]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
3ee3ed7d2b2e28354f54de459e281ee00f739d63
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 #!/usr/bin/env python3
2 # coding: utf-8
3
4 from __future__ import absolute_import, unicode_literals
5
6 import collections
7 import contextlib
8 import datetime
9 import errno
10 import fileinput
11 import functools
12 import io
13 import itertools
14 import json
15 import locale
16 import operator
17 import os
18 import platform
19 import re
20 import shutil
21 import subprocess
22 import sys
23 import tempfile
24 import time
25 import tokenize
26 import traceback
27 import random
28 import unicodedata
29
30 from enum import Enum
31 from string import ascii_letters
32
33 from .compat import (
34 compat_basestring,
35 compat_brotli,
36 compat_get_terminal_size,
37 compat_kwargs,
38 compat_numeric_types,
39 compat_os_name,
40 compat_pycrypto_AES,
41 compat_shlex_quote,
42 compat_str,
43 compat_tokenize_tokenize,
44 compat_urllib_error,
45 compat_urllib_request,
46 compat_urllib_request_DataHandler,
47 windows_enable_vt_mode,
48 )
49 from .cookies import load_cookies
50 from .utils import (
51 age_restricted,
52 args_to_str,
53 ContentTooShortError,
54 date_from_str,
55 DateRange,
56 DEFAULT_OUTTMPL,
57 determine_ext,
58 determine_protocol,
59 DownloadCancelled,
60 DownloadError,
61 encode_compat_str,
62 encodeFilename,
63 EntryNotInPlaylist,
64 error_to_compat_str,
65 ExistingVideoReached,
66 expand_path,
67 ExtractorError,
68 float_or_none,
69 format_bytes,
70 format_field,
71 format_decimal_suffix,
72 formatSeconds,
73 GeoRestrictedError,
74 get_domain,
75 HEADRequest,
76 InAdvancePagedList,
77 int_or_none,
78 iri_to_uri,
79 ISO3166Utils,
80 join_nonempty,
81 LazyList,
82 LINK_TEMPLATES,
83 locked_file,
84 make_dir,
85 make_HTTPS_handler,
86 MaxDownloadsReached,
87 merge_headers,
88 network_exceptions,
89 number_of_digits,
90 orderedSet,
91 OUTTMPL_TYPES,
92 PagedList,
93 parse_filesize,
94 PerRequestProxyHandler,
95 platform_name,
96 Popen,
97 POSTPROCESS_WHEN,
98 PostProcessingError,
99 preferredencoding,
100 prepend_extension,
101 ReExtractInfo,
102 register_socks_protocols,
103 RejectedVideoReached,
104 remove_terminal_sequences,
105 render_table,
106 replace_extension,
107 SameFileError,
108 sanitize_filename,
109 sanitize_path,
110 sanitize_url,
111 sanitized_Request,
112 std_headers,
113 STR_FORMAT_RE_TMPL,
114 STR_FORMAT_TYPES,
115 str_or_none,
116 strftime_or_none,
117 subtitles_filename,
118 supports_terminal_sequences,
119 timetuple_from_msec,
120 to_high_limit_path,
121 traverse_obj,
122 try_get,
123 UnavailableVideoError,
124 url_basename,
125 variadic,
126 version_tuple,
127 write_json_file,
128 write_string,
129 YoutubeDLCookieProcessor,
130 YoutubeDLHandler,
131 YoutubeDLRedirectHandler,
132 )
133 from .cache import Cache
134 from .minicurses import format_text
135 from .extractor import (
136 gen_extractor_classes,
137 get_info_extractor,
138 _LAZY_LOADER,
139 _PLUGIN_CLASSES as plugin_extractors
140 )
141 from .extractor.openload import PhantomJSwrapper
142 from .downloader import (
143 FFmpegFD,
144 get_suitable_downloader,
145 shorten_protocol_name
146 )
147 from .downloader.rtmp import rtmpdump_version
148 from .postprocessor import (
149 get_postprocessor,
150 EmbedThumbnailPP,
151 FFmpegFixupDuplicateMoovPP,
152 FFmpegFixupDurationPP,
153 FFmpegFixupM3u8PP,
154 FFmpegFixupM4aPP,
155 FFmpegFixupStretchedPP,
156 FFmpegFixupTimestampPP,
157 FFmpegMergerPP,
158 FFmpegPostProcessor,
159 MoveFilesAfterDownloadPP,
160 _PLUGIN_CLASSES as plugin_postprocessors
161 )
162 from .update import detect_variant
163 from .version import __version__, RELEASE_GIT_HEAD
164
165 if compat_os_name == 'nt':
166 import ctypes
167
168
169 class YoutubeDL(object):
170 """YoutubeDL class.
171
172 YoutubeDL objects are the ones responsible of downloading the
173 actual video file and writing it to disk if the user has requested
174 it, among some other tasks. In most cases there should be one per
175 program. As, given a video URL, the downloader doesn't know how to
176 extract all the needed information, task that InfoExtractors do, it
177 has to pass the URL to one of them.
178
179 For this, YoutubeDL objects have a method that allows
180 InfoExtractors to be registered in a given order. When it is passed
181 a URL, the YoutubeDL object handles it to the first InfoExtractor it
182 finds that reports being able to handle it. The InfoExtractor extracts
183 all the information about the video or videos the URL refers to, and
184 YoutubeDL process the extracted information, possibly using a File
185 Downloader to download the video.
186
187 YoutubeDL objects accept a lot of parameters. In order not to saturate
188 the object constructor with arguments, it receives a dictionary of
189 options instead. These options are available through the params
190 attribute for the InfoExtractors to use. The YoutubeDL also
191 registers itself as the downloader in charge for the InfoExtractors
192 that are added to it, so this is a "mutual registration".
193
194 Available options:
195
196 username: Username for authentication purposes.
197 password: Password for authentication purposes.
198 videopassword: Password for accessing a video.
199 ap_mso: Adobe Pass multiple-system operator identifier.
200 ap_username: Multiple-system operator account username.
201 ap_password: Multiple-system operator account password.
202 usenetrc: Use netrc for authentication instead.
203 verbose: Print additional info to stdout.
204 quiet: Do not print messages to stdout.
205 no_warnings: Do not print out anything for warnings.
206 forceprint: A dict with keys WHEN mapped to a list of templates to
207 print to stdout. The allowed keys are video or any of the
208 items in utils.POSTPROCESS_WHEN.
209 For compatibility, a single list is also accepted
210 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
211 a list of tuples with (template, filename)
212 forceurl: Force printing final URL. (Deprecated)
213 forcetitle: Force printing title. (Deprecated)
214 forceid: Force printing ID. (Deprecated)
215 forcethumbnail: Force printing thumbnail URL. (Deprecated)
216 forcedescription: Force printing description. (Deprecated)
217 forcefilename: Force printing final filename. (Deprecated)
218 forceduration: Force printing duration. (Deprecated)
219 forcejson: Force printing info_dict as JSON.
220 dump_single_json: Force printing the info_dict of the whole playlist
221 (or video) as a single JSON line.
222 force_write_download_archive: Force writing download archive regardless
223 of 'skip_download' or 'simulate'.
224 simulate: Do not download the video files. If unset (or None),
225 simulate only if listsubtitles, listformats or list_thumbnails is used
226 format: Video format code. see "FORMAT SELECTION" for more details.
227 You can also pass a function. The function takes 'ctx' as
228 argument and returns the formats to download.
229 See "build_format_selector" for an implementation
230 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
231 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
232 extracting metadata even if the video is not actually
233 available for download (experimental)
234 format_sort: A list of fields by which to sort the video formats.
235 See "Sorting Formats" for more details.
236 format_sort_force: Force the given format_sort. see "Sorting Formats"
237 for more details.
238 prefer_free_formats: Whether to prefer video formats with free containers
239 over non-free ones of same quality.
240 allow_multiple_video_streams: Allow multiple video streams to be merged
241 into a single file
242 allow_multiple_audio_streams: Allow multiple audio streams to be merged
243 into a single file
244 check_formats Whether to test if the formats are downloadable.
245 Can be True (check all), False (check none),
246 'selected' (check selected formats),
247 or None (check only if requested by extractor)
248 paths: Dictionary of output paths. The allowed keys are 'home'
249 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
250 outtmpl: Dictionary of templates for output names. Allowed keys
251 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
252 For compatibility with youtube-dl, a single string can also be used
253 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
254 restrictfilenames: Do not allow "&" and spaces in file names
255 trim_file_name: Limit length of filename (extension excluded)
256 windowsfilenames: Force the filenames to be windows compatible
257 ignoreerrors: Do not stop on download/postprocessing errors.
258 Can be 'only_download' to ignore only download errors.
259 Default is 'only_download' for CLI, but False for API
260 skip_playlist_after_errors: Number of allowed failures until the rest of
261 the playlist is skipped
262 force_generic_extractor: Force downloader to use the generic extractor
263 overwrites: Overwrite all video and metadata files if True,
264 overwrite only non-video files if None
265 and don't overwrite any file if False
266 For compatibility with youtube-dl,
267 "nooverwrites" may also be used instead
268 playliststart: Playlist item to start at.
269 playlistend: Playlist item to end at.
270 playlist_items: Specific indices of playlist to download.
271 playlistreverse: Download playlist items in reverse order.
272 playlistrandom: Download playlist items in random order.
273 matchtitle: Download only matching titles.
274 rejecttitle: Reject downloads for matching titles.
275 logger: Log messages to a logging.Logger instance.
276 logtostderr: Log messages to stderr instead of stdout.
277 consoletitle: Display progress in console window's titlebar.
