 self.params['logger'].error(message)
807 else:
808 self._write_string('%s\n' % self._bidi_workaround(message), self._out_files['error'], only_once=only_once)
809
810 def _send_console_code(self, code):
811 if compat_os_name == 'nt' or not self._out_files['console']:
812 return
813 self._write_string(code, self._out_files['console'])
814
815 def to_console_title(self, message):
816 if not self.params.get('consoletitle', False):
817 return
818 message = remove_terminal_sequences(message)
819 if compat_os_name == 'nt':
820 if ctypes.windll.kernel32.GetConsoleWindow():
821 # c_wchar_p() might not be necessary if `message` is
822 # already of type unicode()
823 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
824 else:
825 self._send_console_code(f'\033]0;{message}\007')
826
827 def save_console_title(self):
828 if not self.params.get('consoletitle') or self.params.get('simulate'):
829 return
830 self._send_console_code('\033[22;0t') # Save the title on stack
831
832 def restore_console_title(self):
833 if not self.params.get('consoletitle') or self.params.get('simulate'):
834 return
835 self._send_console_code('\033[23;0t') # Restore the title from stack
836
837 def __enter__(self):
838 self.save_console_title()
839 return self
840
841 def __exit__(self, *args):
842 self.restore_console_title()
843
844 if self.params.get('cookiefile') is not None:
845 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
846
847 def trouble(self, message=None, tb=None, is_error=True):
848 """Determine action to take when a download problem appears.
849
850 Depending on if the downloader has been configured to ignore
851 download errors or not, this method may throw an exception or
852 not when errors are found, after printing the message.
853
854 @param tb If given, is additional traceback information
855 @param is_error Whether to raise error according to ignorerrors
856 """
857 if message is not None:
858 self.to_stderr(message)
859 if self.params.get('verbose'):
860 if tb is None:
861 if sys.exc_info()[0]: # if .trouble has been called from an except block
862 tb = ''
863 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
864 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
865 tb += encode_compat_str(traceback.format_exc())
866 else:
867 tb_data = traceback.format_list(traceback.extract_stack())
868 tb = ''.join(tb_data)
869 if tb:
870 self.to_stderr(tb)
871 if not is_error:
872 return
873 if not self.params.get('ignoreerrors'):
874 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
875 exc_info = sys.exc_info()[1].exc_info
876 else:
877 exc_info = sys.exc_info()
878 raise DownloadError(message, exc_info)
879 self._download_retcode = 1
880
881 class Styles(Enum):
882 HEADERS = 'yellow'
883 EMPHASIS = 'light blue'
884 ID = 'green'
885 DELIM = 'blue'
886 ERROR = 'red'
887 WARNING = 'yellow'
888 SUPPRESS = 'light black'
889
890 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
891 text = str(text)
892 if test_encoding:
893 original_text = text
894 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
895 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
896 text = text.encode(encoding, 'ignore').decode(encoding)
897 if fallback is not None and text != original_text:
898 text = fallback
899 if isinstance(f, Enum):
900 f = f.value
901 return format_text(text, f) if allow_colors else text if fallback is None else fallback
902
903 def _format_screen(self, *args, **kwargs):
904 return self._format_text(
905 self._out_files['screen'], self._allow_colors['screen'], *args, **kwargs)
906
907 def _format_err(self, *args, **kwargs):
908 return self._format_text(
909 self._out_files['error'], self._allow_colors['error'], *args, **kwargs)
910
911 def report_warning(self, message, only_once=False):
912 '''
913 Print the message to stderr, it will be prefixed with 'WARNING:'
914 If stderr is a tty file the 'WARNING:' will be colored
915 '''
916 if self.params.get('logger') is not None:
917 self.params['logger'].warning(message)
918 else:
919 if self.params.get('no_warnings'):
920 return
921 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
922
923 def deprecation_warning(self, message):
924 if self.params.get('logger') is not None:
925 self.params['logger'].warning(f'DeprecationWarning: {message}')
926 else:
927 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
928
929 def report_error(self, message, *args, **kwargs):
930 '''
931 Do the same as trouble, but prefixes the message with 'ERROR:', colored
932 in red if stderr is a tty file.
933 '''
934 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
935
936 def write_debug(self, message, only_once=False):
937 '''Log debug message or Print message to stderr'''
938 if not self.params.get('verbose', False):
939 return
940 message = '[debug] %s' % message
941 if self.params.get('logger'):
942 self.params['logger'].debug(message)
943 else:
944 self.to_stderr(message, only_once)
945
946 def report_file_already_downloaded(self, file_name):
947 """Report file has already been fully downloaded."""
948 try:
949 self.to_screen('[download] %s has already been downloaded' % file_name)
950 except UnicodeEncodeError:
951 self.to_screen('[download] The file has already been downloaded')
952
953 def report_file_delete(self, file_name):
954 """Report that existing file will be deleted."""
955 try:
956 self.to_screen('Deleting existing file %s' % file_name)
957 except UnicodeEncodeError:
958 self.to_screen('Deleting existing file')
959
960 def raise_no_formats(self, info, forced=False, *, msg=None):
961 has_drm = info.get('__has_drm')
962 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
963 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
964 if forced or not ignored:
965 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
966 expected=has_drm or ignored or expected)
967 else:
968 self.report_warning(msg)
969
970 def parse_outtmpl(self):
971 outtmpl_dict = self.params.get('outtmpl', {})
972 if not isinstance(outtmpl_dict, dict):
973 outtmpl_dict = {'default': outtmpl_dict}
974 # Remove spaces in the default template
975 if self.params.get('restrictfilenames'):
976 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
977 else:
978 sanitize = lambda x: x
979 outtmpl_dict.update({
980 k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items()
981 if outtmpl_dict.get(k) is None})
982 for _, val in outtmpl_dict.items():
983 if isinstance(val, bytes):
984 self.report_warning('Parameter outtmpl is bytes, but should be a unicode string')
985 return outtmpl_dict
986
987 def get_output_path(self, dir_type='', filename=None):
988 paths = self.params.get('paths', {})
989 assert isinstance(paths, dict)
990 path = os.path.join(
991 expand_path(paths.get('home', '').strip()),
992 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
993 filename or '')
994 return sanitize_path(path, force=self.params.get('windowsfilenames'))
995
996 @staticmethod
997 def _outtmpl_expandpath(outtmpl):
998 # expand_path translates '%%' into '%' and '$$' into '$'
999 # correspondingly that is not what we want since we need to keep
1000 # '%%' intact for template dict substitution step. Working around
1001 # with boundary-alike separator hack.
1002 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1003 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1004
1005 # outtmpl should be expand_path'ed before template dict substitution
1006 # because meta fields may contain env variables we don't want to
1007 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1008 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1009 return expand_path(outtmpl).replace(sep, '')
1010
1011 @staticmethod
1012 def escape_outtmpl(outtmpl):
1013 ''' Escape any remaining strings like %s, %abc% etc. '''
1014 return re.sub(
1015 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1016 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1017 outtmpl)
1018
1019 @classmethod
1020 def validate_outtmpl(cls, outtmpl):
1021 ''' @return None or Exception object '''
1022 outtmpl = re.sub(
1023 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1024 lambda mobj: f'{mobj.group(0)[:-1]}s',
1025 cls._outtmpl_expandpath(outtmpl))
1026 try:
1027 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1028 return None
1029 except ValueError as err:
1030 return err
1031
1032 @staticmethod
1033 def _copy_infodict(info_dict):
1034 info_dict = dict(info_dict)
1035 info_dict.pop('__postprocessors', None)
1036 return info_dict
1037
1038 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1039 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1040 @param sanitize Whether to sanitize the output as a filename.
1041 For backward compatibility, a function can also be passed
1042 """
1043
1044 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1045
1046 info_dict = self._copy_infodict(info_dict)
1047 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1048 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1049 if info_dict.get('duration', None) is not None
1050 else None)
1051 info_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
1052 info_dict['video_autonumber'] = self._num_videos
1053 if info_dict.get('resolution') is None:
1054 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1055
1056 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1057 # of %(field)s to %(field)0Nd for backward compatibility
1058 field_size_compat_map = {
1059 'playlist_index': number_of_digits(info_dict.get('_last_playlist_index') or 0),
1060 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1061 'autonumber': self.params.get('autonumber_size') or 5,
1062 }
1063
1064 TMPL_DICT = {}
1065 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1066 MATH_FUNCTIONS = {
1067 '+': float.__add__,
1068 '-': float.__sub__,
1069 }
1070 # Field is of the form key1.key2...
1071 # where keys (except first) can be string, int or slice
1072 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1073 MATH_FIELD_RE = r'''(?:{field}|{num})'''.format(field=FIELD_RE, num=r'-?\d+(?:.\d+)?')
1074 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1075 INTERNAL_FORMAT_RE = re.compile(r'''(?x)
1076 (?P<negate>-)?
1077 (?P<fields>{field})
1078 (?P<maths>(?:{math_op}{math_field})*)
1079 (?:>(?P<strf_format>.+?))?
1080 (?P<remaining>
1081 (?P<alternate>(?<!\\),[^|&)]+)?
1082 (?:&(?P<replacement>.*?))?
1083 (?:\|(?P<default>.*?))?
1084 )$'''.format(field=FIELD_RE, math_op=MATH_OPERATORS_RE, math_field=MATH_FIELD_RE))
1085
1086 def _traverse_infodict(k):
1087 k = k.split('.')