278 writedescription: Write the video description to a .description file
279 writeinfojson: Write the video description to a .info.json file
280 clean_infojson: Remove private fields from the infojson
281 getcomments: Extract video comments. This will not be written to disk
282 unless writeinfojson is also given
283 writeannotations: Write the video annotations to a .annotations.xml file
284 writethumbnail: Write the thumbnail image to a file
285 allow_playlist_files: Whether to write playlists' description, infojson etc
286 also to disk when using the 'write*' options
287 write_all_thumbnails: Write all thumbnail formats to files
288 writelink: Write an internet shortcut file, depending on the
289 current platform (.url/.webloc/.desktop)
290 writeurllink: Write a Windows internet shortcut file (.url)
291 writewebloclink: Write a macOS internet shortcut file (.webloc)
292 writedesktoplink: Write a Linux internet shortcut file (.desktop)
293 writesubtitles: Write the video subtitles to a file
294 writeautomaticsub: Write the automatically generated subtitles to a file
295 allsubtitles: Deprecated - Use subtitleslangs = ['all']
296 Downloads all the subtitles of the video
297 (requires writesubtitles or writeautomaticsub)
298 listsubtitles: Lists all available subtitles for the video
299 subtitlesformat: The format code for subtitles
300 subtitleslangs: List of languages of the subtitles to download (can be regex).
301 The list may contain "all" to refer to all the available
302 subtitles. The language can be prefixed with a "-" to
303 exclude it from the requested languages. Eg: ['all', '-live_chat']
304 keepvideo: Keep the video file after post-processing
305 daterange: A DateRange object, download only if the upload_date is in the range.
306 skip_download: Skip the actual download of the video file
307 cachedir: Location of the cache files in the filesystem.
308 False to disable filesystem cache.
309 noplaylist: Download single video instead of a playlist if in doubt.
310 age_limit: An integer representing the user's age in years.
311 Unsuitable videos for the given age are skipped.
312 min_views: An integer representing the minimum view count the video
313 must have in order to not be skipped.
314 Videos without view count information are always
315 downloaded. None for no limit.
316 max_views: An integer representing the maximum view count.
317 Videos that are more popular than that are not
318 downloaded.
319 Videos without view count information are always
320 downloaded. None for no limit.
321 download_archive: File name of a file where all downloads are recorded.
322 Videos already present in the file are not downloaded
323 again.
324 break_on_existing: Stop the download process after attempting to download a
325 file that is in the archive.
326 break_on_reject: Stop the download process when encountering a video that
327 has been filtered out.
328 break_per_url: Whether break_on_reject and break_on_existing
329 should act on each input URL as opposed to for the entire queue
330 cookiefile: File name where cookies should be read from and dumped to
331 cookiesfrombrowser: A tuple containing the name of the browser, the profile
332 name/pathfrom where cookies are loaded, and the name of the
333 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
334 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
335 support RFC 5746 secure renegotiation
336 nocheckcertificate: Do not verify SSL certificates
337 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
338 At the moment, this is only supported by YouTube.
339 http_headers: A dictionary of custom headers to be used for all requests
340 proxy: URL of the proxy server to use
341 geo_verification_proxy: URL of the proxy to use for IP address verification
342 on geo-restricted sites.
343 socket_timeout: Time to wait for unresponsive hosts, in seconds
344 bidi_workaround: Work around buggy terminals without bidirectional text
345 support, using fridibi
346 debug_printtraffic:Print out sent and received HTTP traffic
347 include_ads: Download ads as well (deprecated)
348 default_search: Prepend this string if an input url is not valid.
349 'auto' for elaborate guessing
350 encoding: Use this encoding instead of the system-specified.
351 extract_flat: Do not resolve URLs, return the immediate result.
352 Pass in 'in_playlist' to only show this behavior for
353 playlist items.
354 wait_for_video: If given, wait for scheduled streams to become available.
355 The value should be a tuple containing the range
356 (min_secs, max_secs) to wait between retries
357 postprocessors: A list of dictionaries, each with an entry
358 * key: The name of the postprocessor. See
359 yt_dlp/postprocessor/__init__.py for a list.
360 * when: When to run the postprocessor. Allowed values are
361 the entries of utils.POSTPROCESS_WHEN
362 Assumed to be 'post_process' if not given
363 post_hooks: Deprecated - Register a custom postprocessor instead
364 A list of functions that get called as the final step
365 for each video file, after all postprocessors have been
366 called. The filename will be passed as the only argument.
367 progress_hooks: A list of functions that get called on download
368 progress, with a dictionary with the entries
369 * status: One of "downloading", "error", or "finished".
370 Check this first and ignore unknown values.
371 * info_dict: The extracted info_dict
372
373 If status is one of "downloading", or "finished", the
374 following properties may also be present:
375 * filename: The final filename (always present)
376 * tmpfilename: The filename we're currently writing to
377 * downloaded_bytes: Bytes on disk
378 * total_bytes: Size of the whole file, None if unknown
379 * total_bytes_estimate: Guess of the eventual file size,
380 None if unavailable.
381 * elapsed: The number of seconds since download started.
382 * eta: The estimated time in seconds, None if unknown
383 * speed: The download speed in bytes/second, None if
384 unknown
385 * fragment_index: The counter of the currently
386 downloaded video fragment.
387 * fragment_count: The number of fragments (= individual
388 files that will be merged)
389
390 Progress hooks are guaranteed to be called at least once
391 (with status "finished") if the download is successful.
392 postprocessor_hooks: A list of functions that get called on postprocessing
393 progress, with a dictionary with the entries
394 * status: One of "started", "processing", or "finished".
395 Check this first and ignore unknown values.
396 * postprocessor: Name of the postprocessor
397 * info_dict: The extracted info_dict
398
399 Progress hooks are guaranteed to be called at least twice
400 (with status "started" and "finished") if the processing is successful.
401 merge_output_format: Extension to use when merging formats.
402 final_ext: Expected final extension; used to detect when the file was
403 already downloaded and converted
404 fixup: Automatically correct known faults of the file.
405 One of:
406 - "never": do nothing
407 - "warn": only emit a warning
408 - "detect_or_warn": check whether we can do anything
409 about it, warn otherwise (default)
410 source_address: Client-side IP address to bind to.
411 call_home: Boolean, true iff we are allowed to contact the
412 yt-dlp servers for debugging. (BROKEN)
413 sleep_interval_requests: Number of seconds to sleep between requests
414 during extraction
415 sleep_interval: Number of seconds to sleep before each download when
416 used alone or a lower bound of a range for randomized
417 sleep before each download (minimum possible number
418 of seconds to sleep) when used along with
419 max_sleep_interval.
420 max_sleep_interval:Upper bound of a range for randomized sleep before each
421 download (maximum possible number of seconds to sleep).
422 Must only be used along with sleep_interval.
423 Actual sleep time will be a random float from range
424 [sleep_interval; max_sleep_interval].
425 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
426 listformats: Print an overview of available video formats and exit.
427 list_thumbnails: Print a table of all thumbnails and exit.
428 match_filter: A function that gets called with the info_dict of
429 every video.
430 If it returns a message, the video is ignored.
431 If it returns None, the video is downloaded.
432 match_filter_func in utils.py is one example for this.
433 no_color: Do not emit color codes in output.
434 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
435 HTTP header
436 geo_bypass_country:
437 Two-letter ISO 3166-2 country code that will be used for
438 explicit geographic restriction bypassing via faking
439 X-Forwarded-For HTTP header
440 geo_bypass_ip_block:
441 IP range in CIDR notation that will be used similarly to
442 geo_bypass_country
443
444 The following options determine which downloader is picked:
445 external_downloader: A dictionary of protocol keys and the executable of the
446 external downloader to use for it. The allowed protocols
447 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
448 Set the value to 'native' to use the native downloader
449 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
450 or {'m3u8': 'ffmpeg'} instead.
451 Use the native HLS downloader instead of ffmpeg/avconv
452 if True, otherwise use ffmpeg/avconv if False, otherwise
453 use downloader suggested by extractor if None.
454 compat_opts: Compatibility options. See "Differences in default behavior".
455 The following options do not work when used through the API:
456 filename, abort-on-error, multistreams, no-live-chat, format-sort
457 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
458 Refer __init__.py for their implementation
459 progress_template: Dictionary of templates for progress outputs.
460 Allowed keys are 'download', 'postprocess',
461 'download-title' (console title) and 'postprocess-title'.
462 The template is mapped on a dictionary with keys 'progress' and 'info'
463
464 The following parameters are not used by YoutubeDL itself, they are used by
465 the downloader (see yt_dlp/downloader/common.py):
466 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
467 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
468 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
469 external_downloader_args, concurrent_fragment_downloads.
470
471 The following options are used by the post processors:
472 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
473 otherwise prefer ffmpeg. (avconv support is deprecated)
474 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
475 to the binary or its containing directory.
476 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
477 and a list of additional command-line arguments for the
478 postprocessor/executable. The dict can also have "PP+EXE" keys
479 which are used when the given exe is used by the given PP.
480 Use 'default' as the name for arguments to passed to all PP
481 For compatibility with youtube-dl, a single list of args
482 can also be used
483
484 The following options are used by the extractors:
485 extractor_retries: Number of times to retry for known errors
486 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
487 hls_split_discontinuity: Split HLS playlists to different formats at
488 discontinuities such as ad breaks (default: False)
489 extractor_args: A dictionary of arguments to be passed to the extractors.
490 See "EXTRACTOR ARGUMENTS" for details.
491 Eg: {'youtube': {'skip': ['dash', 'hls']}}
492 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
493 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
494 If True (default), DASH manifests and related
495 data will be downloaded and processed by extractor.