1088 if k[0] == '':
1089 k.pop(0)
1090 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1091
1092 def get_value(mdict):
1093 # Object traversal
1094 value = _traverse_infodict(mdict['fields'])
1095 # Negative
1096 if mdict['negate']:
1097 value = float_or_none(value)
1098 if value is not None:
1099 value *= -1
1100 # Do maths
1101 offset_key = mdict['maths']
1102 if offset_key:
1103 value = float_or_none(value)
1104 operator = None
1105 while offset_key:
1106 item = re.match(
1107 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1108 offset_key).group(0)
1109 offset_key = offset_key[len(item):]
1110 if operator is None:
1111 operator = MATH_FUNCTIONS[item]
1112 continue
1113 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1114 offset = float_or_none(item)
1115 if offset is None:
1116 offset = float_or_none(_traverse_infodict(item))
1117 try:
1118 value = operator(value, multiplier * offset)
1119 except (TypeError, ZeroDivisionError):
1120 return None
1121 operator = None
1122 # Datetime formatting
1123 if mdict['strf_format']:
1124 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1125
1126 return value
1127
1128 na = self.params.get('outtmpl_na_placeholder', 'NA')
1129
1130 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1131 return sanitize_filename(str(value), restricted=restricted, is_id=(
1132 bool(re.search(r'(^|[_.])id(\.|$)', key))
1133 if 'filename-sanitization' in self.params.get('compat_opts', [])
1134 else NO_DEFAULT))
1135
1136 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1137 sanitize = bool(sanitize)
1138
1139 def _dumpjson_default(obj):
1140 if isinstance(obj, (set, LazyList)):
1141 return list(obj)
1142 return repr(obj)
1143
1144 def create_key(outer_mobj):
1145 if not outer_mobj.group('has_key'):
1146 return outer_mobj.group(0)
1147 key = outer_mobj.group('key')
1148 mobj = re.match(INTERNAL_FORMAT_RE, key)
1149 initial_field = mobj.group('fields') if mobj else ''
1150 value, replacement, default = None, None, na
1151 while mobj:
1152 mobj = mobj.groupdict()
1153 default = mobj['default'] if mobj['default'] is not None else default
1154 value = get_value(mobj)
1155 replacement = mobj['replacement']
1156 if value is None and mobj['alternate']:
1157 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1158 else:
1159 break
1160
1161 fmt = outer_mobj.group('format')
1162 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1163 fmt = f'0{field_size_compat_map[key]:d}d'
1164
1165 value = default if value is None else value if replacement is None else replacement
1166
1167 flags = outer_mobj.group('conversion') or ''
1168 str_fmt = f'{fmt[:-1]}s'
1169 if fmt[-1] == 'l': # list
1170 delim = '\n' if '#' in flags else ', '
1171 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1172 elif fmt[-1] == 'j': # json
1173 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1174 elif fmt[-1] == 'q': # quoted
1175 value = map(str, variadic(value) if '#' in flags else [value])
1176 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1177 elif fmt[-1] == 'B': # bytes
1178 value = f'%{str_fmt}'.encode() % str(value).encode('utf-8')
1179 value, fmt = value.decode('utf-8', 'ignore'), 's'
1180 elif fmt[-1] == 'U': # unicode normalized
1181 value, fmt = unicodedata.normalize(
1182 # "+" = compatibility equivalence, "#" = NFD
1183 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1184 value), str_fmt
1185 elif fmt[-1] == 'D': # decimal suffix
1186 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1187 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1188 factor=1024 if '#' in flags else 1000)
1189 elif fmt[-1] == 'S': # filename sanitization
1190 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1191 elif fmt[-1] == 'c':
1192 if value:
1193 value = str(value)[0]
1194 else:
1195 fmt = str_fmt
1196 elif fmt[-1] not in 'rs': # numeric
1197 value = float_or_none(value)
1198 if value is None:
1199 value, fmt = default, 's'
1200
1201 if sanitize:
1202 if fmt[-1] == 'r':
1203 # If value is an object, sanitize might convert it to a string
1204 # So we convert it to repr first
1205 value, fmt = repr(value), str_fmt
1206 if fmt[-1] in 'csr':
1207 value = sanitizer(initial_field, value)
1208
1209 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1210 TMPL_DICT[key] = value
1211 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1212
1213 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1214
1215 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1216 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1217 return self.escape_outtmpl(outtmpl) % info_dict
1218
1219 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1220 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1221 if outtmpl is None:
1222 outtmpl = self.outtmpl_dict.get(tmpl_type or 'default', self.outtmpl_dict['default'])
1223 try:
1224 outtmpl = self._outtmpl_expandpath(outtmpl)
1225 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1226 if not filename:
1227 return None
1228
1229 if tmpl_type in ('', 'temp'):
1230 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1231 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1232 filename = replace_extension(filename, ext, final_ext)
1233 elif tmpl_type:
1234 force_ext = OUTTMPL_TYPES[tmpl_type]
1235 if force_ext:
1236 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1237
1238 # https://github.com/blackjack4494/youtube-dlc/issues/85
1239 trim_file_name = self.params.get('trim_file_name', False)
1240 if trim_file_name:
1241 no_ext, *ext = filename.rsplit('.', 2)
1242 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1243
1244 return filename
1245 except ValueError as err:
1246 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1247 return None
1248
1249 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1250 """Generate the output filename"""
1251 if outtmpl:
1252 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1253 dir_type = None
1254 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1255 if not filename and dir_type not in ('', 'temp'):
1256 return ''
1257
1258 if warn:
1259 if not self.params.get('paths'):
1260 pass
1261 elif filename == '-':
1262 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1263 elif os.path.isabs(filename):
1264 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1265 if filename == '-' or not filename:
1266 return filename
1267
1268 return self.get_output_path(dir_type, filename)
1269
1270 def _match_entry(self, info_dict, incomplete=False, silent=False):
1271 """ Returns None if the file should be downloaded """
1272
1273 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1274
1275 def check_filter():
1276 if 'title' in info_dict:
1277 # This can happen when we're just evaluating the playlist
1278 title = info_dict['title']
1279 matchtitle = self.params.get('matchtitle', False)
1280 if matchtitle:
1281 if not re.search(matchtitle, title, re.IGNORECASE):
1282 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1283 rejecttitle = self.params.get('rejecttitle', False)
1284 if rejecttitle:
1285 if re.search(rejecttitle, title, re.IGNORECASE):
1286 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1287 date = info_dict.get('upload_date')
1288 if date is not None:
1289 dateRange = self.params.get('daterange', DateRange())
1290 if date not in dateRange:
1291 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1292 view_count = info_dict.get('view_count')
1293 if view_count is not None:
1294 min_views = self.params.get('min_views')
1295 if min_views is not None and view_count < min_views:
1296 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1297 max_views = self.params.get('max_views')
1298 if max_views is not None and view_count > max_views:
1299 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1300 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1301 return 'Skipping "%s" because it is age restricted' % video_title
1302
1303 match_filter = self.params.get('match_filter')
1304 if match_filter is not None:
1305 try:
1306 ret = match_filter(info_dict, incomplete=incomplete)
1307 except TypeError:
1308 # For backward compatibility
1309 ret = None if incomplete else match_filter(info_dict)
1310 if ret is not None:
1311 return ret
1312 return None
1313
1314 if self.in_download_archive(info_dict):
1315 reason = '%s has already been recorded in the archive' % video_title
1316 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1317 else:
1318 reason = check_filter()
1319 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1320 if reason is not None:
1321 if not silent:
1322 self.to_screen('[download] ' + reason)
1323 if self.params.get(break_opt, False):
1324 raise break_err()
1325 return reason
1326
1327 @staticmethod
1328 def add_extra_info(info_dict, extra_info):
1329 '''Set the keys from extra_info in info dict if they are missing'''
1330 for key, value in extra_info.items():
1331 info_dict.setdefault(key, value)
1332
1333 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1334 process=True, force_generic_extractor=False):
1335 """
1336 Return a list with a dictionary for each video extracted.
1337
1338 Arguments:
1339 url -- URL to extract
1340
1341 Keyword arguments:
1342 download -- whether to download videos during extraction
1343 ie_key -- extractor key hint
1344 extra_info -- dictionary containing the extra values to add to each result
1345 process -- whether to resolve all unresolved references (URLs, playlist items),
1346 must be True for download to work.
1347 force_generic_extractor -- force using the generic extractor
1348 """
1349
1350 if extra_info is None:
1351 extra_info = {}
1352
1353 if not ie_key and force_generic_extractor:
1354 ie_key = 'Generic'
1355
1356 if ie_key:
1357 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1358 else:
1359 ies = self._ies
1360
1361 for ie_key, ie in ies.items():
1362 if not ie.suitable(url):
1363 continue
1364
1365 if not ie.working():
1366 self.report_warning('The program functionality for this site has been marked as broken, '
1367 'and will probably not work.')
1368
1369 temp_id = ie.get_temp_id(url)
1370 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1371 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1372 if self.params.get('break_on_existing', False):
1373 raise ExistingVideoReached()
1374 break
1375 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1376 else:
1377 self.report_error('no suitable InfoExtractor for URL %s' % url)
1378
1379 def __handle_extraction_exceptions(func):
1380 @functools.wraps(func)
1381 def wrapper(self, *args, **kwargs):
1382 while True:
1383 try:
1384 return func(self, *args, **kwargs)
1385 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1386 raise
1387 except ReExtractInfo as e:
1388 if e.expected:
1389 self.to_screen(f'{e}; Re-extracting data')
1390 else:
1391 self.to_stderr('\r')
1392 self.report_warning(f'{e}; Re-extracting data')
1393 continue
1394 except GeoRestrictedError as e:
1395 msg = e.msg
1396 if e.countries:
1397 msg += '\nThis video is available in %s.' % ', '.join(
1398 map(ISO3166Utils.short2full, e.countries))
1399 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1400 self.report_error(msg)
1401 except ExtractorError as e: # An error we somewhat expected
1402 self.report_error(str(e), e.format_traceback())
1403 except Exception as e:
1404 if self.params.get('ignoreerrors'):
1405 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1406 else:
1407 raise
1408 break
1409 return wrapper
1410
1411 def _wait_for_video(self, ie_result):
1412 if (not self.params.get('wait_for_video')
1413 or ie_result.get('_type', 'video') != 'video'
1414 or ie_result.get('formats') or ie_result.get('url')):
1415 return
1416
1417 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1418 last_msg = ''
1419
1420 def progress(msg):
1421 nonlocal last_msg
1422 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1423 last_msg = msg
1424
1425 min_wait, max_wait = self.params.get('wait_for_video')
1426 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1427 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1428 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1429 self.report_warning('Release time of video is not known')
1430 elif (diff or 0) <= 0:
1431 self.report_warning('Video should already be available according to extracted info')
1432 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1433 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1434
1435 wait_till = time.time() + diff
1436 try:
1437 while True:
1438 diff = wait_till - time.time()
1439 if diff <= 0:
1440 progress('')
1441 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1442 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1443 time.sleep(1)
1444 except KeyboardInterrupt:
1445 progress('')
1446 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1447 except BaseException as e:
1448 if not isinstance(e, ReExtractInfo):
1449 self.to_screen('')
1450 raise
1451
1452 @__handle_extraction_exceptions
1453 def __extract_info(self, url, ie, download, extra_info, process):
1454 ie_result = ie.extract(url)
1455 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1456 return
1457 if isinstance(ie_result, list):
1458 # Backwards compatibility: old IE result format
1459 ie_result = {
1460 '_type': 'compat_list',
1461 'entries': ie_result,
1462 }
1463 if extra_info.get('original_url'):
1464 ie_result.setdefault('original_url', extra_info['original_url'])
1465 self.add_default_extra_info(ie_result, ie, url)
1466 if process:
1467 self._wait_for_video(ie_result)
1468 return self.process_ie_result(ie_result, download, extra_info)
1469 else:
1470 return ie_result
1471
1472 def add_default_extra_info(self, ie_result, ie, url):
1473 if url is not None:
1474 self.add_extra_info(ie_result, {
1475 'webpage_url': url,
1476 'original_url': url,
1477 })
1478 webpage_url = ie_result.get('webpage_url')
1479 if webpage_url:
1480 self.add_extra_info(ie_result, {
1481 'webpage_url_basename': url_basename(webpage_url),
1482 'webpage_url_domain': get_domain(webpage_url),
1483 })
1484 if ie is not None:
1485 self.add_extra_info(ie_result, {
1486 'extractor': ie.IE_NAME,
1487 'extractor_key': ie.ie_key(),
1488 })
1489
1490 def process_ie_result(self, ie_result, download=True, extra_info=None):
1491 """
1492 Take the result of the ie(may be modified) and resolve all unresolved
1493 references (URLs, playlist items).
1494
1495 It will also download the videos if 'download'.
1496 Returns the resolved ie_result.
1497 """
1498 if extra_info is None:
1499 extra_info = {}
1500 result_type = ie_result.get('_type', 'video')
1501
1502 if result_type in ('url', 'url_transparent'):
1503 ie_result['url'] = sanitize_url(ie_result['url'])
1504 if ie_result.get('original_url'):
1505 extra_info.setdefault('original_url', ie_result['original_url'])
1506
1507 extract_flat = self.params.get('extract_flat', False)
1508 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1509 or extract_flat is True):
1510 info_copy = ie_result.copy()
1511 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1512 if ie and not ie_result.get('id'):
1513 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1514 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1515 self.add_extra_info(info_copy, extra_info)
1516 info_copy, _ = self.pre_process(info_copy)
1517 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1518 if self.params.get('force_write_download_archive', False):
1519 self.record_download_archive(info_copy)
1520 return ie_result
1521
1522 if result_type == 'video':
1523 self.add_extra_info(ie_result, extra_info)
1524 ie_result = self.process_video_result(ie_result, download=download)
1525 additional_urls = (ie_result or {}).get('additional_urls')
1526 if additional_urls:
1527 # TODO: Improve MetadataParserPP to allow setting a list
1528 if isinstance(additional_urls, compat_str):
1529 additional_urls = [additional_urls]
1530 self.to_screen(
1531 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1532 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1533 ie_result['additional_entries'] = [
1534 self.extract_info(
1535 url, download, extra_info=extra_info,
1536 force_generic_extractor=self.params.get('force_generic_extractor'))
1537 for url in additional_urls
1538 ]
1539 return ie_result
1540 elif result_type == 'url':
1541 # We have to add extra_info to the results because it may be
1542 # contained in a playlist
1543 return self.extract_info(
1544 ie_result['url'], download,
1545 ie_key=ie_result.get('ie_key'),
1546 extra_info=extra_info)
1547 elif result_type == 'url_transparent':
1548 # Use the information from the embedding page
1549 info = self.extract_info(
1550 ie_result['url'], ie_key=ie_result.get('ie_key'),
1551 extra_info=extra_info, download=False, process=False)
1552
1553 # extract_info may return None when ignoreerrors is enabled and
1554 # extraction failed with an error, don't crash and return early
1555 # in this case
1556 if not info:
1557 return info
1558
1559 new_result = info.copy()
1560 new_result.update(filter_dict(ie_result, lambda k, v: (
1561 v is not None and k not in {'_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'})))
1562
1563 # Extracted info may not be a video result (i.e.
1564 # info.get('_type', 'video') != video) but rather an url or
1565 # url_transparent. In such cases outer metadata (from ie_result)
1566 # should be propagated to inner one (info). For this to happen
1567 # _type of info should be overridden with url_transparent. This
1568 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1569 if new_result.get('_type') == 'url':
1570 new_result['_type'] = 'url_transparent'
1571
1572 return self.process_ie_result(
1573 new_result, download=download, extra_info=extra_info)
1574 elif result_type in ('playlist', 'multi_video'):
1575 # Protect from infinite recursion due to recursively nested playlists
1576 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1577 webpage_url = ie_result['webpage_url']
1578 if webpage_url in self._playlist_urls:
1579 self.to_screen(
1580 '[download] Skipping already downloaded playlist: %s'
1581 % ie_result.get('title') or ie_result.get('id'))
1582 return
1583
1584 self._playlist_level += 1
1585 self._playlist_urls.add(webpage_url)
1586 self._fill_common_fields(ie_result, False)
1587 self._sanitize_thumbnails(ie_result)
1588 try:
1589 return self.__process_playlist(ie_result, download)
1590 finally:
1591 self._playlist_level -= 1
1592 if not self._playlist_level:
1593 self._playlist_urls.clear()
1594 elif result_type == 'compat_list':
1595 self.report_warning(
1596 'Extractor %s returned a compat_list result. '
1597 'It needs to be updated.' % ie_result.get('extractor'))
1598
1599 def _fixup(r):
1600 self.add_extra_info(r, {
1601 'extractor': ie_result['extractor'],
1602 'webpage_url': ie_result['webpage_url'],
1603 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1604 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1605 'extractor_key': ie_result['extractor_key'],
1606 })
1607 return r
1608 ie_result['entries'] = [
1609 self.process_ie_result(_fixup(r), download, extra_info)
1610 for r in ie_result['entries']
1611 ]
1612 return ie_result
1613 else:
1614 raise Exception('Invalid result type: %s' % result_type)
1615
1616 def _ensure_dir_exists(self, path):
1617 return make_dir(path, self.report_error)
1618
1619 @staticmethod
1620 def _playlist_infodict(ie_result, **kwargs):
1621 return {
1622 **ie_result,
1623 'playlist': ie_result.get('title') or ie_result.get('id'),
1624 'playlist_id': ie_result.get('id'),
1625 'playlist_title': ie_result.get('title'),
1626 'playlist_uploader': ie_result.get('uploader'),
1627 'playlist_uploader_id': ie_result.get('uploader_id'),
1628 'playlist_index': 0,
1629 **kwargs,
1630 }
1631
1632 def __process_playlist(self, ie_result, download):
1633 # We process each entry in the playlist
1634 playlist = ie_result.get('title') or ie_result.get('id')
1635 self.to_screen('[download] Downloading playlist: %s' % playlist)
1636
1637 if 'entries' not in ie_result:
1638 raise EntryNotInPlaylist('There are no entries')
1639
1640 MissingEntry = object()
1641 incomplete_entries = bool(ie_result.get('requested_entries'))
1642 if incomplete_entries:
1643 def fill_missing_entries(entries, indices):
1644 ret = [MissingEntry] * max(indices)
1645 for i, entry in zip(indices, entries):
1646 ret[i - 1] = entry
1647 return ret
1648 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1649
1650 playlist_results = []
1651
1652 playliststart = self.params.get('playliststart', 1)
1653 playlistend = self.params.get('playlistend')
1654 # For backwards compatibility, interpret -1 as whole list
1655 if playlistend == -1:
1656 playlistend = None
1657
1658 playlistitems_str = self.params.get('playlist_items')
1659 playlistitems = None
1660 if playlistitems_str is not None:
1661 def iter_playlistitems(format):
1662 for string_segment in format.split(','):
1663 if '-' in string_segment:
1664 start, end = string_segment.split('-')
1665 for item in range(int(start), int(end) + 1):
1666 yield int(item)
1667 else:
1668 yield int(string_segment)
1669 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1670
1671 ie_entries = ie_result['entries']
1672 if isinstance(ie_entries, list):
1673 playlist_count = len(ie_entries)
1674 msg = f'Collected {playlist_count} videos; downloading %d of them'
1675 ie_result['playlist_count'] = ie_result.get('playlist_count') or playlist_count
1676
1677 def get_entry(i):
1678 return ie_entries[i - 1]
1679 else:
1680 msg = 'Downloading %d videos'
1681 if not isinstance(ie_entries, (PagedList, LazyList)):
1682 ie_entries = LazyList(ie_entries)
1683 elif isinstance(ie_entries, InAdvancePagedList):
1684 if ie_entries._pagesize == 1:
1685 playlist_count = ie_entries._pagecount
1686
1687 def get_entry(i):
1688 return YoutubeDL.__handle_extraction_exceptions(
1689 lambda self, i: ie_entries[i - 1]
1690 )(self, i)
1691
1692 entries, broken = [], False
1693 items = playlistitems if playlistitems is not None else itertools.count(playliststart)
1694 for i in items:
1695 if i == 0:
1696 continue
1697 if playlistitems is None and playlistend is not None and playlistend < i:
1698 break
1699 entry = None
1700 try:
1701 entry = get_entry(i)
1702 if entry is MissingEntry:
1703 raise EntryNotInPlaylist()
1704 except (IndexError, EntryNotInPlaylist):
1705 if incomplete_entries:
1706 raise EntryNotInPlaylist(f'Entry {i} cannot be found')
1707 elif not playlistitems:
1708 break
1709 entries.append(entry)
1710 try:
1711 if entry is not None:
1712 # TODO: Add auto-generated fields
1713 self._match_entry(entry, incomplete=True, silent=True)
1714 except (ExistingVideoReached, RejectedVideoReached):
1715 broken = True
1716 break
1717 ie_result['entries'] = entries
1718
1719 # Save playlist_index before re-ordering
1720 entries = [
1721 ((playlistitems[i - 1] if playlistitems else i + playliststart - 1), entry)
1722 for i, entry in enumerate(entries, 1)
1723 if entry is not None]
1724 n_entries = len(entries)
1725
1726 if not (ie_result.get('playlist_count') or broken or playlistitems or playlistend):
1727 ie_result['playlist_count'] = n_entries
1728
1729 if not playlistitems and (playliststart != 1 or playlistend):
1730 playlistitems = list(range(playliststart, playliststart + n_entries))
1731 ie_result['requested_entries'] = playlistitems
1732
1733 _infojson_written = False
1734 write_playlist_files = self.params.get('allow_playlist_files', True)
1735 if write_playlist_files and self.params.get('list_thumbnails'):
1736 self.list_thumbnails(ie_result)
1737 if write_playlist_files and not self.params.get('simulate'):
1738 ie_copy = self._playlist_infodict(ie_result, n_entries=n_entries)
1739 _infojson_written = self._write_info_json(
1740 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1741 if _infojson_written is None:
1742 return
1743 if self._write_description('playlist', ie_result,
1744 self.prepare_filename(ie_copy, 'pl_description')) is None:
1745 return
1746 # TODO: This should be passed to ThumbnailsConvertor if necessary
1747 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1748
1749 if self.params.get('playlistreverse', False):
1750 entries = entries[::-1]
1751 if self.params.get('playlistrandom', False):
1752 random.shuffle(entries)
1753
1754 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1755
1756 self.to_screen(f'[{ie_result["extractor"]}] playlist {playlist}: {msg % n_entries}')
1757 failures = 0
1758 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1759 for i, entry_tuple in enumerate(entries, 1):
1760 playlist_index, entry = entry_tuple
1761 if 'playlist-index' in self.params.get('compat_opts', []):
1762 playlist_index = playlistitems[i - 1] if playlistitems else i + playliststart - 1
1763 self.to_screen(f'[download] Downloading video {i} of {n_entries}')
1764 # This __x_forwarded_for_ip thing is a bit ugly but requires
1765 # minimal changes
1766 if x_forwarded_for:
1767 entry['__x_forwarded_for_ip'] = x_forwarded_for
1768 extra = {
1769 'n_entries': n_entries,
1770 '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1771 'playlist_count': ie_result.get('playlist_count'),
1772 'playlist_index': playlist_index,
1773 'playlist_autonumber': i,
1774 'playlist': playlist,
1775 'playlist_id': ie_result.get('id'),
1776 'playlist_title': ie_result.get('title'),
1777 'playlist_uploader': ie_result.get('uploader'),
1778 'playlist_uploader_id': ie_result.get('uploader_id'),
1779 'extractor': ie_result['extractor'],
1780 'webpage_url': ie_result['webpage_url'],
1781 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1782 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1783 'extractor_key': ie_result['extractor_key'],
1784 }
1785
1786 if self._match_entry(entry, incomplete=True) is not None:
1787 continue
1788
1789 entry_result = self.__process_iterable_entry(entry, download, extra)
1790 if not entry_result:
1791 failures += 1
1792 if failures >= max_failures:
1793 self.report_error(
1794 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1795 break
1796 playlist_results.append(entry_result)
1797 ie_result['entries'] = playlist_results
1798
1799 # Write the updated info to json
1800 if _infojson_written is True and self._write_info_json(
1801 'updated playlist', ie_result,
1802 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1803 return
1804
1805 ie_result = self.run_all_pps('playlist', ie_result)
1806 self.to_screen(f'[download] Finished downloading playlist: {playlist}')
1807 return ie_result
1808
1809 @__handle_extraction_exceptions
1810 def __process_iterable_entry(self, entry, download, extra_info):
1811 return self.process_ie_result(
1812 entry, download=download, extra_info=extra_info)
1813
1814 def _build_format_filter(self, filter_spec):
1815 " Returns a function to filter the formats according to the filter_spec "
1816
1817 OPERATORS = {
1818 '<': operator.lt,
1819 '<=': operator.le,
1820 '>': operator.gt,
1821 '>=': operator.ge,
1822 '=': operator.eq,
1823 '!=': operator.ne,
1824 }
1825 operator_rex = re.compile(r'''(?x)\s*
1826 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1827 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1828 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1829 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1830 m = operator_rex.fullmatch(filter_spec)
1831 if m:
1832 try:
1833 comparison_value = int(m.group('value'))
1834 except ValueError:
1835 comparison_value = parse_filesize(m.group('value'))
1836 if comparison_value is None:
1837 comparison_value = parse_filesize(m.group('value') + 'B')
1838 if comparison_value is None:
1839 raise ValueError(
1840 'Invalid value %r in format specification %r' % (
1841 m.group('value'), filter_spec))
1842 op = OPERATORS[m.group('op')]
1843
1844 if not m:
1845 STR_OPERATORS = {
1846 '=': operator.eq,
1847 '^=': lambda attr, value: attr.startswith(value),
1848 '$=': lambda attr, value: attr.endswith(value),
1849 '*=': lambda attr, value: value in attr,
1850 '~=': lambda attr, value: value.search(attr) is not None
1851 }
1852 str_operator_rex = re.compile(r'''(?x)\s*
1853 (?P<key>[a-zA-Z0-9._-]+)\s*
1854 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1855 (?P<quote>["'])?