496 You can reduce network I/O by disabling it if you don't
497 care about DASH. (only for youtube)
498 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
499 If True (default), HLS manifests and related
500 data will be downloaded and processed by extractor.
501 You can reduce network I/O by disabling it if you don't
502 care about HLS. (only for youtube)
503 """
504
505 _NUMERIC_FIELDS = set((
506 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
507 'timestamp', 'release_timestamp',
508 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
509 'average_rating', 'comment_count', 'age_limit',
510 'start_time', 'end_time',
511 'chapter_number', 'season_number', 'episode_number',
512 'track_number', 'disc_number', 'release_year',
513 ))
514
515 _format_selection_exts = {
516 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
517 'video': {'mp4', 'flv', 'webm', '3gp'},
518 'storyboards': {'mhtml'},
519 }
520
521 params = None
522 _ies = {}
523 _pps = {k: [] for k in POSTPROCESS_WHEN}
524 _printed_messages = set()
525 _first_webpage_request = True
526 _download_retcode = None
527 _num_downloads = None
528 _playlist_level = 0
529 _playlist_urls = set()
530 _screen_file = None
531
532 def __init__(self, params=None, auto_init=True):
533 """Create a FileDownloader object with the given options.
534 @param auto_init Whether to load the default extractors and print header (if verbose).
535 Set to 'no_verbose_header' to not print the header
536 """
537 if params is None:
538 params = {}
539 self._ies = {}
540 self._ies_instances = {}
541 self._pps = {k: [] for k in POSTPROCESS_WHEN}
542 self._printed_messages = set()
543 self._first_webpage_request = True
544 self._post_hooks = []
545 self._progress_hooks = []
546 self._postprocessor_hooks = []
547 self._download_retcode = 0
548 self._num_downloads = 0
549 self._num_videos = 0
550 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
551 self._err_file = sys.stderr
552 self.params = params
553 self.cache = Cache(self)
554
555 windows_enable_vt_mode()
556 self._allow_colors = {
557 'screen': not self.params.get('no_color') and supports_terminal_sequences(self._screen_file),
558 'err': not self.params.get('no_color') and supports_terminal_sequences(self._err_file),
559 }
560
561 if sys.version_info < (3, 6):
562 self.report_warning(
563 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
564
565 if self.params.get('allow_unplayable_formats'):
566 self.report_warning(
567 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
568 'This is a developer option intended for debugging. \n'
569 ' If you experience any issues while using this option, '
570 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
571
572 def check_deprecated(param, option, suggestion):
573 if self.params.get(param) is not None:
574 self.report_warning('%s is deprecated. Use %s instead' % (option, suggestion))
575 return True
576 return False
577
578 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
579 if self.params.get('geo_verification_proxy') is None:
580 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
581
582 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
583 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
584 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
585
586 for msg in self.params.get('_warnings', []):
587 self.report_warning(msg)
588 for msg in self.params.get('_deprecation_warnings', []):
589 self.deprecation_warning(msg)
590
591 if 'list-formats' in self.params.get('compat_opts', []):
592 self.params['listformats_table'] = False
593
594 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
595 # nooverwrites was unnecessarily changed to overwrites
596 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
597 # This ensures compatibility with both keys
598 self.params['overwrites'] = not self.params['nooverwrites']
599 elif self.params.get('overwrites') is None:
600 self.params.pop('overwrites', None)
601 else:
602 self.params['nooverwrites'] = not self.params['overwrites']
603
604 self.params.setdefault('forceprint', {})
605 self.params.setdefault('print_to_file', {})
606
607 # Compatibility with older syntax
608 if not isinstance(params['forceprint'], dict):
609 self.params['forceprint'] = {'video': params['forceprint']}
610
611 if self.params.get('bidi_workaround', False):
612 try:
613 import pty
614 master, slave = pty.openpty()
615 width = compat_get_terminal_size().columns
616 if width is None:
617 width_args = []
618 else:
619 width_args = ['-w', str(width)]
620 sp_kwargs = dict(
621 stdin=subprocess.PIPE,
622 stdout=slave,
623 stderr=self._err_file)
624 try:
625 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
626 except OSError:
627 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
628 self._output_channel = os.fdopen(master, 'rb')
629 except OSError as ose:
630 if ose.errno == errno.ENOENT:
631 self.report_warning(
632 'Could not find fribidi executable, ignoring --bidi-workaround. '
633 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
634 else:
635 raise
636
637 if (sys.platform != 'win32'
638 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
639 and not self.params.get('restrictfilenames', False)):
640 # Unicode filesystem API will throw errors (#1474, #13027)
641 self.report_warning(
642 'Assuming --restrict-filenames since file system encoding '
643 'cannot encode all characters. '
644 'Set the LC_ALL environment variable to fix this.')
645 self.params['restrictfilenames'] = True
646
647 self.outtmpl_dict = self.parse_outtmpl()
648
649 # Creating format selector here allows us to catch syntax errors before the extraction
650 self.format_selector = (
651 self.params.get('format') if self.params.get('format') in (None, '-')
652 else self.params['format'] if callable(self.params['format'])
653 else self.build_format_selector(self.params['format']))
654
655 # Set http_headers defaults according to std_headers
656 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
657
658 self._setup_opener()
659
660 if auto_init:
661 if auto_init != 'no_verbose_header':
662 self.print_debug_header()
663 self.add_default_info_extractors()
664
665 hooks = {
666 'post_hooks': self.add_post_hook,
667 'progress_hooks': self.add_progress_hook,
668 'postprocessor_hooks': self.add_postprocessor_hook,
669 }
670 for opt, fn in hooks.items():
671 for ph in self.params.get(opt, []):
672 fn(ph)
673
674 for pp_def_raw in self.params.get('postprocessors', []):
675 pp_def = dict(pp_def_raw)
676 when = pp_def.pop('when', 'post_process')
677 self.add_post_processor(
678 get_postprocessor(pp_def.pop('key'))(self, **compat_kwargs(pp_def)),
679 when=when)
680
681 register_socks_protocols()
682
683 def preload_download_archive(fn):
684 """Preload the archive, if any is specified"""
685 if fn is None:
686 return False
687 self.write_debug(f'Loading archive file {fn!r}')
688 try:
689 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
690 for line in archive_file:
691 self.archive.add(line.strip())
692 except IOError as ioe:
693 if ioe.errno != errno.ENOENT:
694 raise
695 return False
696 return True
697
698 self.archive = set()
699 preload_download_archive(self.params.get('download_archive'))
700
701 def warn_if_short_id(self, argv):
702 # short YouTube ID starting with dash?
703 idxs = [
704 i for i, a in enumerate(argv)
705 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
706 if idxs:
707 correct_argv = (
708 ['yt-dlp']
709 + [a for i, a in enumerate(argv) if i not in idxs]
710 + ['--'] + [argv[i] for i in idxs]
711 )
712 self.report_warning(
713 'Long argument string detected. '
714 'Use -- to separate parameters and URLs, like this:\n%s' %
715 args_to_str(correct_argv))
716
717 def add_info_extractor(self, ie):
718 """Add an InfoExtractor object to the end of the list."""
719 ie_key = ie.ie_key()
720 self._ies[ie_key] = ie
721 if not isinstance(ie, type):
722 self._ies_instances[ie_key] = ie
723 ie.set_downloader(self)
724
725 def _get_info_extractor_class(self, ie_key):
726 ie = self._ies.get(ie_key)
727 if ie is None:
728 ie = get_info_extractor(ie_key)
729 self.add_info_extractor(ie)
730 return ie
731
732 def get_info_extractor(self, ie_key):
733 """
734 Get an instance of an IE with name ie_key, it will try to get one from
735 the _ies list, if there's no instance it will create a new one and add
736 it to the extractor list.
737 """
738 ie = self._ies_instances.get(ie_key)
739 if ie is None:
740 ie = get_info_extractor(ie_key)()
741 self.add_info_extractor(ie)
742 return ie
743
744 def add_default_info_extractors(self):
745 """
746 Add the InfoExtractors returned by gen_extractors to the end of the list
747 """
748 for ie in gen_extractor_classes():
749 self.add_info_extractor(ie)
750
751 def add_post_processor(self, pp, when='post_process'):
752 """Add a PostProcessor object to the end of the chain."""