1856 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1857 (?(quote)(?P=quote))\s*
1858 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1859 m = str_operator_rex.fullmatch(filter_spec)
1860 if m:
1861 if m.group('op') == '~=':
1862 comparison_value = re.compile(m.group('value'))
1863 else:
1864 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1865 str_op = STR_OPERATORS[m.group('op')]
1866 if m.group('negation'):
1867 op = lambda attr, value: not str_op(attr, value)
1868 else:
1869 op = str_op
1870
1871 if not m:
1872 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1873
1874 def _filter(f):
1875 actual_value = f.get(m.group('key'))
1876 if actual_value is None:
1877 return m.group('none_inclusive')
1878 return op(actual_value, comparison_value)
1879 return _filter
1880
1881 def _check_formats(self, formats):
1882 for f in formats:
1883 self.to_screen('[info] Testing format %s' % f['format_id'])
1884 path = self.get_output_path('temp')
1885 if not self._ensure_dir_exists(f'{path}/'):
1886 continue
1887 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1888 temp_file.close()
1889 try:
1890 success, _ = self.dl(temp_file.name, f, test=True)
1891 except (DownloadError, IOError, OSError, ValueError) + network_exceptions:
1892 success = False
1893 finally:
1894 if os.path.exists(temp_file.name):
1895 try:
1896 os.remove(temp_file.name)
1897 except OSError:
1898 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1899 if success:
1900 yield f
1901 else:
1902 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1903
1904 def _default_format_spec(self, info_dict, download=True):
1905
1906 def can_merge():
1907 merger = FFmpegMergerPP(self)
1908 return merger.available and merger.can_merge()
1909
1910 prefer_best = (
1911 not self.params.get('simulate')
1912 and download
1913 and (
1914 not can_merge()
1915 or info_dict.get('is_live', False)
1916 or self.outtmpl_dict['default'] == '-'))
1917 compat = (
1918 prefer_best
1919 or self.params.get('allow_multiple_audio_streams', False)
1920 or 'format-spec' in self.params.get('compat_opts', []))
1921
1922 return (
1923 'best/bestvideo+bestaudio' if prefer_best
1924 else 'bestvideo*+bestaudio/best' if not compat
1925 else 'bestvideo+bestaudio/best')
1926
1927 def build_format_selector(self, format_spec):
1928 def syntax_error(note, start):
1929 message = (
1930 'Invalid format specification: '
1931 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1932 return SyntaxError(message)
1933
1934 PICKFIRST = 'PICKFIRST'
1935 MERGE = 'MERGE'
1936 SINGLE = 'SINGLE'
1937 GROUP = 'GROUP'
1938 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1939
1940 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1941 'video': self.params.get('allow_multiple_video_streams', False)}
1942
1943 check_formats = self.params.get('check_formats') == 'selected'
1944
1945 def _parse_filter(tokens):
1946 filter_parts = []
1947 for type, string, start, _, _ in tokens:
1948 if type == tokenize.OP and string == ']':
1949 return ''.join(filter_parts)
1950 else:
1951 filter_parts.append(string)
1952
1953 def _remove_unused_ops(tokens):
1954 # Remove operators that we don't use and join them with the surrounding strings
1955 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1956 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1957 last_string, last_start, last_end, last_line = None, None, None, None
1958 for type, string, start, end, line in tokens:
1959 if type == tokenize.OP and string == '[':
1960 if last_string:
1961 yield tokenize.NAME, last_string, last_start, last_end, last_line
1962 last_string = None
1963 yield type, string, start, end, line
1964 # everything inside brackets will be handled by _parse_filter
1965 for type, string, start, end, line in tokens:
1966 yield type, string, start, end, line
1967 if type == tokenize.OP and string == ']':
1968 break
1969 elif type == tokenize.OP and string in ALLOWED_OPS:
1970 if last_string:
1971 yield tokenize.NAME, last_string, last_start, last_end, last_line
1972 last_string = None
1973 yield type, string, start, end, line
1974 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1975 if not last_string:
1976 last_string = string
1977 last_start = start
1978 last_end = end
1979 else:
1980 last_string += string
1981 if last_string:
1982 yield tokenize.NAME, last_string, last_start, last_end, last_line
1983
1984 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1985 selectors = []
1986 current_selector = None
1987 for type, string, start, _, _ in tokens:
1988 # ENCODING is only defined in python 3.x
1989 if type == getattr(tokenize, 'ENCODING', None):
1990 continue
1991 elif type in [tokenize.NAME, tokenize.NUMBER]:
1992 current_selector = FormatSelector(SINGLE, string, [])
1993 elif type == tokenize.OP:
1994 if string == ')':
1995 if not inside_group:
1996 # ')' will be handled by the parentheses group
1997 tokens.restore_last_token()
1998 break
1999 elif inside_merge and string in ['/', ',']:
2000 tokens.restore_last_token()
2001 break
2002 elif inside_choice and string == ',':
2003 tokens.restore_last_token()
2004 break
2005 elif string == ',':
2006 if not current_selector:
2007 raise syntax_error('"," must follow a format selector', start)
2008 selectors.append(current_selector)
2009 current_selector = None
2010 elif string == '/':
2011 if not current_selector:
2012 raise syntax_error('"/" must follow a format selector', start)
2013 first_choice = current_selector
2014 second_choice = _parse_format_selection(tokens, inside_choice=True)
2015 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2016 elif string == '[':
2017 if not current_selector:
2018 current_selector = FormatSelector(SINGLE, 'best', [])
2019 format_filter = _parse_filter(tokens)
2020 current_selector.filters.append(format_filter)
2021 elif string == '(':
2022 if current_selector:
2023 raise syntax_error('Unexpected "("', start)
2024 group = _parse_format_selection(tokens, inside_group=True)
2025 current_selector = FormatSelector(GROUP, group, [])
2026 elif string == '+':
2027 if not current_selector:
2028 raise syntax_error('Unexpected "+"', start)
2029 selector_1 = current_selector
2030 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2031 if not selector_2:
2032 raise syntax_error('Expected a selector', start)
2033 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2034 else:
2035 raise syntax_error(f'Operator not recognized: "{string}"', start)
2036 elif type == tokenize.ENDMARKER:
2037 break
2038 if current_selector:
2039 selectors.append(current_selector)
2040 return selectors
2041
2042 def _merge(formats_pair):
2043 format_1, format_2 = formats_pair
2044
2045 formats_info = []
2046 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2047 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2048
2049 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2050 get_no_more = {'video': False, 'audio': False}
2051 for (i, fmt_info) in enumerate(formats_info):
2052 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2053 formats_info.pop(i)
2054 continue
2055 for aud_vid in ['audio', 'video']:
2056 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2057 if get_no_more[aud_vid]:
2058 formats_info.pop(i)
2059 break
2060 get_no_more[aud_vid] = True
2061
2062 if len(formats_info) == 1:
2063 return formats_info[0]
2064
2065 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2066 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2067
2068 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2069 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2070
2071 output_ext = self.params.get('merge_output_format')
2072 if not output_ext:
2073 if the_only_video:
2074 output_ext = the_only_video['ext']
2075 elif the_only_audio and not video_fmts:
2076 output_ext = the_only_audio['ext']
2077 else:
2078 output_ext = 'mkv'
2079
2080 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2081
2082 new_dict = {
2083 'requested_formats': formats_info,
2084 'format': '+'.join(filtered('format')),
2085 'format_id': '+'.join(filtered('format_id')),
2086 'ext': output_ext,
2087 'protocol': '+'.join(map(determine_protocol, formats_info)),
2088 'language': '+'.join(orderedSet(filtered('language'))) or None,
2089 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2090 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2091 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2092 }
2093
2094 if the_only_video:
2095 new_dict.update({
2096 'width': the_only_video.get('width'),
2097 'height': the_only_video.get('height'),
2098 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2099 'fps': the_only_video.get('fps'),
2100 'dynamic_range': the_only_video.get('dynamic_range'),
2101 'vcodec': the_only_video.get('vcodec'),
2102 'vbr': the_only_video.get('vbr'),
2103 'stretched_ratio': the_only_video.get('stretched_ratio'),
2104 })
2105
2106 if the_only_audio:
2107 new_dict.update({
2108 'acodec': the_only_audio.get('acodec'),
2109 'abr': the_only_audio.get('abr'),
2110 'asr': the_only_audio.get('asr'),
2111 })
2112
2113 return new_dict
2114
2115 def _check_formats(formats):
2116 if not check_formats:
2117 yield from formats
2118 return
2119 yield from self._check_formats(formats)
2120
2121 def _build_selector_function(selector):
2122 if isinstance(selector, list): # ,
2123 fs = [_build_selector_function(s) for s in selector]
2124
2125 def selector_function(ctx):
2126 for f in fs:
2127 yield from f(ctx)
2128 return selector_function
2129
2130 elif selector.type == GROUP: # ()
2131 selector_function = _build_selector_function(selector.selector)
2132
2133 elif selector.type == PICKFIRST: # /
2134 fs = [_build_selector_function(s) for s in selector.selector]
2135
2136 def selector_function(ctx):
2137 for f in fs:
2138 picked_formats = list(f(ctx))
2139 if picked_formats:
2140 return picked_formats
2141 return []
2142
2143 elif selector.type == MERGE: # +
2144 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2145
2146 def selector_function(ctx):
2147 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2148 yield _merge(pair)
2149
2150 elif selector.type == SINGLE: # atom
2151 format_spec = selector.selector or 'best'
2152
2153 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2154 if format_spec == 'all':
2155 def selector_function(ctx):
2156 yield from _check_formats(ctx['formats'][::-1])
2157 elif format_spec == 'mergeall':
2158 def selector_function(ctx):
2159 formats = list(_check_formats(
2160 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2161 if not formats:
2162 return
2163 merged_format = formats[-1]
2164 for f in formats[-2::-1]:
2165 merged_format = _merge((merged_format, f))
2166 yield merged_format
2167
2168 else:
2169 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2170 mobj = re.match(
2171 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2172 format_spec)
2173 if mobj is not None:
2174 format_idx = int_or_none(mobj.group('n'), default=1)
2175 format_reverse = mobj.group('bw')[0] == 'b'
2176 format_type = (mobj.group('type') or [None])[0]
2177 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2178 format_modified = mobj.group('mod') is not None
2179
2180 format_fallback = not format_type and not format_modified # for b, w
2181 _filter_f = (
2182 (lambda f: f.get('%scodec' % format_type) != 'none')
2183 if format_type and format_modified # bv*, ba*, wv*, wa*
2184 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2185 if format_type # bv, ba, wv, wa
2186 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2187 if not format_modified # b, w
2188 else lambda f: True) # b*, w*
2189 filter_f = lambda f: _filter_f(f) and (
2190 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2191 else:
2192 if format_spec in self._format_selection_exts['audio']:
2193 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2194 elif format_spec in self._format_selection_exts['video']:
2195 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2196 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2197 elif format_spec in self._format_selection_exts['storyboards']:
2198 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2199 else:
2200 filter_f = lambda f: f.get('format_id') == format_spec # id
2201
2202 def selector_function(ctx):
2203 formats = list(ctx['formats'])
2204 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2205 if not matches:
2206 if format_fallback and ctx['incomplete_formats']:
2207 # for extractors with incomplete formats (audio only (soundcloud)
2208 # or video only (imgur)) best/worst will fallback to
2209 # best/worst {video,audio}-only format
2210 matches = formats
2211 elif seperate_fallback and not ctx['has_merged_format']:
2212 # for compatibility with youtube-dl when there is no pre-merged format
2213 matches = list(filter(seperate_fallback, formats))
2214 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2215 try:
2216 yield matches[format_idx - 1]
2217 except LazyList.IndexError:
2218 return
2219
2220 filters = [self._build_format_filter(f) for f in selector.filters]
2221
2222 def final_selector(ctx):
2223 ctx_copy = dict(ctx)
2224 for _filter in filters:
2225 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2226 return selector_function(ctx_copy)
2227 return final_selector
2228
2229 stream = io.BytesIO(format_spec.encode('utf-8'))
2230 try:
2231 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2232 except tokenize.TokenError:
2233 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2234
2235 class TokenIterator:
2236 def __init__(self, tokens):
2237 self.tokens = tokens
2238 self.counter = 0
2239
2240 def __iter__(self):
2241 return self
2242
2243 def __next__(self):
2244 if self.counter >= len(self.tokens):
2245 raise StopIteration()
2246 value = self.tokens[self.counter]
2247 self.counter += 1
2248 return value
2249
2250 next = __next__
2251
2252 def restore_last_token(self):
2253 self.counter -= 1
2254
2255 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2256 return _build_selector_function(parsed_selector)
2257
2258 def _calc_headers(self, info_dict):
2259 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2260
2261 cookies = self._calc_cookies(info_dict)
2262 if cookies:
2263 res['Cookie'] = cookies
2264
2265 if 'X-Forwarded-For' not in res:
2266 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2267 if x_forwarded_for_ip:
2268 res['X-Forwarded-For'] = x_forwarded_for_ip
2269
2270 return res
2271
2272 def _calc_cookies(self, info_dict):
2273 pr = sanitized_Request(info_dict['url'])
2274 self.cookiejar.add_cookie_header(pr)
2275 return pr.get_header('Cookie')
2276
2277 def _sort_thumbnails(self, thumbnails):
2278 thumbnails.sort(key=lambda t: (
2279 t.get('preference') if t.get('preference') is not None else -1,
2280 t.get('width') if t.get('width') is not None else -1,
2281 t.get('height') if t.get('height') is not None else -1,
2282 t.get('id') if t.get('id') is not None else '',
2283 t.get('url')))
2284
2285 def _sanitize_thumbnails(self, info_dict):
2286 thumbnails = info_dict.get('thumbnails')
2287 if thumbnails is None:
2288 thumbnail = info_dict.get('thumbnail')
2289 if thumbnail:
2290 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2291 if not thumbnails:
2292 return
2293
2294 def check_thumbnails(thumbnails):
2295 for t in thumbnails:
2296 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2297 try:
2298 self.urlopen(HEADRequest(t['url']))
2299 except network_exceptions as err:
2300 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2301 continue
2302 yield t
2303
2304 self._sort_thumbnails(thumbnails)
2305 for i, t in enumerate(thumbnails):
2306 if t.get('id') is None:
2307 t['id'] = '%d' % i
2308 if t.get('width') and t.get('height'):
2309 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2310 t['url'] = sanitize_url(t['url'])
2311
2312 if self.params.get('check_formats') is True:
2313 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2314 else:
2315 info_dict['thumbnails'] = thumbnails
2316
2317 def _fill_common_fields(self, info_dict, is_video=True):
2318 # TODO: move sanitization here
2319 if is_video:
2320 # playlists are allowed to lack "title"
2321 info_dict['fulltitle'] = info_dict.get('title')
2322 if 'title' not in info_dict:
2323 raise ExtractorError('Missing "title" field in extractor result',
2324 video_id=info_dict['id'], ie=info_dict['extractor'])
2325 elif not info_dict.get('title'):
2326 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2327 info_dict['title'] = f'{info_dict["extractor"]} video #{info_dict["id"]}'
2328
2329 if info_dict.get('duration') is not None:
2330 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2331
2332 for ts_key, date_key in (
2333 ('timestamp', 'upload_date'),
2334 ('release_timestamp', 'release_date'),
2335 ('modified_timestamp', 'modified_date'),
2336 ):
2337 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2338 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2339 # see http://bugs.python.org/issue1646728)
2340 try:
2341 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2342 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2343 except (ValueError, OverflowError, OSError):
2344 pass
2345
2346 live_keys = ('is_live', 'was_live')
2347 live_status = info_dict.get('live_status')
2348 if live_status is None:
2349 for key in live_keys:
2350 if info_dict.get(key) is False:
2351 continue
2352 if info_dict.get(key):
2353 live_status = key
2354 break
2355 if all(info_dict.get(key) is False for key in live_keys):
2356 live_status = 'not_live'
2357 if live_status:
2358 info_dict['live_status'] = live_status
2359 for key in live_keys:
2360 if info_dict.get(key) is None:
2361 info_dict[key] = (live_status == key)
2362
2363 # Auto generate title fields corresponding to the *_number fields when missing
2364 # in order to always have clean titles. This is very common for TV series.
2365 for field in ('chapter', 'season', 'episode'):
2366 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2367 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2368
2369 def process_video_result(self, info_dict, download=True):
2370 assert info_dict.get('_type', 'video') == 'video'
2371 self._num_videos += 1
2372
2373 if 'id' not in info_dict:
2374 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2375 elif not info_dict.get('id'):
2376 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2377
2378 def report_force_conversion(field, field_not, conversion):
2379 self.report_warning(
2380 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2381 % (field, field_not, conversion))
2382
2383 def sanitize_string_field(info, string_field):
2384 field = info.get(string_field)
2385 if field is None or isinstance(field, compat_str):
2386 return
2387 report_force_conversion(string_field, 'a string', 'string')
2388 info[string_field] = compat_str(field)
2389
2390 def sanitize_numeric_fields(info):
2391 for numeric_field in self._NUMERIC_FIELDS:
2392 field = info.get(numeric_field)
2393 if field is None or isinstance(field, (int, float)):
2394 continue
2395 report_force_conversion(numeric_field, 'numeric', 'int')
2396 info[numeric_field] = int_or_none(field)
2397
2398 sanitize_string_field(info_dict, 'id')
2399 sanitize_numeric_fields(info_dict)
2400 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2401 self.report_warning('"duration" field is negative, there is an error in extractor')
2402
2403 if 'playlist' not in info_dict:
2404 # It isn't part of a playlist
2405 info_dict['playlist'] = None
2406 info_dict['playlist_index'] = None
2407
2408 self._sanitize_thumbnails(info_dict)
2409
2410 thumbnail = info_dict.get('thumbnail')
2411 thumbnails = info_dict.get('thumbnails')
2412 if thumbnail:
2413 info_dict['thumbnail'] = sanitize_url(thumbnail)
2414 elif thumbnails:
2415 info_dict['thumbnail'] = thumbnails[-1]['url']
2416
2417 if info_dict.get('display_id') is None and 'id' in info_dict:
2418 info_dict['display_id'] = info_dict['id']
2419
2420 self._fill_common_fields(info_dict)
2421
2422 for cc_kind in ('subtitles', 'automatic_captions'):
2423 cc = info_dict.get(cc_kind)
2424 if cc:
2425 for _, subtitle in cc.items():
2426 for subtitle_format in subtitle:
2427 if subtitle_format.get('url'):
2428 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2429 if subtitle_format.get('ext') is None:
2430 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2431
2432 automatic_captions = info_dict.get('automatic_captions')
2433 subtitles = info_dict.get('subtitles')
2434
2435 info_dict['requested_subtitles'] = self.process_subtitles(
2436 info_dict['id'], subtitles, automatic_captions)
2437
2438 if info_dict.get('formats') is None:
2439 # There's only one format available
2440 formats = [info_dict]
2441 else:
2442 formats = info_dict['formats']
2443
2444 info_dict['__has_drm'] = any(f.get('has_drm') for f in formats)
2445 if not self.params.get('allow_unplayable_formats'):
2446 formats = [f for f in formats if not f.get('has_drm')]
2447 if info_dict['__has_drm'] and all(
2448 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2449 self.report_warning(
2450 'This video is DRM protected and only images are available for download. '
2451 'Use --list-formats to see them')
2452
2453 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2454 if not get_from_start:
2455 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2456 if info_dict.get('is_live') and formats:
2457 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2458 if get_from_start and not formats:
2459 self.raise_no_formats(info_dict, msg=(
2460 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2461 'If you want to download from the current time, use --no-live-from-start'))
2462
2463 if not formats:
2464 self.raise_no_formats(info_dict)
2465
2466 def is_wellformed(f):
2467 url = f.get('url')
2468 if not url:
2469 self.report_warning(
2470 '"url" field is missing or empty - skipping format, '
2471 'there is an error in extractor')
2472 return False
2473 if isinstance(url, bytes):
2474 sanitize_string_field(f, 'url')
2475 return True
2476
2477 # Filter out malformed formats for better extraction robustness
2478 formats = list(filter(is_wellformed, formats))
2479
2480 formats_dict = {}
2481
2482 # We check that all the formats have the format and format_id fields
2483 for i, format in enumerate(formats):
2484 sanitize_string_field(format, 'format_id')
2485 sanitize_numeric_fields(format)
2486 format['url'] = sanitize_url(format['url'])
2487 if not format.get('format_id'):
2488 format['format_id'] = compat_str(i)
2489 else:
2490 # Sanitize format_id from characters used in format selector expression
2491 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2492 format_id = format['format_id']
2493 if format_id not in formats_dict:
2494 formats_dict[format_id] = []
2495 formats_dict[format_id].append(format)
2496
2497 # Make sure all formats have unique format_id
2498 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2499 for format_id, ambiguous_formats in formats_dict.items():
2500 ambigious_id = len(ambiguous_formats) > 1
2501 for i, format in enumerate(ambiguous_formats):
2502 if ambigious_id:
2503 format['format_id'] = '%s-%d' % (format_id, i)
2504 if format.get('ext') is None:
2505 format['ext'] = determine_ext(format['url']).lower()
2506 # Ensure there is no conflict between id and ext in format selection
2507 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2508 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2509 format['format_id'] = 'f%s' % format['format_id']
2510
2511 for i, format in enumerate(formats):
2512 if format.get('format') is None:
2513 format['format'] = '{id} - {res}{note}'.format(
2514 id=format['format_id'],
2515 res=self.format_resolution(format),
2516 note=format_field(format, 'format_note', ' (%s)'),
2517 )
2518 if format.get('protocol') is None:
2519 format['protocol'] = determine_protocol(format)
2520 if format.get('resolution') is None:
2521 format['resolution'] = self.format_resolution(format, default=None)
2522 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2523 format['dynamic_range'] = 'SDR'
2524 if (info_dict.get('duration') and format.get('tbr')
2525 and not format.get('filesize') and not format.get('filesize_approx')):
2526 format['filesize_approx'] = info_dict['duration'] * format['tbr'] * (1024 / 8)
2527
2528 # Add HTTP headers, so that external programs can use them from the
2529 # json output
2530 full_format_info = info_dict.copy()
2531 full_format_info.update(format)
2532 format['http_headers'] = self._calc_headers(full_format_info)
2533 # Remove private housekeeping stuff
2534 if '__x_forwarded_for_ip' in info_dict:
2535 del info_dict['__x_forwarded_for_ip']
2536
2537 if self.params.get('check_formats') is True:
2538 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2539
2540 if not formats or formats[0] is not info_dict:
2541 # only set the 'formats' fields if the original info_dict list them
2542 # otherwise we end up with a circular reference, the first (and unique)
2543 # element in the 'formats' field in info_dict is info_dict itself,
2544 # which can't be exported to json
2545 info_dict['formats'] = formats
2546
2547 info_dict, _ = self.pre_process(info_dict)
2548
2549 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2550 return info_dict
2551
2552 self.post_extract(info_dict)
2553 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2554
2555 # The pre-processors may have modified the formats
2556 formats = info_dict.get('formats', [info_dict])
2557
2558 list_only = self.params.get('simulate') is None and (
2559 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2560 interactive_format_selection = not list_only and self.format_selector == '-'
2561 if self.params.get('list_thumbnails'):
2562 self.list_thumbnails(info_dict)
2563 if self.params.get('listsubtitles'):
2564 if 'automatic_captions' in info_dict:
2565 self.list_subtitles(
2566 info_dict['id'], automatic_captions, 'automatic captions')
2567 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2568 if self.params.get('listformats') or interactive_format_selection:
2569 self.list_formats(info_dict)
2570 if list_only:
2571 # Without this printing, -F --print-json will not work
2572 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2573 return
2574
2575 format_selector = self.format_selector
2576 if format_selector is None:
2577 req_format = self._default_format_spec(info_dict, download=download)
2578 self.write_debug('Default format spec: %s' % req_format)
2579 format_selector = self.build_format_selector(req_format)
2580
2581 while True:
2582 if interactive_format_selection:
2583 req_format = input(
2584 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2585 try:
2586 format_selector = self.build_format_selector(req_format)
2587 except SyntaxError as err:
2588 self.report_error(err, tb=False, is_error=False)
2589 continue
2590
2591 formats_to_download = list(format_selector({
2592 'formats': formats,
2593 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2594 'incomplete_formats': (
2595 # All formats are video-only or
2596 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2597 # all formats are audio-only
2598 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2599 }))
2600 if interactive_format_selection and not formats_to_download:
2601 self.report_error('Requested format is not available', tb=False, is_error=False)
2602 continue
2603 break
2604
2605 if not formats_to_download:
2606 if not self.params.get('ignore_no_formats_error'):
2607 raise ExtractorError(
2608 'Requested format is not available. Use --list-formats for a list of available formats',
2609 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2610 self.report_warning('Requested format is not available')
2611 # Process what we can, even without any available formats.
2612 formats_to_download = [{}]
2613
2614 best_format = formats_to_download[-1]
2615 if download:
2616 if best_format:
2617 self.to_screen(
2618 f'[info] {info_dict["id"]}: Downloading {len(formats_to_download)} format(s): '
2619 + ', '.join([f['format_id'] for f in formats_to_download]))
2620 max_downloads_reached = False
2621 for i, fmt in enumerate(formats_to_download):
2622 formats_to_download[i] = new_info = self._copy_infodict(info_dict)
2623 new_info.update(fmt)
2624 try:
2625 self.process_info(new_info)
2626 except MaxDownloadsReached:
2627 max_downloads_reached = True
2628 # Remove copied info
2629 for key, val in tuple(new_info.items()):
2630 if info_dict.get(key) == val:
2631 new_info.pop(key)
2632 if max_downloads_reached:
2633 break
2634
2635 write_archive = {f.get('__write_download_archive', False) for f in formats_to_download}
2636 assert write_archive.issubset({True, False, 'ignore'})
2637 if True in write_archive and False not in write_archive:
2638 self.record_download_archive(info_dict)
2639
2640 info_dict['requested_downloads'] = formats_to_download
2641 info_dict = self.run_all_pps('after_video', info_dict)
2642 if max_downloads_reached:
2643 raise MaxDownloadsReached()
2644
2645 # We update the info dict with the selected best quality format (backwards compatibility)
2646 info_dict.update(best_format)
2647 return info_dict
2648
2649 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2650 """Select the requested subtitles and their format"""
2651 available_subs, normal_sub_langs = {}, []
2652 if normal_subtitles and self.params.get('writesubtitles'):
2653 available_subs.update(normal_subtitles)
2654 normal_sub_langs = tuple(normal_subtitles.keys())
2655 if automatic_captions and self.params.get('writeautomaticsub'):
2656 for lang, cap_info in automatic_captions.items():
2657 if lang not in available_subs:
2658 available_subs[lang] = cap_info
2659
2660 if (not self.params.get('writesubtitles') and not
2661 self.params.get('writeautomaticsub') or not
2662 available_subs):
2663 return None
2664
2665 all_sub_langs = tuple(available_subs.keys())
2666 if self.params.get('allsubtitles', False):
2667 requested_langs = all_sub_langs
2668 elif self.params.get('subtitleslangs', False):
2669 # A list is used so that the order of languages will be the same as
2670 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2671 requested_langs = []
2672 for lang_re in self.params.get('subtitleslangs'):
2673 discard = lang_re[0] == '-'
2674 if discard:
2675 lang_re = lang_re[1:]
2676 if lang_re == 'all':
2677 if discard:
2678 requested_langs = []
2679 else:
2680 requested_langs.extend(all_sub_langs)
2681 continue
2682 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2683 if discard:
2684 for lang in current_langs:
2685 while lang in requested_langs:
2686 requested_langs.remove(lang)
2687 else:
2688 requested_langs.extend(current_langs)
2689 requested_langs = orderedSet(requested_langs)
2690 elif normal_sub_langs:
2691 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2692 else:
2693 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2694 if requested_langs:
2695 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2696
2697 formats_query = self.params.get('subtitlesformat', 'best')
2698 formats_preference = formats_query.split('/') if formats_query else []
2699 subs = {}
2700 for lang in requested_langs:
2701 formats = available_subs.get(lang)