753 self._pps[when].append(pp)
754 pp.set_downloader(self)
755
756 def add_post_hook(self, ph):
757 """Add the post hook"""
758 self._post_hooks.append(ph)
759
760 def add_progress_hook(self, ph):
761 """Add the download progress hook"""
762 self._progress_hooks.append(ph)
763
764 def add_postprocessor_hook(self, ph):
765 """Add the postprocessing progress hook"""
766 self._postprocessor_hooks.append(ph)
767 for pps in self._pps.values():
768 for pp in pps:
769 pp.add_progress_hook(ph)
770
771 def _bidi_workaround(self, message):
772 if not hasattr(self, '_output_channel'):
773 return message
774
775 assert hasattr(self, '_output_process')
776 assert isinstance(message, compat_str)
777 line_count = message.count('\n') + 1
778 self._output_process.stdin.write((message + '\n').encode('utf-8'))
779 self._output_process.stdin.flush()
780 res = ''.join(self._output_channel.readline().decode('utf-8')
781 for _ in range(line_count))
782 return res[:-len('\n')]
783
784 def _write_string(self, message, out=None, only_once=False):
785 if only_once:
786 if message in self._printed_messages:
787 return
788 self._printed_messages.add(message)
789 write_string(message, out=out, encoding=self.params.get('encoding'))
790
791 def to_stdout(self, message, skip_eol=False, quiet=False):
792 """Print message to stdout"""
793 if self.params.get('logger'):
794 self.params['logger'].debug(message)
795 elif not quiet or self.params.get('verbose'):
796 self._write_string(
797 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
798 self._err_file if quiet else self._screen_file)
799
800 def to_stderr(self, message, only_once=False):
801 """Print message to stderr"""
802 assert isinstance(message, compat_str)
803 if self.params.get('logger'):
804 self.params['logger'].error(message)
805 else:
806 self._write_string('%s\n' % self._bidi_workaround(message), self._err_file, only_once=only_once)
807
808 def to_console_title(self, message):
809 if not self.params.get('consoletitle', False):
810 return
811 message = remove_terminal_sequences(message)
812 if compat_os_name == 'nt':
813 if ctypes.windll.kernel32.GetConsoleWindow():
814 # c_wchar_p() might not be necessary if `message` is
815 # already of type unicode()
816 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
817 elif 'TERM' in os.environ:
818 self._write_string('\033]0;%s\007' % message, self._screen_file)
819
820 def save_console_title(self):
821 if not self.params.get('consoletitle', False):
822 return
823 if self.params.get('simulate'):
824 return
825 if compat_os_name != 'nt' and 'TERM' in os.environ:
826 # Save the title on stack
827 self._write_string('\033[22;0t', self._screen_file)
828
829 def restore_console_title(self):
830 if not self.params.get('consoletitle', False):
831 return
832 if self.params.get('simulate'):
833 return
834 if compat_os_name != 'nt' and 'TERM' in os.environ:
835 # Restore the title from stack
836 self._write_string('\033[23;0t', self._screen_file)
837
838 def __enter__(self):
839 self.save_console_title()
840 return self
841
842 def __exit__(self, *args):
843 self.restore_console_title()
844
845 if self.params.get('cookiefile') is not None:
846 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
847
848 def trouble(self, message=None, tb=None, is_error=True):
849 """Determine action to take when a download problem appears.
850
851 Depending on if the downloader has been configured to ignore
852 download errors or not, this method may throw an exception or
853 not when errors are found, after printing the message.
854
855 @param tb If given, is additional traceback information
856 @param is_error Whether to raise error according to ignorerrors
857 """
858 if message is not None:
859 self.to_stderr(message)
860 if self.params.get('verbose'):
861 if tb is None:
862 if sys.exc_info()[0]: # if .trouble has been called from an except block
863 tb = ''
864 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
865 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
866 tb += encode_compat_str(traceback.format_exc())
867 else:
868 tb_data = traceback.format_list(traceback.extract_stack())
869 tb = ''.join(tb_data)
870 if tb:
871 self.to_stderr(tb)
872 if not is_error:
873 return
874 if not self.params.get('ignoreerrors'):
875 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
876 exc_info = sys.exc_info()[1].exc_info
877 else:
878 exc_info = sys.exc_info()
879 raise DownloadError(message, exc_info)
880 self._download_retcode = 1
881
882 def to_screen(self, message, skip_eol=False):
883 """Print message to stdout if not in quiet mode"""
884 self.to_stdout(
885 message, skip_eol, quiet=self.params.get('quiet', False))
886
887 class Styles(Enum):
888 HEADERS = 'yellow'
889 EMPHASIS = 'light blue'
890 ID = 'green'
891 DELIM = 'blue'
892 ERROR = 'red'
893 WARNING = 'yellow'
894 SUPPRESS = 'light black'
895
896 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
897 if test_encoding:
898 original_text = text
899 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
900 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
901 text = text.encode(encoding, 'ignore').decode(encoding)
902 if fallback is not None and text != original_text:
903 text = fallback
904 if isinstance(f, self.Styles):
905 f = f.value
906 return format_text(text, f) if allow_colors else text if fallback is None else fallback
907
908 def _format_screen(self, *args, **kwargs):
909 return self._format_text(
910 self._screen_file, self._allow_colors['screen'], *args, **kwargs)
911
912 def _format_err(self, *args, **kwargs):
913 return self._format_text(
914 self._err_file, self._allow_colors['err'], *args, **kwargs)
915
916 def report_warning(self, message, only_once=False):
917 '''
918 Print the message to stderr, it will be prefixed with 'WARNING:'
919 If stderr is a tty file the 'WARNING:' will be colored
920 '''
921 if self.params.get('logger') is not None:
922 self.params['logger'].warning(message)
923 else:
924 if self.params.get('no_warnings'):
925 return
926 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
927
928 def deprecation_warning(self, message):
929 if self.params.get('logger') is not None:
930 self.params['logger'].warning('DeprecationWarning: {message}')
931 else:
932 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
933
934 def report_error(self, message, *args, **kwargs):
935 '''
936 Do the same as trouble, but prefixes the message with 'ERROR:', colored
937 in red if stderr is a tty file.
938 '''
939 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
940
941 def write_debug(self, message, only_once=False):
942 '''Log debug message or Print message to stderr'''
943 if not self.params.get('verbose', False):
944 return
945 message = '[debug] %s' % message
946 if self.params.get('logger'):
947 self.params['logger'].debug(message)
948 else:
949 self.to_stderr(message, only_once)
950
951 def report_file_already_downloaded(self, file_name):
952 """Report file has already been fully downloaded."""
953 try:
954 self.to_screen('[download] %s has already been downloaded' % file_name)
955 except UnicodeEncodeError:
956 self.to_screen('[download] The file has already been downloaded')
957
958 def report_file_delete(self, file_name):
959 """Report that existing file will be deleted."""
960 try:
961 self.to_screen('Deleting existing file %s' % file_name)
962 except UnicodeEncodeError:
963 self.to_screen('Deleting existing file')
964
965 def raise_no_formats(self, info, forced=False, *, msg=None):
966 has_drm = info.get('__has_drm')
967 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
968 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
969 if forced or not ignored:
970 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
971 expected=has_drm or ignored or expected)
972 else:
973 self.report_warning(msg)
974
975 def parse_outtmpl(self):
976 outtmpl_dict = self.params.get('outtmpl', {})
977 if not isinstance(outtmpl_dict, dict):
978 outtmpl_dict = {'default': outtmpl_dict}
979 # Remove spaces in the default template
980 if self.params.get('restrictfilenames'):
981 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
982 else:
983 sanitize = lambda x: x
984 outtmpl_dict.update({
985 k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items()
986 if outtmpl_dict.get(k) is None})
987 for key, val in outtmpl_dict.items():
988 if isinstance(val, bytes):
989 self.report_warning(
990 'Parameter outtmpl is bytes, but should be a unicode string. '
991 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
992 return outtmpl_dict
993
994 def get_output_path(self, dir_type='', filename=None):
995 paths = self.params.get('paths', {})
996 assert isinstance(paths, dict)
997 path = os.path.join(
998 expand_path(paths.get('home', '').strip()),
999 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1000 filename or '')
1001
1002 # Temporary fix for #4787
1003 # 'Treat' all problem characters by passing filename through preferredencoding
1004 # to workaround encoding issues with subprocess on python2 @ Windows
1005 if sys.version_info < (3, 0) and sys.platform == 'win32':
1006 path = encodeFilename(path, True).decode(preferredencoding())
1007 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1008
1009 @staticmethod
1010 def _outtmpl_expandpath(outtmpl):
1011 # expand_path translates '%%' into '%' and '$$' into '$'
1012 # correspondingly that is not what we want since we need to keep
1013 # '%%' intact for template dict substitution step. Working around
1014 # with boundary-alike separator hack.
1015 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1016 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
1017
1018 # outtmpl should be expand_path'ed before template dict substitution
1019 # because meta fields may contain env variables we don't want to
1020 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1021 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1022 return expand_path(outtmpl).replace(sep, '')
1023
1024 @staticmethod
1025 def escape_outtmpl(outtmpl):
1026 ''' Escape any remaining strings like %s, %abc% etc. '''
1027 return re.sub(
1028 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1029 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1030 outtmpl)
1031
1032 @classmethod
1033 def validate_outtmpl(cls, outtmpl):
1034 ''' @return None or Exception object '''
1035 outtmpl = re.sub(
1036 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1037 lambda mobj: f'{mobj.group(0)[:-1]}s',
1038 cls._outtmpl_expandpath(outtmpl))
1039 try:
1040 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1041 return None
1042 except ValueError as err:
1043 return err
1044
1045 @staticmethod
1046 def _copy_infodict(info_dict):
1047 info_dict = dict(info_dict)
1048 info_dict.pop('__postprocessors', None)
1049 return info_dict
1050
1051 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1052 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1053 @param sanitize Whether to sanitize the output as a filename.
1054 For backward compatibility, a function can also be passed
1055 """
1056
1057 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1058
1059 info_dict = self._copy_infodict(info_dict)
1060 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1061 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1062 if info_dict.get('duration', None) is not None
1063 else None)
1064 info_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
1065 info_dict['video_autonumber'] = self._num_videos
1066 if info_dict.get('resolution') is None:
1067 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1068
1069 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1070 # of %(field)s to %(field)0Nd for backward compatibility
1071 field_size_compat_map = {
1072 'playlist_index': number_of_digits(info_dict.get('_last_playlist_index') or 0),
1073 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1074 'autonumber': self.params.get('autonumber_size') or 5,
1075 }
1076
1077 TMPL_DICT = {}
1078 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1079 MATH_FUNCTIONS = {
1080 '+': float.__add__,
1081 '-': float.__sub__,
1082 }
1083 # Field is of the form key1.key2...
1084 # where keys (except first) can be string, int or slice
1085 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1086 MATH_FIELD_RE = r'''(?:{field}|{num})'''.format(field=FIELD_RE, num=r'-?\d+(?:.\d+)?')
1087 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1088 INTERNAL_FORMAT_RE = re.compile(r'''(?x)
1089 (?P<negate>-)?
1090 (?P<fields>{field})
1091 (?P<maths>(?:{math_op}{math_field})*)
1092 (?:>(?P<strf_format>.+?))?
1093 (?P<alternate>(?<!\\),[^|&)]+)?
1094 (?:&(?P<replacement>.*?))?
1095 (?:\|(?P<default>.*?))?
1096 $'''.format(field=FIELD_RE, math_op=MATH_OPERATORS_RE, math_field=MATH_FIELD_RE))
1097
1098 def _traverse_infodict(k):
1099 k = k.split('.')
1100 if k[0] == '':
1101 k.pop(0)
1102 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1103
1104 def get_value(mdict):
1105 # Object traversal
1106 value = _traverse_infodict(mdict['fields'])
1107 # Negative
1108 if mdict['negate']:
1109 value = float_or_none(value)
1110 if value is not None:
1111 value *= -1
1112 # Do maths
1113 offset_key = mdict['maths']
1114 if offset_key:
1115 value = float_or_none(value)
1116 operator = None
1117 while offset_key:
1118 item = re.match(
1119 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1120 offset_key).group(0)
1121 offset_key = offset_key[len(item):]
1122 if operator is None:
1123 operator = MATH_FUNCTIONS[item]
1124 continue
1125 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1126 offset = float_or_none(item)
1127 if offset is None:
1128 offset = float_or_none(_traverse_infodict(item))
1129 try:
1130 value = operator(value, multiplier * offset)
1131 except (TypeError, ZeroDivisionError):
1132 return None
1133 operator = None
1134 # Datetime formatting
1135 if mdict['strf_format']:
1136 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1137
1138 return value
1139
1140 na = self.params.get('outtmpl_na_placeholder', 'NA')
1141
1142 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1143 return sanitize_filename(str(value), restricted=restricted,
1144 is_id=re.search(r'(^|[_.])id(\.|$)', key))
1145
1146 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1147 sanitize = bool(sanitize)
1148
1149 def _dumpjson_default(obj):
1150 if isinstance(obj, (set, LazyList)):
1151 return list(obj)
1152 return repr(obj)
1153
1154 def create_key(outer_mobj):
1155 if not outer_mobj.group('has_key'):
1156 return outer_mobj.group(0)
1157 key = outer_mobj.group('key')
1158 mobj = re.match(INTERNAL_FORMAT_RE, key)
1159 initial_field = mobj.group('fields') if mobj else ''
1160 value, replacement, default = None, None, na
1161 while mobj:
1162 mobj = mobj.groupdict()
1163 default = mobj['default'] if mobj['default'] is not None else default
1164 value = get_value(mobj)
1165 replacement = mobj['replacement']
1166 if value is None and mobj['alternate']:
1167 mobj = re.match(INTERNAL_FORMAT_RE, mobj['alternate'][1:])
1168 else:
1169 break
1170
1171 fmt = outer_mobj.group('format')
1172 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1173 fmt = '0{:d}d'.format(field_size_compat_map[key])
1174
1175 value = default if value is None else value if replacement is None else replacement
1176
1177 flags = outer_mobj.group('conversion') or ''
1178 str_fmt = f'{fmt[:-1]}s'
1179 if fmt[-1] == 'l': # list
1180 delim = '\n' if '#' in flags else ', '
1181 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1182 elif fmt[-1] == 'j': # json
1183 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1184 elif fmt[-1] == 'q': # quoted
1185 value = map(str, variadic(value) if '#' in flags else [value])
1186 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1187 elif fmt[-1] == 'B': # bytes
1188 value = f'%{str_fmt}'.encode('utf-8') % str(value).encode('utf-8')
1189 value, fmt = value.decode('utf-8', 'ignore'), 's'
1190 elif fmt[-1] == 'U': # unicode normalized
1191 value, fmt = unicodedata.normalize(
1192 # "+" = compatibility equivalence, "#" = NFD
1193 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1194 value), str_fmt
1195 elif fmt[-1] == 'D': # decimal suffix
1196 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1197 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1198 factor=1024 if '#' in flags else 1000)
1199 elif fmt[-1] == 'S': # filename sanitization
1200 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1201 elif fmt[-1] == 'c':
1202 if value:
1203 value = str(value)[0]
1204 else:
1205 fmt = str_fmt
1206 elif fmt[-1] not in 'rs': # numeric
1207 value = float_or_none(value)
1208 if value is None:
1209 value, fmt = default, 's'
1210
1211 if sanitize:
1212 if fmt[-1] == 'r':
1213 # If value is an object, sanitize might convert it to a string
1214 # So we convert it to repr first
1215 value, fmt = repr(value), str_fmt
1216 if fmt[-1] in 'csr':
1217 value = sanitizer(initial_field, value)
1218
1219 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1220 TMPL_DICT[key] = value
1221 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1222
1223 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1224
1225 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1226 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1227 return self.escape_outtmpl(outtmpl) % info_dict
1228
1229 def _prepare_filename(self, info_dict, tmpl_type='default'):
1230 try:
1231 outtmpl = self._outtmpl_expandpath(self.outtmpl_dict.get(tmpl_type, self.outtmpl_dict['default']))
1232 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1233 if not filename:
1234 return None
1235
1236 if tmpl_type in ('default', 'temp'):
1237 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1238 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1239 filename = replace_extension(filename, ext, final_ext)
1240 else:
1241 force_ext = OUTTMPL_TYPES[tmpl_type]
1242 if force_ext:
1243 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1244
1245 # https://github.com/blackjack4494/youtube-dlc/issues/85
1246 trim_file_name = self.params.get('trim_file_name', False)
1247 if trim_file_name:
1248 no_ext, *ext = filename.rsplit('.', 2)
1249 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1250
1251 return filename
1252 except ValueError as err:
1253 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1254 return None
1255
1256 def prepare_filename(self, info_dict, dir_type='', warn=False):
1257 """Generate the output filename."""
1258
1259 filename = self._prepare_filename(info_dict, dir_type or 'default')
1260 if not filename and dir_type not in ('', 'temp'):
1261 return ''
1262
1263 if warn:
1264 if not self.params.get('paths'):
1265 pass
1266 elif filename == '-':
1267 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1268 elif os.path.isabs(filename):
1269 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1270 if filename == '-' or not filename:
1271 return filename
1272
1273 return self.get_output_path(dir_type, filename)
1274
1275 def _match_entry(self, info_dict, incomplete=False, silent=False):
1276 """ Returns None if the file should be downloaded """
1277
1278 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1279
1280 def check_filter():
1281 if 'title' in info_dict:
1282 # This can happen when we're just evaluating the playlist
1283 title = info_dict['title']
1284 matchtitle = self.params.get('matchtitle', False)
1285 if matchtitle:
1286 if not re.search(matchtitle, title, re.IGNORECASE):
1287 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1288 rejecttitle = self.params.get('rejecttitle', False)
1289 if rejecttitle:
1290 if re.search(rejecttitle, title, re.IGNORECASE):
1291 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1292 date = info_dict.get('upload_date')
1293 if date is not None:
1294 dateRange = self.params.get('daterange', DateRange())
1295 if date not in dateRange:
1296 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
1297 view_count = info_dict.get('view_count')
1298 if view_count is not None:
1299 min_views = self.params.get('min_views')
1300 if min_views is not None and view_count < min_views:
1301 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1302 max_views = self.params.get('max_views')
1303 if max_views is not None and view_count > max_views:
1304 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1305 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1306 return 'Skipping "%s" because it is age restricted' % video_title
1307
1308 match_filter = self.params.get('match_filter')
1309 if match_filter is not None:
1310 try:
1311 ret = match_filter(info_dict, incomplete=incomplete)
1312 except TypeError:
1313 # For backward compatibility
1314 ret = None if incomplete else match_filter(info_dict)
1315 if ret is not None:
1316 return ret
1317 return None
1318
1319 if self.in_download_archive(info_dict):
1320 reason = '%s has already been recorded in the archive' % video_title
1321 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1322 else:
1323 reason = check_filter()
1324 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1325 if reason is not None:
1326 if not silent:
1327 self.to_screen('[download] ' + reason)
1328 if self.params.get(break_opt, False):
1329 raise break_err()
1330 return reason
1331
1332 @staticmethod
1333 def add_extra_info(info_dict, extra_info):
1334 '''Set the keys from extra_info in info dict if they are missing'''
1335 for key, value in extra_info.items():
1336 info_dict.setdefault(key, value)
1337
1338 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1339 process=True, force_generic_extractor=False):
1340 """
1341 Return a list with a dictionary for each video extracted.
1342
1343 Arguments:
1344 url -- URL to extract
1345
1346 Keyword arguments:
1347 download -- whether to download videos during extraction
1348 ie_key -- extractor key hint
1349 extra_info -- dictionary containing the extra values to add to each result
1350 process -- whether to resolve all unresolved references (URLs, playlist items),
1351 must be True for download to work.
1352 force_generic_extractor -- force using the generic extractor
1353 """
1354
1355 if extra_info is None:
1356 extra_info = {}
1357
1358 if not ie_key and force_generic_extractor:
1359 ie_key = 'Generic'
1360
1361 if ie_key:
1362 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1363 else:
1364 ies = self._ies
1365
1366 for ie_key, ie in ies.items():
1367 if not ie.suitable(url):
1368 continue
1369
1370 if not ie.working():
1371 self.report_warning('The program functionality for this site has been marked as broken, '
1372 'and will probably not work.')