2702 if formats is None:
2703 self.report_warning(f'{lang} subtitles not available for {video_id}')
2704 continue
2705 for ext in formats_preference:
2706 if ext == 'best':
2707 f = formats[-1]
2708 break
2709 matches = list(filter(lambda f: f['ext'] == ext, formats))
2710 if matches:
2711 f = matches[-1]
2712 break
2713 else:
2714 f = formats[-1]
2715 self.report_warning(
2716 'No subtitle format found matching "%s" for language %s, '
2717 'using %s' % (formats_query, lang, f['ext']))
2718 subs[lang] = f
2719 return subs
2720
2721 def _forceprint(self, key, info_dict):
2722 if info_dict is None:
2723 return
2724 info_copy = info_dict.copy()
2725 info_copy['formats_table'] = self.render_formats_table(info_dict)
2726 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2727 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2728 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2729
2730 def format_tmpl(tmpl):
2731 mobj = re.match(r'\w+(=?)$', tmpl)
2732 if mobj and mobj.group(1):
2733 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2734 elif mobj:
2735 return f'%({tmpl})s'
2736 return tmpl
2737
2738 for tmpl in self.params['forceprint'].get(key, []):
2739 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2740
2741 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2742 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2743 tmpl = format_tmpl(tmpl)
2744 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2745 if self._ensure_dir_exists(filename):
2746 with open(filename, 'a', encoding='utf-8') as f:
2747 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2748
2749 def __forced_printings(self, info_dict, filename, incomplete):
2750 def print_mandatory(field, actual_field=None):
2751 if actual_field is None:
2752 actual_field = field
2753 if (self.params.get('force%s' % field, False)
2754 and (not incomplete or info_dict.get(actual_field) is not None)):
2755 self.to_stdout(info_dict[actual_field])
2756
2757 def print_optional(field):
2758 if (self.params.get('force%s' % field, False)
2759 and info_dict.get(field) is not None):
2760 self.to_stdout(info_dict[field])
2761
2762 info_dict = info_dict.copy()
2763 if filename is not None:
2764 info_dict['filename'] = filename
2765 if info_dict.get('requested_formats') is not None:
2766 # For RTMP URLs, also include the playpath
2767 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2768 elif info_dict.get('url'):
2769 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2770
2771 if (self.params.get('forcejson')
2772 or self.params['forceprint'].get('video')
2773 or self.params['print_to_file'].get('video')):
2774 self.post_extract(info_dict)
2775 self._forceprint('video', info_dict)
2776
2777 print_mandatory('title')
2778 print_mandatory('id')
2779 print_mandatory('url', 'urls')
2780 print_optional('thumbnail')
2781 print_optional('description')
2782 print_optional('filename')
2783 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2784 self.to_stdout(formatSeconds(info_dict['duration']))
2785 print_mandatory('format')
2786
2787 if self.params.get('forcejson'):
2788 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2789
2790 def dl(self, name, info, subtitle=False, test=False):
2791 if not info.get('url'):
2792 self.raise_no_formats(info, True)
2793
2794 if test:
2795 verbose = self.params.get('verbose')
2796 params = {
2797 'test': True,
2798 'quiet': self.params.get('quiet') or not verbose,
2799 'verbose': verbose,
2800 'noprogress': not verbose,
2801 'nopart': True,
2802 'skip_unavailable_fragments': False,
2803 'keep_fragments': False,
2804 'overwrites': True,
2805 '_no_ytdl_file': True,
2806 }
2807 else:
2808 params = self.params
2809 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2810 if not test:
2811 for ph in self._progress_hooks:
2812 fd.add_progress_hook(ph)
2813 urls = '", "'.join(
2814 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2815 for f in info.get('requested_formats', []) or [info])
2816 self.write_debug('Invoking downloader on "%s"' % urls)
2817
2818 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2819 # But it may contain objects that are not deep-copyable
2820 new_info = self._copy_infodict(info)
2821 if new_info.get('http_headers') is None:
2822 new_info['http_headers'] = self._calc_headers(new_info)
2823 return fd.download(name, new_info, subtitle)
2824
2825 def existing_file(self, filepaths, *, default_overwrite=True):
2826 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2827 if existing_files and not self.params.get('overwrites', default_overwrite):
2828 return existing_files[0]
2829
2830 for file in existing_files:
2831 self.report_file_delete(file)
2832 os.remove(file)
2833 return None
2834
2835 def process_info(self, info_dict):
2836 """Process a single resolved IE result. (Modifies it in-place)"""
2837
2838 assert info_dict.get('_type', 'video') == 'video'
2839 original_infodict = info_dict
2840
2841 if 'format' not in info_dict and 'ext' in info_dict:
2842 info_dict['format'] = info_dict['ext']
2843
2844 # This is mostly just for backward compatibility of process_info
2845 # As a side-effect, this allows for format-specific filters
2846 if self._match_entry(info_dict) is not None:
2847 info_dict['__write_download_archive'] = 'ignore'
2848 return
2849
2850 # Does nothing under normal operation - for backward compatibility of process_info
2851 self.post_extract(info_dict)
2852 self._num_downloads += 1
2853
2854 # info_dict['_filename'] needs to be set for backward compatibility
2855 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2856 temp_filename = self.prepare_filename(info_dict, 'temp')
2857 files_to_move = {}
2858
2859 # Forced printings
2860 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2861
2862 if self.params.get('simulate'):
2863 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2864 return
2865
2866 if full_filename is None:
2867 return
2868 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2869 return
2870 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2871 return
2872
2873 if self._write_description('video', info_dict,
2874 self.prepare_filename(info_dict, 'description')) is None:
2875 return
2876
2877 sub_files = self._write_subtitles(info_dict, temp_filename)
2878 if sub_files is None:
2879 return
2880 files_to_move.update(dict(sub_files))
2881
2882 thumb_files = self._write_thumbnails(
2883 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2884 if thumb_files is None:
2885 return
2886 files_to_move.update(dict(thumb_files))
2887
2888 infofn = self.prepare_filename(info_dict, 'infojson')
2889 _infojson_written = self._write_info_json('video', info_dict, infofn)
2890 if _infojson_written:
2891 info_dict['infojson_filename'] = infofn
2892 # For backward compatibility, even though it was a private field
2893 info_dict['__infojson_filename'] = infofn
2894 elif _infojson_written is None:
2895 return
2896
2897 # Note: Annotations are deprecated
2898 annofn = None
2899 if self.params.get('writeannotations', False):
2900 annofn = self.prepare_filename(info_dict, 'annotation')
2901 if annofn:
2902 if not self._ensure_dir_exists(encodeFilename(annofn)):
2903 return
2904 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2905 self.to_screen('[info] Video annotations are already present')
2906 elif not info_dict.get('annotations'):
2907 self.report_warning('There are no annotations to write.')
2908 else:
2909 try:
2910 self.to_screen('[info] Writing video annotations to: ' + annofn)
2911 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2912 annofile.write(info_dict['annotations'])
2913 except (KeyError, TypeError):
2914 self.report_warning('There are no annotations to write.')
2915 except OSError:
2916 self.report_error('Cannot write annotations file: ' + annofn)
2917 return
2918
2919 # Write internet shortcut files
2920 def _write_link_file(link_type):
2921 url = try_get(info_dict['webpage_url'], iri_to_uri)
2922 if not url:
2923 self.report_warning(
2924 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2925 return True
2926 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2927 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2928 return False
2929 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2930 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2931 return True
2932 try:
2933 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2934 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2935 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2936 template_vars = {'url': url}
2937 if link_type == 'desktop':
2938 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2939 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2940 except OSError:
2941 self.report_error(f'Cannot write internet shortcut {linkfn}')
2942 return False
2943 return True
2944
2945 write_links = {
2946 'url': self.params.get('writeurllink'),
2947 'webloc': self.params.get('writewebloclink'),
2948 'desktop': self.params.get('writedesktoplink'),
2949 }
2950 if self.params.get('writelink'):
2951 link_type = ('webloc' if sys.platform == 'darwin'
2952 else 'desktop' if sys.platform.startswith('linux')
2953 else 'url')
2954 write_links[link_type] = True
2955
2956 if any(should_write and not _write_link_file(link_type)
2957 for link_type, should_write in write_links.items()):
2958 return
2959
2960 def replace_info_dict(new_info):
2961 nonlocal info_dict
2962 if new_info == info_dict:
2963 return
2964 info_dict.clear()
2965 info_dict.update(new_info)
2966
2967 try:
2968 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2969 replace_info_dict(new_info)
2970 except PostProcessingError as err:
2971 self.report_error('Preprocessing: %s' % str(err))
2972 return
2973
2974 if self.params.get('skip_download'):
2975 info_dict['filepath'] = temp_filename
2976 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2977 info_dict['__files_to_move'] = files_to_move
2978 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2979 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2980 else:
2981 # Download
2982 info_dict.setdefault('__postprocessors', [])
2983 try:
2984
2985 def existing_video_file(*filepaths):
2986 ext = info_dict.get('ext')
2987 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
2988 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
2989 default_overwrite=False)
2990 if file:
2991 info_dict['ext'] = os.path.splitext(file)[1][1:]
2992 return file
2993
2994 success = True
2995 if info_dict.get('requested_formats') is not None:
2996
2997 def compatible_formats(formats):
2998 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2999 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3000 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3001 if len(video_formats) > 2 or len(audio_formats) > 2:
3002 return False
3003
3004 # Check extension
3005 exts = {format.get('ext') for format in formats}
3006 COMPATIBLE_EXTS = (
3007 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3008 {'webm'},
3009 )
3010 for ext_sets in COMPATIBLE_EXTS:
3011 if ext_sets.issuperset(exts):
3012 return True
3013 # TODO: Check acodec/vcodec
3014 return False
3015
3016 requested_formats = info_dict['requested_formats']
3017 old_ext = info_dict['ext']
3018 if self.params.get('merge_output_format') is None:
3019 if not compatible_formats(requested_formats):
3020 info_dict['ext'] = 'mkv'
3021 self.report_warning(
3022 'Requested formats are incompatible for merge and will be merged into mkv')
3023 if (info_dict['ext'] == 'webm'
3024 and info_dict.get('thumbnails')
3025 # check with type instead of pp_key, __name__, or isinstance
3026 # since we dont want any custom PPs to trigger this
3027 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])):
3028 info_dict['ext'] = 'mkv'
3029 self.report_warning(
3030 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3031 new_ext = info_dict['ext']
3032
3033 def correct_ext(filename, ext=new_ext):
3034 if filename == '-':
3035 return filename
3036 filename_real_ext = os.path.splitext(filename)[1][1:]
3037 filename_wo_ext = (
3038 os.path.splitext(filename)[0]
3039 if filename_real_ext in (old_ext, new_ext)
3040 else filename)
3041 return f'{filename_wo_ext}.{ext}'
3042
3043 # Ensure filename always has a correct extension for successful merge
3044 full_filename = correct_ext(full_filename)
3045 temp_filename = correct_ext(temp_filename)
3046 dl_filename = existing_video_file(full_filename, temp_filename)
3047 info_dict['__real_download'] = False
3048
3049 downloaded = []
3050 merger = FFmpegMergerPP(self)
3051
3052 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3053 if dl_filename is not None:
3054 self.report_file_already_downloaded(dl_filename)
3055 elif fd:
3056 for f in requested_formats if fd != FFmpegFD else []:
3057 f['filepath'] = fname = prepend_extension(
3058 correct_ext(temp_filename, info_dict['ext']),
3059 'f%s' % f['format_id'], info_dict['ext'])
3060 downloaded.append(fname)
3061 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3062 success, real_download = self.dl(temp_filename, info_dict)
3063 info_dict['__real_download'] = real_download
3064 else:
3065 if self.params.get('allow_unplayable_formats'):
3066 self.report_warning(
3067 'You have requested merging of multiple formats '
3068 'while also allowing unplayable formats to be downloaded. '
3069 'The formats won\'t be merged to prevent data corruption.')
3070 elif not merger.available:
3071 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3072 if not self.params.get('ignoreerrors'):
3073 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3074 return
3075 self.report_warning(f'{msg}. The formats won\'t be merged')
3076
3077 if temp_filename == '-':
3078 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3079 else 'but the formats are incompatible for simultaneous download' if merger.available
3080 else 'but ffmpeg is not installed')
3081 self.report_warning(
3082 f'You have requested downloading multiple formats to stdout {reason}. '
3083 'The formats will be streamed one after the other')
3084 fname = temp_filename
3085 for f in requested_formats:
3086 new_info = dict(info_dict)
3087 del new_info['requested_formats']
3088 new_info.update(f)
3089 if temp_filename != '-':
3090 fname = prepend_extension(
3091 correct_ext(temp_filename, new_info['ext']),
3092 'f%s' % f['format_id'], new_info['ext'])
3093 if not self._ensure_dir_exists(fname):
3094 return
3095 f['filepath'] = fname
3096 downloaded.append(fname)
3097 partial_success, real_download = self.dl(fname, new_info)
3098 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3099 success = success and partial_success
3100
3101 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3102 info_dict['__postprocessors'].append(merger)
3103 info_dict['__files_to_merge'] = downloaded
3104 # Even if there were no downloads, it is being merged only now
3105 info_dict['__real_download'] = True
3106 else:
3107 for file in downloaded:
3108 files_to_move[file] = None
3109 else:
3110 # Just a single file
3111 dl_filename = existing_video_file(full_filename, temp_filename)
3112 if dl_filename is None or dl_filename == temp_filename:
3113 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3114 # So we should try to resume the download
3115 success, real_download = self.dl(temp_filename, info_dict)
3116 info_dict['__real_download'] = real_download
3117 else:
3118 self.report_file_already_downloaded(dl_filename)
3119
3120 dl_filename = dl_filename or temp_filename
3121 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3122
3123 except network_exceptions as err:
3124 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3125 return
3126 except OSError as err:
3127 raise UnavailableVideoError(err)
3128 except (ContentTooShortError, ) as err:
3129 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3130 return
3131
3132 if success and full_filename != '-':
3133
3134 def fixup():
3135 do_fixup = True
3136 fixup_policy = self.params.get('fixup')
3137 vid = info_dict['id']
3138
3139 if fixup_policy in ('ignore', 'never'):
3140 return
3141 elif fixup_policy == 'warn':
3142 do_fixup = False
3143 elif fixup_policy != 'force':
3144 assert fixup_policy in ('detect_or_warn', None)
3145 if not info_dict.get('__real_download'):
3146 do_fixup = False
3147
3148 def ffmpeg_fixup(cndn, msg, cls):
3149 if not cndn:
3150 return
3151 if not do_fixup:
3152 self.report_warning(f'{vid}: {msg}')
3153 return
3154 pp = cls(self)
3155 if pp.available:
3156 info_dict['__postprocessors'].append(pp)
3157 else:
3158 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3159
3160 stretched_ratio = info_dict.get('stretched_ratio')
3161 ffmpeg_fixup(
3162 stretched_ratio not in (1, None),
3163 f'Non-uniform pixel ratio {stretched_ratio}',
3164 FFmpegFixupStretchedPP)
3165
3166 ffmpeg_fixup(
3167 (info_dict.get('requested_formats') is None
3168 and info_dict.get('container') == 'm4a_dash'
3169 and info_dict.get('ext') == 'm4a'),
3170 'writing DASH m4a. Only some players support this container',
3171 FFmpegFixupM4aPP)
3172
3173 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3174 downloader = downloader.__name__ if downloader else None
3175
3176 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3177 ffmpeg_fixup(downloader == 'HlsFD',
3178 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3179 FFmpegFixupM3u8PP)
3180 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3181 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3182
3183 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3184 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed duration detected', FFmpegFixupDurationPP)
3185
3186 fixup()
3187 try:
3188 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3189 except PostProcessingError as err:
3190 self.report_error('Postprocessing: %s' % str(err))
3191 return
3192 try:
3193 for ph in self._post_hooks:
3194 ph(info_dict['filepath'])
3195 except Exception as err:
3196 self.report_error('post hooks: %s' % str(err))
3197 return
3198 info_dict['__write_download_archive'] = True
3199
3200 if self.params.get('force_write_download_archive'):
3201 info_dict['__write_download_archive'] = True
3202
3203 # Make sure the info_dict was modified in-place
3204 assert info_dict is original_infodict
3205
3206 max_downloads = self.params.get('max_downloads')
3207 if max_downloads is not None and self._num_downloads >= int(max_downloads):
3208 raise MaxDownloadsReached()
3209
3210 def __download_wrapper(self, func):
3211 @functools.wraps(func)
3212 def wrapper(*args, **kwargs):
3213 try:
3214 res = func(*args, **kwargs)
3215 except UnavailableVideoError as e:
3216 self.report_error(e)
3217 except MaxDownloadsReached as e:
3218 self.to_screen(f'[info] {e}')
3219 raise
3220 except DownloadCancelled as e:
3221 self.to_screen(f'[info] {e}')
3222 if not self.params.get('break_per_url'):
3223 raise
3224 else:
3225 if self.params.get('dump_single_json', False):
3226 self.post_extract(res)
3227 self.to_stdout(json.dumps(self.sanitize_info(res)))
3228 return wrapper
3229
3230 def download(self, url_list):
3231 """Download a given list of URLs."""
3232 url_list = variadic(url_list) # Passing a single URL is a common mistake
3233 outtmpl = self.outtmpl_dict['default']
3234 if (len(url_list) > 1
3235 and outtmpl != '-'
3236 and '%' not in outtmpl
3237 and self.params.get('max_downloads') != 1):
3238 raise SameFileError(outtmpl)
3239
3240 for url in url_list:
3241 self.__download_wrapper(self.extract_info)(
3242 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3243
3244 return self._download_retcode
3245
3246 def download_with_info_file(self, info_filename):
3247 with contextlib.closing(fileinput.FileInput(
3248 [info_filename], mode='r',
3249 openhook=fileinput.hook_encoded('utf-8'))) as f:
3250 # FileInput doesn't have a read method, we can't call json.load
3251 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3252 try:
3253 self.__download_wrapper(self.process_ie_result)(info, download=True)
3254 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3255 if not isinstance(e, EntryNotInPlaylist):
3256 self.to_stderr('\r')
3257 webpage_url = info.get('webpage_url')
3258 if webpage_url is not None:
3259 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3260 return self.download([webpage_url])
3261 else:
3262 raise
3263 return self._download_retcode
3264
3265 @staticmethod
3266 def sanitize_info(info_dict, remove_private_keys=False):
3267 ''' Sanitize the infodict for converting to json '''
3268 if info_dict is None:
3269 return info_dict
3270 info_dict.setdefault('epoch', int(time.time()))
3271 info_dict.setdefault('_type', 'video')
3272
3273 if remove_private_keys:
3274 reject = lambda k, v: v is None or (k.startswith('_') and k != '_type') or k in {
3275 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3276 'entries', 'filepath', 'infojson_filename', 'original_url', 'playlist_autonumber',
3277 }
3278 else:
3279 reject = lambda k, v: False
3280
3281 def filter_fn(obj):
3282 if isinstance(obj, dict):
3283 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3284 elif isinstance(obj, (list, tuple, set, LazyList)):
3285 return list(map(filter_fn, obj))
3286 elif obj is None or isinstance(obj, (str, int, float, bool)):
3287 return obj
3288 else:
3289 return repr(obj)
3290
3291 return filter_fn(info_dict)
3292
3293 @staticmethod
3294 def filter_requested_info(info_dict, actually_filter=True):
3295 ''' Alias of sanitize_info for backward compatibility '''
3296 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3297
3298 @staticmethod
3299 def post_extract(info_dict):
3300 def actual_post_extract(info_dict):
3301 if info_dict.get('_type') in ('playlist', 'multi_video'):
3302 for video_dict in info_dict.get('entries', {}):
3303 actual_post_extract(video_dict or {})
3304 return
3305
3306 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3307 info_dict.update(post_extractor())
3308
3309 actual_post_extract(info_dict or {})
3310
3311 def run_pp(self, pp, infodict):
3312 files_to_delete = []
3313 if '__files_to_move' not in infodict:
3314 infodict['__files_to_move'] = {}
3315 try:
3316 files_to_delete, infodict = pp.run(infodict)
3317 except PostProcessingError as e:
3318 # Must be True and not 'only_download'
3319 if self.params.get('ignoreerrors') is True:
3320 self.report_error(e)
3321 return infodict
3322 raise
3323
3324 if not files_to_delete:
3325 return infodict
3326 if self.params.get('keepvideo', False):
3327 for f in files_to_delete:
3328 infodict['__files_to_move'].setdefault(f, '')
3329 else:
3330 for old_filename in set(files_to_delete):
3331 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
3332 try:
3333 os.remove(encodeFilename(old_filename))
3334 except OSError:
3335 self.report_warning('Unable to remove downloaded original file')
3336 if old_filename in infodict['__files_to_move']:
3337 del infodict['__files_to_move'][old_filename]
3338 return infodict
3339
3340 def run_all_pps(self, key, info, *, additional_pps=None):
3341 self._forceprint(key, info)
3342 for pp in (additional_pps or []) + self._pps[key]:
3343 info = self.run_pp(pp, info)
3344 return info
3345
3346 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3347 info = dict(ie_info)
3348 info['__files_to_move'] = files_to_move or {}
3349 info = self.run_all_pps(key, info)
3350 return info, info.pop('__files_to_move', None)
3351
3352 def post_process(self, filename, info, files_to_move=None):
3353 """Run all the postprocessors on the given file."""
3354 info['filepath'] = filename
3355 info['__files_to_move'] = files_to_move or {}
3356 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3357 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3358 del info['__files_to_move']
3359 return self.run_all_pps('after_move', info)
3360
3361 def _make_archive_id(self, info_dict):
3362 video_id = info_dict.get('id')
3363 if not video_id:
3364 return
3365 # Future-proof against any change in case
3366 # and backwards compatibility with prior versions
3367 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3368 if extractor is None:
3369 url = str_or_none(info_dict.get('url'))
3370 if not url:
3371 return
3372 # Try to find matching extractor for the URL and take its ie_key
3373 for ie_key, ie in self._ies.items():
3374 if ie.suitable(url):
3375 extractor = ie_key
3376 break
3377 else:
3378 return
3379 return f'{extractor.lower()} {video_id}'
3380
3381 def in_download_archive(self, info_dict):
3382 fn = self.params.get('download_archive')
3383 if fn is None:
3384 return False
3385
3386 vid_id = self._make_archive_id(info_dict)
3387 if not vid_id:
3388 return False # Incomplete video information
3389
3390 return vid_id in self.archive
3391
3392 def record_download_archive(self, info_dict):
3393 fn = self.params.get('download_archive')
3394 if fn is None:
3395 return
3396 vid_id = self._make_archive_id(info_dict)
3397 assert vid_id
3398 self.write_debug(f'Adding to archive: {vid_id}')
3399 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3400 archive_file.write(vid_id + '\n')
3401 self.archive.add(vid_id)
3402
3403 @staticmethod
3404 def format_resolution(format, default='unknown'):
3405 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3406 return 'audio only'
3407 if format.get('resolution') is not None:
3408 return format['resolution']
3409 if format.get('width') and format.get('height'):
3410 return '%dx%d' % (format['width'], format['height'])
3411 elif format.get('height'):
3412 return '%sp' % format['height']
3413 elif format.get('width'):
3414 return '%dx?' % format['width']
3415 return default
3416
3417 def _list_format_headers(self, *headers):
3418 if self.params.get('listformats_table', True) is not False:
3419 return [self._format_screen(header, self.Styles.HEADERS) for header in headers]
3420 return headers
3421
3422 def _format_note(self, fdict):
3423 res = ''
3424 if fdict.get('ext') in ['f4f', 'f4m']:
3425 res += '(unsupported)'
3426 if fdict.get('language'):
3427 if res:
3428 res += ' '
3429 res += '[%s]' % fdict['language']
3430 if fdict.get('format_note') is not None:
3431 if res:
3432 res += ' '
3433 res += fdict['format_note']
3434 if fdict.get('tbr') is not None:
3435 if res:
3436 res += ', '
3437 res += '%4dk' % fdict['tbr']
3438 if fdict.get('container') is not None:
3439 if res:
3440 res += ', '
3441 res += '%s container' % fdict['container']
3442 if (fdict.get('vcodec') is not None
3443 and fdict.get('vcodec') != 'none'):
3444 if res:
3445 res += ', '
3446 res += fdict['vcodec']
3447 if fdict.get('vbr') is not None:
3448 res += '@'
3449 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3450 res += 'video@'
3451 if fdict.get('vbr') is not None:
3452 res += '%4dk' % fdict['vbr']
3453 if fdict.get('fps') is not None:
3454 if res:
3455 res += ', '
3456 res += '%sfps' % fdict['fps']
3457 if fdict.get('acodec') is not None:
3458 if res:
3459 res += ', '
3460 if fdict['acodec'] == 'none':
3461 res += 'video only'
3462 else:
3463 res += '%-5s' % fdict['acodec']
3464 elif fdict.get('abr') is not None:
3465 if res:
3466 res += ', '
3467 res += 'audio'
3468 if fdict.get('abr') is not None:
3469 res += '@%3dk' % fdict['abr']
3470 if fdict.get('asr') is not None:
3471 res += ' (%5dHz)' % fdict['asr']
3472 if fdict.get('filesize') is not None:
3473 if res:
3474 res += ', '
3475 res += format_bytes(fdict['filesize'])
3476 elif fdict.get('filesize_approx') is not None:
3477 if res:
3478 res += ', '
3479 res += '~' + format_bytes(fdict['filesize_approx'])
3480 return res
3481
3482 def render_formats_table(self, info_dict):
3483 if not info_dict.get('formats') and not info_dict.get('url'):
3484 return None
3485
3486 formats = info_dict.get('formats', [info_dict])
3487 if not self.params.get('listformats_table', True) is not False:
3488 table = [
3489 [
3490 format_field(f, 'format_id'),
3491 format_field(f, 'ext'),
3492 self.format_resolution(f),
3493 self._format_note(f)
3494 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3495 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3496
3497 delim = self._format_screen('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3498 table = [
3499 [
3500 self._format_screen(format_field(f, 'format_id'), self.Styles.ID),
3501 format_field(f, 'ext'),
3502 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3503 format_field(f, 'fps', '\t%d'),
3504 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3505 delim,
3506 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3507 format_field(f, 'tbr', '\t%dk'),
3508 shorten_protocol_name(f.get('protocol', '')),
3509 delim,