1373
1374 temp_id = ie.get_temp_id(url)
1375 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1376 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1377 if self.params.get('break_on_existing', False):
1378 raise ExistingVideoReached()
1379 break
1380 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1381 else:
1382 self.report_error('no suitable InfoExtractor for URL %s' % url)
1383
1384 def __handle_extraction_exceptions(func):
1385 @functools.wraps(func)
1386 def wrapper(self, *args, **kwargs):
1387 while True:
1388 try:
1389 return func(self, *args, **kwargs)
1390 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1391 raise
1392 except ReExtractInfo as e:
1393 if e.expected:
1394 self.to_screen(f'{e}; Re-extracting data')
1395 else:
1396 self.to_stderr('\r')
1397 self.report_warning(f'{e}; Re-extracting data')
1398 continue
1399 except GeoRestrictedError as e:
1400 msg = e.msg
1401 if e.countries:
1402 msg += '\nThis video is available in %s.' % ', '.join(
1403 map(ISO3166Utils.short2full, e.countries))
1404 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1405 self.report_error(msg)
1406 except ExtractorError as e: # An error we somewhat expected
1407 self.report_error(str(e), e.format_traceback())
1408 except Exception as e:
1409 if self.params.get('ignoreerrors'):
1410 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1411 else:
1412 raise
1413 break
1414 return wrapper
1415
1416 def _wait_for_video(self, ie_result):
1417 if (not self.params.get('wait_for_video')
1418 or ie_result.get('_type', 'video') != 'video'
1419 or ie_result.get('formats') or ie_result.get('url')):
1420 return
1421
1422 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1423 last_msg = ''
1424
1425 def progress(msg):
1426 nonlocal last_msg
1427 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1428 last_msg = msg
1429
1430 min_wait, max_wait = self.params.get('wait_for_video')
1431 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1432 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1433 diff = random.randrange(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait)
1434 self.report_warning('Release time of video is not known')
1435 elif (diff or 0) <= 0:
1436 self.report_warning('Video should already be available according to extracted info')
1437 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1438 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1439
1440 wait_till = time.time() + diff
1441 try:
1442 while True:
1443 diff = wait_till - time.time()
1444 if diff <= 0:
1445 progress('')
1446 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1447 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1448 time.sleep(1)
1449 except KeyboardInterrupt:
1450 progress('')
1451 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1452 except BaseException as e:
1453 if not isinstance(e, ReExtractInfo):
1454 self.to_screen('')
1455 raise
1456
1457 @__handle_extraction_exceptions
1458 def __extract_info(self, url, ie, download, extra_info, process):
1459 ie_result = ie.extract(url)
1460 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1461 return
1462 if isinstance(ie_result, list):
1463 # Backwards compatibility: old IE result format
1464 ie_result = {
1465 '_type': 'compat_list',
1466 'entries': ie_result,
1467 }
1468 if extra_info.get('original_url'):
1469 ie_result.setdefault('original_url', extra_info['original_url'])
1470 self.add_default_extra_info(ie_result, ie, url)
1471 if process:
1472 self._wait_for_video(ie_result)
1473 return self.process_ie_result(ie_result, download, extra_info)
1474 else:
1475 return ie_result
1476
1477 def add_default_extra_info(self, ie_result, ie, url):
1478 if url is not None:
1479 self.add_extra_info(ie_result, {
1480 'webpage_url': url,
1481 'original_url': url,
1482 })
1483 webpage_url = ie_result.get('webpage_url')
1484 if webpage_url:
1485 self.add_extra_info(ie_result, {
1486 'webpage_url_basename': url_basename(webpage_url),
1487 'webpage_url_domain': get_domain(webpage_url),
1488 })
1489 if ie is not None:
1490 self.add_extra_info(ie_result, {
1491 'extractor': ie.IE_NAME,
1492 'extractor_key': ie.ie_key(),
1493 })
1494
1495 def process_ie_result(self, ie_result, download=True, extra_info=None):
1496 """
1497 Take the result of the ie(may be modified) and resolve all unresolved
1498 references (URLs, playlist items).
1499
1500 It will also download the videos if 'download'.
1501 Returns the resolved ie_result.
1502 """
1503 if extra_info is None:
1504 extra_info = {}
1505 result_type = ie_result.get('_type', 'video')
1506
1507 if result_type in ('url', 'url_transparent'):
1508 ie_result['url'] = sanitize_url(ie_result['url'])
1509 if ie_result.get('original_url'):
1510 extra_info.setdefault('original_url', ie_result['original_url'])
1511
1512 extract_flat = self.params.get('extract_flat', False)
1513 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1514 or extract_flat is True):
1515 info_copy = ie_result.copy()
1516 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1517 if ie and not ie_result.get('id'):
1518 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1519 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1520 self.add_extra_info(info_copy, extra_info)
1521 info_copy, _ = self.pre_process(info_copy)
1522 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1523 if self.params.get('force_write_download_archive', False):
1524 self.record_download_archive(info_copy)
1525 return ie_result
1526
1527 if result_type == 'video':
1528 self.add_extra_info(ie_result, extra_info)
1529 ie_result = self.process_video_result(ie_result, download=download)
1530 additional_urls = (ie_result or {}).get('additional_urls')
1531 if additional_urls:
1532 # TODO: Improve MetadataParserPP to allow setting a list
1533 if isinstance(additional_urls, compat_str):
1534 additional_urls = [additional_urls]
1535 self.to_screen(
1536 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1537 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1538 ie_result['additional_entries'] = [
1539 self.extract_info(
1540 url, download, extra_info=extra_info,
1541 force_generic_extractor=self.params.get('force_generic_extractor'))
1542 for url in additional_urls
1543 ]
1544 return ie_result
1545 elif result_type == 'url':
1546 # We have to add extra_info to the results because it may be
1547 # contained in a playlist
1548 return self.extract_info(
1549 ie_result['url'], download,
1550 ie_key=ie_result.get('ie_key'),
1551 extra_info=extra_info)
1552 elif result_type == 'url_transparent':
1553 # Use the information from the embedding page
1554 info = self.extract_info(
1555 ie_result['url'], ie_key=ie_result.get('ie_key'),
1556 extra_info=extra_info, download=False, process=False)
1557
1558 # extract_info may return None when ignoreerrors is enabled and
1559 # extraction failed with an error, don't crash and return early
1560 # in this case
1561 if not info:
1562 return info
1563
1564 force_properties = dict(
1565 (k, v) for k, v in ie_result.items() if v is not None)
1566 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1567 if f in force_properties:
1568 del force_properties[f]
1569 new_result = info.copy()
1570 new_result.update(force_properties)
1571
1572 # Extracted info may not be a video result (i.e.
1573 # info.get('_type', 'video') != video) but rather an url or
1574 # url_transparent. In such cases outer metadata (from ie_result)
1575 # should be propagated to inner one (info). For this to happen
1576 # _type of info should be overridden with url_transparent. This
1577 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1578 if new_result.get('_type') == 'url':
1579 new_result['_type'] = 'url_transparent'
1580
1581 return self.process_ie_result(
1582 new_result, download=download, extra_info=extra_info)
1583 elif result_type in ('playlist', 'multi_video'):
1584 # Protect from infinite recursion due to recursively nested playlists
1585 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1586 webpage_url = ie_result['webpage_url']
1587 if webpage_url in self._playlist_urls:
1588 self.to_screen(
1589 '[download] Skipping already downloaded playlist: %s'
1590 % ie_result.get('title') or ie_result.get('id'))
1591 return
1592
1593 self._playlist_level += 1
1594 self._playlist_urls.add(webpage_url)
1595 self._fill_common_fields(ie_result, False)
1596 self._sanitize_thumbnails(ie_result)
1597 try:
1598 return self.__process_playlist(ie_result, download)
1599 finally:
1600 self._playlist_level -= 1
1601 if not self._playlist_level:
1602 self._playlist_urls.clear()
1603 elif result_type == 'compat_list':
1604 self.report_warning(
1605 'Extractor %s returned a compat_list result. '
1606 'It needs to be updated.' % ie_result.get('extractor'))
1607
1608 def _fixup(r):
1609 self.add_extra_info(r, {
1610 'extractor': ie_result['extractor'],
1611 'webpage_url': ie_result['webpage_url'],
1612 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1613 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1614 'extractor_key': ie_result['extractor_key'],
1615 })
1616 return r
1617 ie_result['entries'] = [