3510 format_field(f, 'vcodec', default='unknown').replace(
3511 'none', 'images' if f.get('acodec') == 'none'
3512 else self._format_screen('audio only', self.Styles.SUPPRESS)),
3513 format_field(f, 'vbr', '\t%dk'),
3514 format_field(f, 'acodec', default='unknown').replace(
3515 'none', '' if f.get('vcodec') == 'none'
3516 else self._format_screen('video only', self.Styles.SUPPRESS)),
3517 format_field(f, 'abr', '\t%dk'),
3518 format_field(f, 'asr', '\t%dHz'),
3519 join_nonempty(
3520 self._format_screen('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3521 format_field(f, 'language', '[%s]'),
3522 join_nonempty(format_field(f, 'format_note'),
3523 format_field(f, 'container', ignore=(None, f.get('ext'))),
3524 delim=', '),
3525 delim=' '),
3526 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3527 header_line = self._list_format_headers(
3528 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3529 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3530
3531 return render_table(
3532 header_line, table, hide_empty=True,
3533 delim=self._format_screen('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3534
3535 def render_thumbnails_table(self, info_dict):
3536 thumbnails = list(info_dict.get('thumbnails') or [])
3537 if not thumbnails:
3538 return None
3539 return render_table(
3540 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3541 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3542
3543 def render_subtitles_table(self, video_id, subtitles):
3544 def _row(lang, formats):
3545 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3546 if len(set(names)) == 1:
3547 names = [] if names[0] == 'unknown' else names[:1]
3548 return [lang, ', '.join(names), ', '.join(exts)]
3549
3550 if not subtitles:
3551 return None
3552 return render_table(
3553 self._list_format_headers('Language', 'Name', 'Formats'),
3554 [_row(lang, formats) for lang, formats in subtitles.items()],
3555 hide_empty=True)
3556
3557 def __list_table(self, video_id, name, func, *args):
3558 table = func(*args)
3559 if not table:
3560 self.to_screen(f'{video_id} has no {name}')
3561 return
3562 self.to_screen(f'[info] Available {name} for {video_id}:')
3563 self.to_stdout(table)
3564
3565 def list_formats(self, info_dict):
3566 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3567
3568 def list_thumbnails(self, info_dict):
3569 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3570
3571 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3572 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3573
3574 def urlopen(self, req):
3575 """ Start an HTTP download """
3576 if isinstance(req, str):
3577 req = sanitized_Request(req)
3578 return self._opener.open(req, timeout=self._socket_timeout)
3579
3580 def print_debug_header(self):
3581 if not self.params.get('verbose'):
3582 return
3583
3584 def get_encoding(stream):
3585 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3586 if not supports_terminal_sequences(stream):
3587 from .compat import WINDOWS_VT_MODE
3588 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3589 return ret
3590
3591 encoding_str = 'Encodings: locale %s, fs %s, out %s, err %s, pref %s' % (
3592 locale.getpreferredencoding(),
3593 sys.getfilesystemencoding(),
3594 get_encoding(self._out_files['screen']), get_encoding(self._out_files['error']),
3595 self.get_encoding())
3596
3597 logger = self.params.get('logger')
3598 if logger:
3599 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3600 write_debug(encoding_str)
3601 else:
3602 write_string(f'[debug] {encoding_str}\n', encoding=None)
3603 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3604
3605 source = detect_variant()
3606 write_debug(join_nonempty(
3607 'yt-dlp version', __version__,
3608 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3609 '' if source == 'unknown' else f'({source})',
3610 delim=' '))
3611 if not _LAZY_LOADER:
3612 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3613 write_debug('Lazy loading extractors is forcibly disabled')
3614 else:
3615 write_debug('Lazy loading extractors is disabled')
3616 if plugin_extractors or plugin_postprocessors:
3617 write_debug('Plugins: %s' % [
3618 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3619 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3620 if self.params.get('compat_opts'):
3621 write_debug('Compatibility options: %s' % ', '.join(self.params.get('compat_opts')))
3622
3623 if source == 'source':
3624 try:
3625 sp = Popen(
3626 ['git', 'rev-parse', '--short', 'HEAD'],
3627 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3628 cwd=os.path.dirname(os.path.abspath(__file__)))
3629 out, err = sp.communicate_or_kill()
3630 out = out.decode().strip()
3631 if re.match('[0-9a-f]+', out):
3632 write_debug('Git HEAD: %s' % out)
3633 except Exception:
3634 try:
3635 sys.exc_clear()
3636 except Exception:
3637 pass
3638
3639 def python_implementation():
3640 impl_name = platform.python_implementation()
3641 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3642 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3643 return impl_name
3644
3645 write_debug('Python version %s (%s %s) - %s' % (
3646 platform.python_version(),
3647 python_implementation(),
3648 platform.architecture()[0],
3649 platform_name()))
3650
3651 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3652 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3653 if ffmpeg_features:
3654 exe_versions['ffmpeg'] += ' (%s)' % ','.join(ffmpeg_features)
3655
3656 exe_versions['rtmpdump'] = rtmpdump_version()
3657 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3658 exe_str = ', '.join(
3659 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3660 ) or 'none'
3661 write_debug('exe versions: %s' % exe_str)
3662
3663 from .cookies import SECRETSTORAGE_AVAILABLE, SQLITE_AVAILABLE
3664 from .downloader.websocket import has_websockets
3665 from .postprocessor.embedthumbnail import has_mutagen
3666
3667 lib_str = join_nonempty(
3668 compat_brotli and compat_brotli.__name__,
3669 has_certifi and 'certifi',
3670 compat_pycrypto_AES and compat_pycrypto_AES.__name__.split('.')[0],
3671 SECRETSTORAGE_AVAILABLE and 'secretstorage',
3672 has_mutagen and 'mutagen',
3673 SQLITE_AVAILABLE and 'sqlite',
3674 has_websockets and 'websockets',
3675 delim=', ') or 'none'
3676 write_debug('Optional libraries: %s' % lib_str)
3677
3678 self._setup_opener()
3679 proxy_map = {}
3680 for handler in self._opener.handlers:
3681 if hasattr(handler, 'proxies'):
3682 proxy_map.update(handler.proxies)
3683 write_debug(f'Proxy map: {proxy_map}')
3684
3685 # Not implemented
3686 if False and self.params.get('call_home'):
3687 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3688 write_debug('Public IP address: %s' % ipaddr)
3689 latest_version = self.urlopen(
3690 'https://yt-dl.org/latest/version').read().decode('utf-8')
3691 if version_tuple(latest_version) > version_tuple(__version__):
3692 self.report_warning(
3693 'You are using an outdated version (newest version: %s)! '
3694 'See https://yt-dl.org/update if you need help updating.' %
3695 latest_version)
3696
3697 def _setup_opener(self):
3698 if hasattr(self, '_opener'):
3699 return
3700 timeout_val = self.params.get('socket_timeout')
3701 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3702
3703 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3704 opts_cookiefile = self.params.get('cookiefile')
3705 opts_proxy = self.params.get('proxy')
3706
3707 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3708
3709 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3710 if opts_proxy is not None:
3711 if opts_proxy == '':
3712 proxies = {}
3713 else:
3714 proxies = {'http': opts_proxy, 'https': opts_proxy}
3715 else:
3716 proxies = compat_urllib_request.getproxies()
3717 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3718 if 'http' in proxies and 'https' not in proxies:
3719 proxies['https'] = proxies['http']
3720 proxy_handler = PerRequestProxyHandler(proxies)
3721
3722 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3723 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3724 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3725 redirect_handler = YoutubeDLRedirectHandler()
3726 data_handler = urllib.request.DataHandler()
3727
3728 # When passing our own FileHandler instance, build_opener won't add the
3729 # default FileHandler and allows us to disable the file protocol, which
3730 # can be used for malicious purposes (see
3731 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3732 file_handler = compat_urllib_request.FileHandler()
3733
3734 def file_open(*args, **kwargs):
3735 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3736 file_handler.file_open = file_open
3737
3738 opener = compat_urllib_request.build_opener(
3739 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3740
3741 # Delete the default user-agent header, which would otherwise apply in
3742 # cases where our custom HTTP handler doesn't come into play
3743 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3744 opener.addheaders = []
3745 self._opener = opener
3746
3747 def encode(self, s):
3748 if isinstance(s, bytes):
3749 return s # Already encoded
3750
3751 try:
3752 return s.encode(self.get_encoding())
3753 except UnicodeEncodeError as err:
3754 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3755 raise
3756
3757 def get_encoding(self):
3758 encoding = self.params.get('encoding')
3759 if encoding is None:
3760 encoding = preferredencoding()
3761 return encoding
3762
3763 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3764 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3765 if overwrite is None:
3766 overwrite = self.params.get('overwrites', True)
3767 if not self.params.get('writeinfojson'):
3768 return False
3769 elif not infofn:
3770 self.write_debug(f'Skipping writing {label} infojson')
3771 return False
3772 elif not self._ensure_dir_exists(infofn):
3773 return None
3774 elif not overwrite and os.path.exists(infofn):
3775 self.to_screen(f'[info] {label.title()} metadata is already present')
3776 return 'exists'
3777
3778 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3779 try:
3780 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3781 return True
3782 except OSError:
3783 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3784 return None
3785
3786 def _write_description(self, label, ie_result, descfn):
3787 ''' Write description and returns True = written, False = skip, None = error '''
3788 if not self.params.get('writedescription'):
3789 return False
3790 elif not descfn:
3791 self.write_debug(f'Skipping writing {label} description')
3792 return False
3793 elif not self._ensure_dir_exists(descfn):
3794 return None
3795 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3796 self.to_screen(f'[info] {label.title()} description is already present')
3797 elif ie_result.get('description') is None:
3798 self.report_warning(f'There\'s no {label} description to write')
3799 return False
3800 else:
3801 try:
3802 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3803 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3804 descfile.write(ie_result['description'])
3805 except OSError:
3806 self.report_error(f'Cannot write {label} description file {descfn}')
3807 return None
3808 return True
3809
3810 def _write_subtitles(self, info_dict, filename):
3811 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3812 ret = []
3813 subtitles = info_dict.get('requested_subtitles')
3814 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3815 # subtitles download errors are already managed as troubles in relevant IE
3816 # that way it will silently go on when used with unsupporting IE
3817 return ret
3818
3819 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3820 if not sub_filename_base:
3821 self.to_screen('[info] Skipping writing video subtitles')
3822 return ret
3823 for sub_lang, sub_info in subtitles.items():
3824 sub_format = sub_info['ext']
3825 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3826 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3827 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3828 if existing_sub:
3829 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3830 sub_info['filepath'] = existing_sub
3831 ret.append((existing_sub, sub_filename_final))
3832 continue
3833
3834 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3835 if sub_info.get('data') is not None:
3836 try:
3837 # Use newline='' to prevent conversion of newline characters
3838 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3839 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3840 subfile.write(sub_info['data'])
3841 sub_info['filepath'] = sub_filename
3842 ret.append((sub_filename, sub_filename_final))
3843 continue
3844 except OSError:
3845 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3846 return None
3847
3848 try:
3849 sub_copy = sub_info.copy()
3850 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3851 self.dl(sub_filename, sub_copy, subtitle=True)
3852 sub_info['filepath'] = sub_filename
3853 ret.append((sub_filename, sub_filename_final))
3854 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3855 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3856 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3857 if not self.params.get('ignoreerrors'):
3858 self.report_error(msg)
3859 raise DownloadError(msg)
3860 self.report_warning(msg)
3861 return ret
3862
3863 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3864 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3865 write_all = self.params.get('write_all_thumbnails', False)
3866 thumbnails, ret = [], []
3867 if write_all or self.params.get('writethumbnail', False):
3868 thumbnails = info_dict.get('thumbnails') or []
3869 multiple = write_all and len(thumbnails) > 1
3870
3871 if thumb_filename_base is None:
3872 thumb_filename_base = filename
3873 if thumbnails and not thumb_filename_base:
3874 self.write_debug(f'Skipping writing {label} thumbnail')
3875 return ret
3876
3877 for idx, t in list(enumerate(thumbnails))[::-1]:
3878 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3879 thumb_display_id = f'{label} thumbnail {t["id"]}'
3880 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3881 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3882
3883 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3884 if existing_thumb:
3885 self.to_screen('[info] %s is already present' % (
3886 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3887 t['filepath'] = existing_thumb
3888 ret.append((existing_thumb, thumb_filename_final))
3889 else:
3890 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3891 try:
3892 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3893 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3894 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3895 shutil.copyfileobj(uf, thumbf)
3896 ret.append((thumb_filename, thumb_filename_final))
3897 t['filepath'] = thumb_filename
3898 except network_exceptions as err:
3899 thumbnails.pop(idx)
3900 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3901 if ret and not write_all:
3902 break
3903 return ret