1618 self.process_ie_result(_fixup(r), download, extra_info)
1619 for r in ie_result['entries']
1620 ]
1621 return ie_result
1622 else:
1623 raise Exception('Invalid result type: %s' % result_type)
1624
1625 def _ensure_dir_exists(self, path):
1626 return make_dir(path, self.report_error)
1627
1628 @staticmethod
1629 def _playlist_infodict(ie_result, **kwargs):
1630 return {
1631 **ie_result,
1632 'playlist': ie_result.get('title') or ie_result.get('id'),
1633 'playlist_id': ie_result.get('id'),
1634 'playlist_title': ie_result.get('title'),
1635 'playlist_uploader': ie_result.get('uploader'),
1636 'playlist_uploader_id': ie_result.get('uploader_id'),
1637 'playlist_index': 0,
1638 **kwargs,
1639 }
1640
1641 def __process_playlist(self, ie_result, download):
1642 # We process each entry in the playlist
1643 playlist = ie_result.get('title') or ie_result.get('id')
1644 self.to_screen('[download] Downloading playlist: %s' % playlist)
1645
1646 if 'entries' not in ie_result:
1647 raise EntryNotInPlaylist('There are no entries')
1648
1649 MissingEntry = object()
1650 incomplete_entries = bool(ie_result.get('requested_entries'))
1651 if incomplete_entries:
1652 def fill_missing_entries(entries, indices):
1653 ret = [MissingEntry] * max(indices)
1654 for i, entry in zip(indices, entries):
1655 ret[i - 1] = entry
1656 return ret
1657 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1658
1659 playlist_results = []
1660
1661 playliststart = self.params.get('playliststart', 1)
1662 playlistend = self.params.get('playlistend')
1663 # For backwards compatibility, interpret -1 as whole list
1664 if playlistend == -1:
1665 playlistend = None
1666
1667 playlistitems_str = self.params.get('playlist_items')
1668 playlistitems = None
1669 if playlistitems_str is not None:
1670 def iter_playlistitems(format):
1671 for string_segment in format.split(','):
1672 if '-' in string_segment:
1673 start, end = string_segment.split('-')
1674 for item in range(int(start), int(end) + 1):
1675 yield int(item)
1676 else:
1677 yield int(string_segment)
1678 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1679
1680 ie_entries = ie_result['entries']
1681 if isinstance(ie_entries, list):
1682 playlist_count = len(ie_entries)
1683 msg = f'Collected {playlist_count} videos; downloading %d of them'
1684 ie_result['playlist_count'] = ie_result.get('playlist_count') or playlist_count
1685
1686 def get_entry(i):
1687 return ie_entries[i - 1]
1688 else:
1689 msg = 'Downloading %d videos'
1690 if not isinstance(ie_entries, (PagedList, LazyList)):
1691 ie_entries = LazyList(ie_entries)
1692 elif isinstance(ie_entries, InAdvancePagedList):
1693 if ie_entries._pagesize == 1:
1694 playlist_count = ie_entries._pagecount
1695
1696 def get_entry(i):
1697 return YoutubeDL.__handle_extraction_exceptions(
1698 lambda self, i: ie_entries[i - 1]
1699 )(self, i)
1700
1701 entries, broken = [], False
1702 items = playlistitems if playlistitems is not None else itertools.count(playliststart)
1703 for i in items:
1704 if i == 0:
1705 continue
1706 if playlistitems is None and playlistend is not None and playlistend < i:
1707 break
1708 entry = None
1709 try:
1710 entry = get_entry(i)
1711 if entry is MissingEntry:
1712 raise EntryNotInPlaylist()
1713 except (IndexError, EntryNotInPlaylist):
1714 if incomplete_entries:
1715 raise EntryNotInPlaylist(f'Entry {i} cannot be found')
1716 elif not playlistitems:
1717 break
1718 entries.append(entry)
1719 try:
1720 if entry is not None:
1721 self._match_entry(entry, incomplete=True, silent=True)
1722 except (ExistingVideoReached, RejectedVideoReached):
1723 broken = True
1724 break
1725 ie_result['entries'] = entries
1726
1727 # Save playlist_index before re-ordering
1728 entries = [
1729 ((playlistitems[i - 1] if playlistitems else i + playliststart - 1), entry)
1730 for i, entry in enumerate(entries, 1)
1731 if entry is not None]
1732 n_entries = len(entries)
1733
1734 if not (ie_result.get('playlist_count') or broken or playlistitems or playlistend):
1735 ie_result['playlist_count'] = n_entries
1736
1737 if not playlistitems and (playliststart != 1 or playlistend):
1738 playlistitems = list(range(playliststart, playliststart + n_entries))
1739 ie_result['requested_entries'] = playlistitems
1740
1741 _infojson_written = False
1742 write_playlist_files = self.params.get('allow_playlist_files', True)
1743 if write_playlist_files and self.params.get('list_thumbnails'):
1744 self.list_thumbnails(ie_result)
1745 if write_playlist_files and not self.params.get('simulate'):
1746 ie_copy = self._playlist_infodict(ie_result, n_entries=n_entries)
1747 _infojson_written = self._write_info_json(
1748 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1749 if _infojson_written is None:
1750 return
1751 if self._write_description('playlist', ie_result,
1752 self.prepare_filename(ie_copy, 'pl_description')) is None:
1753 return
1754 # TODO: This should be passed to ThumbnailsConvertor if necessary
1755 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1756
1757 if self.params.get('playlistreverse', False):
1758 entries = entries[::-1]
1759 if self.params.get('playlistrandom', False):
1760 random.shuffle(entries)
1761
1762 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1763
1764 self.to_screen('[%s] playlist %s: %s' % (ie_result['extractor'], playlist, msg % n_entries))
1765 failures = 0
1766 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1767 for i, entry_tuple in enumerate(entries, 1):
1768 playlist_index, entry = entry_tuple
1769 if 'playlist-index' in self.params.get('compat_opts', []):
1770 playlist_index = playlistitems[i - 1] if playlistitems else i + playliststart - 1
1771 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1772 # This __x_forwarded_for_ip thing is a bit ugly but requires
1773 # minimal changes
1774 if x_forwarded_for:
1775 entry['__x_forwarded_for_ip'] = x_forwarded_for
1776 extra = {
1777 'n_entries': n_entries,
1778 '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1779 'playlist_count': ie_result.get('playlist_count'),
1780 'playlist_index': playlist_index,
1781 'playlist_autonumber': i,
1782 'playlist': playlist,
1783 'playlist_id': ie_result.get('id'),
1784 'playlist_title': ie_result.get('title'),
1785 'playlist_uploader': ie_result.get('uploader'),
1786 'playlist_uploader_id': ie_result.get('uploader_id'),
1787 'extractor': ie_result['extractor'],
1788 'webpage_url': ie_result['webpage_url'],
1789 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1790 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1791 'extractor_key': ie_result['extractor_key'],
1792 }
1793
1794 if self._match_entry(entry, incomplete=True) is not None:
1795 continue
1796
1797 entry_result = self.__process_iterable_entry(entry, download, extra)
1798 if not entry_result:
1799 failures += 1
1800 if failures >= max_failures:
1801 self.report_error(
1802 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1803 break
1804 playlist_results.append(entry_result)
1805 ie_result['entries'] = playlist_results
1806
1807 # Write the updated info to json
1808 if _infojson_written and self._write_info_json(
1809 'updated playlist', ie_result,
1810 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1811 return
1812
1813 ie_result = self.run_all_pps('playlist', ie_result)
1814 self.to_screen(f'[download] Finished downloading playlist: {playlist}')
1815 return ie_result
1816
1817 @__handle_extraction_exceptions
1818 def __process_iterable_entry(self, entry, download, extra_info):
1819 return self.process_ie_result(
1820 entry, download=download, extra_info=extra_info)
1821
1822 def _build_format_filter(self, filter_spec):
1823 " Returns a function to filter the formats according to the filter_spec "
1824
1825 OPERATORS = {
1826 '<': operator.lt,
1827 '<=': operator.le,
1828 '>': operator.gt,
1829 '>=': operator.ge,
1830 '=': operator.eq,
1831 '!=': operator.ne,
1832 }
1833 operator_rex = re.compile(r'''(?x)\s*
1834 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1835 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1836 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1837 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1838 m = operator_rex.fullmatch(filter_spec)
1839 if m:
1840 try:
1841 comparison_value = int(m.group('value'))
1842 except ValueError:
1843 comparison_value = parse_filesize(m.group('value'))
1844 if comparison_value is None:
1845 comparison_value = parse_filesize(m.group('value') + 'B')
1846 if comparison_value is None:
1847 raise ValueError(
1848 'Invalid value %r in format specification %r' % (
1849 m.group('value'), filter_spec))
1850 op = OPERATORS[m.group('op')]
1851
1852 if not m:
1853 STR_OPERATORS = {
1854 '=': operator.eq,
1855 '^=': lambda attr, value: attr.startswith(value),
1856 '$=': lambda attr, value: attr.endswith(value),
1857 '*=': lambda attr, value: value in attr,
1858 '~=': lambda attr, value: value.search(attr) is not None
1859 }
1860 str_operator_rex = re.compile(r'''(?x)\s*
1861 (?P<key>[a-zA-Z0-9._-]+)\s*
1862 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1863 (?P<quote>["'])?
1864 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1865 (?(quote)(?P=quote))\s*
1866 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1867 m = str_operator_rex.fullmatch(filter_spec)
1868 if m:
1869 if m.group('op') == '~=':
1870 comparison_value = re.compile(m.group('value'))
1871 else:
1872 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1873 str_op = STR_OPERATORS[m.group('op')]
1874 if m.group('negation'):
1875 op = lambda attr, value: not str_op(attr, value)
1876 else:
1877 op = str_op
1878
1879 if not m:
1880 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1881
1882 def _filter(f):
1883 actual_value = f.get(m.group('key'))
1884 if actual_value is None:
1885 return m.group('none_inclusive')
1886 return op(actual_value, comparison_value)
1887 return _filter
1888
1889 def _check_formats(self, formats):
1890 for f in formats:
1891 self.to_screen('[info] Testing format %s' % f['format_id'])
1892 path = self.get_output_path('temp')
1893 if not self._ensure_dir_exists(f'{path}/'):
1894 continue
1895 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1896 temp_file.close()
1897 try:
1898 success, _ = self.dl(temp_file.name, f, test=True)
1899 except (DownloadError, IOError, OSError, ValueError) + network_exceptions:
1900 success = False
1901 finally:
1902 if os.path.exists(temp_file.name):
1903 try:
1904 os.remove(temp_file.name)
1905 except OSError:
1906 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1907 if success:
1908 yield f
1909 else:
1910 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1911
1912 def _default_format_spec(self, info_dict, download=True):
1913
1914 def can_merge():
1915 merger = FFmpegMergerPP(self)
1916 return merger.available and merger.can_merge()
1917
1918 prefer_best = (
1919 not self.params.get('simulate')
1920 and download
1921 and (
1922 not can_merge()
1923 or info_dict.get('is_live', False)
1924 or self.outtmpl_dict['default'] == '-'))
1925 compat = (
1926 prefer_best
1927 or self.params.get('allow_multiple_audio_streams', False)
1928 or 'format-spec' in self.params.get('compat_opts', []))
1929
1930 return (
1931 'best/bestvideo+bestaudio' if prefer_best
1932 else 'bestvideo*+bestaudio/best' if not compat
1933 else 'bestvideo+bestaudio/best')
1934
1935 def build_format_selector(self, format_spec):
1936 def syntax_error(note, start):
1937 message = (
1938 'Invalid format specification: '
1939 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1940 return SyntaxError(message)
1941
1942 PICKFIRST = 'PICKFIRST'
1943 MERGE = 'MERGE'
1944 SINGLE = 'SINGLE'
1945 GROUP = 'GROUP'
1946 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1947
1948 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1949 'video': self.params.get('allow_multiple_video_streams', False)}
1950
1951 check_formats = self.params.get('check_formats') == 'selected'
1952
1953 def _parse_filter(tokens):
1954 filter_parts = []
1955 for type, string, start, _, _ in tokens:
1956 if type == tokenize.OP and string == ']':
1957 return ''.join(filter_parts)
1958 else:
1959 filter_parts.append(string)
1960
1961 def _remove_unused_ops(tokens):
1962 # Remove operators that we don't use and join them with the surrounding strings
1963 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1964 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1965 last_string, last_start, last_end, last_line = None, None, None, None
1966 for type, string, start, end, line in tokens:
1967 if type == tokenize.OP and string == '[':
1968 if last_string:
1969 yield tokenize.NAME, last_string, last_start, last_end, last_line
1970 last_string = None
1971 yield type, string, start, end, line
1972 # everything inside brackets will be handled by _parse_filter
1973 for type, string, start, end, line in tokens:
1974 yield type, string, start, end, line
1975 if type == tokenize.OP and string == ']':
1976 break
1977 elif type == tokenize.OP and string in ALLOWED_OPS:
1978 if last_string:
1979 yield tokenize.NAME, last_string, last_start, last_end, last_line
1980 last_string = None
1981 yield type, string, start, end, line
1982 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1983 if not last_string:
1984 last_string = string
1985 last_start = start
1986 last_end = end
1987 else:
1988 last_string += string
1989 if last_string:
1990 yield tokenize.NAME, last_string, last_start, last_end, last_line
1991
1992 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1993 selectors = []
1994 current_selector = None
1995 for type, string, start, _, _ in tokens:
1996 # ENCODING is only defined in python 3.x
1997 if type == getattr(tokenize, 'ENCODING', None):
1998 continue
1999 elif type in [tokenize.NAME, tokenize.NUMBER]:
2000 current_selector = FormatSelector(SINGLE, string, [])
2001 elif type == tokenize.OP:
2002 if string == ')':
2003 if not inside_group:
2004 # ')' will be handled by the parentheses group
2005 tokens.restore_last_token()
2006 break
2007 elif inside_merge and string in ['/', ',']:
2008 tokens.restore_last_token()
2009 break
2010 elif inside_choice and string == ',':
2011 tokens.restore_last_token()
2012 break
2013 elif string == ',':
2014 if not current_selector:
2015 raise syntax_error('"," must follow a format selector', start)
2016 selectors.append(current_selector)
2017 current_selector = None
2018 elif string == '/':
2019 if not current_selector:
2020 raise syntax_error('"/" must follow a format selector', start)
2021 first_choice = current_selector
2022 second_choice = _parse_format_selection(tokens, inside_choice=True)
2023 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2024 elif string == '[':
2025 if not current_selector:
2026 current_selector = FormatSelector(SINGLE, 'best', [])
2027 format_filter = _parse_filter(tokens)
2028 current_selector.filters.append(format_filter)
2029 elif string == '(':
2030 if current_selector:
2031 raise syntax_error('Unexpected "("', start)
2032 group = _parse_format_selection(tokens, inside_group=True)
2033 current_selector = FormatSelector(GROUP, group, [])
2034 elif string == '+':
2035 if not current_selector:
2036 raise syntax_error('Unexpected "+"', start)
2037 selector_1 = current_selector
2038 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2039 if not selector_2:
2040 raise syntax_error('Expected a selector', start)
2041 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2042 else:
2043 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
2044 elif type == tokenize.ENDMARKER:
2045 break
2046 if current_selector:
2047 selectors.append(current_selector)
2048 return selectors
2049
2050 def _merge(formats_pair):
2051 format_1, format_2 = formats_pair
2052
2053 formats_info = []
2054 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2055 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2056
2057 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2058 get_no_more = {'video': False, 'audio': False}
2059 for (i, fmt_info) in enumerate(formats_info):
2060 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2061 formats_info.pop(i)
2062 continue
2063 for aud_vid in ['audio', 'video']:
2064 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2065 if get_no_more[aud_vid]:
2066 formats_info.pop(i)
2067 break
2068 get_no_more[aud_vid] = True
2069
2070 if len(formats_info) == 1:
2071 return formats_info[0]
2072
2073 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2074 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2075
2076 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2077 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2078
2079 output_ext = self.params.get('merge_output_format')
2080 if not output_ext:
2081 if the_only_video:
2082 output_ext = the_only_video['ext']
2083 elif the_only_audio and not video_fmts:
2084 output_ext = the_only_audio['ext']
2085 else:
2086 output_ext = 'mkv'
2087
2088 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2089
2090 new_dict = {
2091 'requested_formats': formats_info,
2092 'format': '+'.join(filtered('format')),
2093 'format_id': '+'.join(filtered('format_id')),
2094 'ext': output_ext,
2095 'protocol': '+'.join(map(determine_protocol, formats_info)),
2096 'language': '+'.join(orderedSet(filtered('language'))) or None,
2097 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2098 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2099 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2100 }
2101
2102 if the_only_video:
2103 new_dict.update({
2104 'width': the_only_video.get('width'),
2105 'height': the_only_video.get('height'),
2106 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2107 'fps': the_only_video.get('fps'),
2108 'dynamic_range': the_only_video.get('dynamic_range'),
2109 'vcodec': the_only_video.get('vcodec'),
2110 'vbr': the_only_video.get('vbr'),
2111 'stretched_ratio': the_only_video.get('stretched_ratio'),
2112 })
2113
2114 if the_only_audio:
2115 new_dict.update({
2116 'acodec': the_only_audio.get('acodec'),
2117 'abr': the_only_audio.get('abr'),
2118 'asr': the_only_audio.get('asr'),
2119 })
2120
2121 return new_dict
2122
2123 def _check_formats(formats):
2124 if not check_formats:
2125 yield from formats
2126 return
2127 yield from self._check_formats(formats)
2128
2129 def _build_selector_function(selector):
2130 if isinstance(selector, list): # ,
2131 fs = [_build_selector_function(s) for s in selector]
2132
2133 def selector_function(ctx):
2134 for f in fs:
2135 yield from f(ctx)
2136 return selector_function
2137
2138 elif selector.type == GROUP: # ()
2139 selector_function = _build_selector_function(selector.selector)
2140
2141 elif selector.type == PICKFIRST: # /
2142 fs = [_build_selector_function(s) for s in selector.selector]
2143
2144 def selector_function(ctx):
2145 for f in fs:
2146 picked_formats = list(f(ctx))
2147 if picked_formats:
2148 return picked_formats
2149 return []
2150
2151 elif selector.type == MERGE: # +
2152 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2153
2154 def selector_function(ctx):
2155 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2156 yield _merge(pair)
2157
2158 elif selector.type == SINGLE: # atom
2159 format_spec = selector.selector or 'best'
2160
2161 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2162 if format_spec == 'all':
2163 def selector_function(ctx):
2164 yield from _check_formats(ctx['formats'][::-1])
2165 elif format_spec == 'mergeall':
2166 def selector_function(ctx):
2167 formats = list(_check_formats(ctx['formats']))
2168 if not formats:
2169 return
2170 merged_format = formats[-1]
2171 for f in formats[-2::-1]:
2172 merged_format = _merge((merged_format, f))
2173 yield merged_format
2174
2175 else:
2176 format_fallback, format_reverse, format_idx = False, True, 1
2177 mobj = re.match(
2178 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2179 format_spec)
2180 if mobj is not None:
2181 format_idx = int_or_none(mobj.group('n'), default=1)
2182 format_reverse = mobj.group('bw')[0] == 'b'
2183 format_type = (mobj.group('type') or [None])[0]
2184 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2185 format_modified = mobj.group('mod') is not None
2186
2187 format_fallback = not format_type and not format_modified # for b, w
2188 _filter_f = (
2189 (lambda f: f.get('%scodec' % format_type) != 'none')
2190 if format_type and format_modified # bv*, ba*, wv*, wa*
2191 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2192 if format_type # bv, ba, wv, wa
2193 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2194 if not format_modified # b, w
2195 else lambda f: True) # b*, w*
2196 filter_f = lambda f: _filter_f(f) and (
2197 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2198 else:
2199 if format_spec in self._format_selection_exts['audio']:
2200 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2201 elif format_spec in self._format_selection_exts['video']:
2202 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2203 elif format_spec in self._format_selection_exts['storyboards']:
2204 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2205 else:
2206 filter_f = lambda f: f.get('format_id') == format_spec # id
2207
2208 def selector_function(ctx):
2209 formats = list(ctx['formats'])
2210 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2211 if format_fallback and ctx['incomplete_formats'] and not matches:
2212 # for extractors with incomplete formats (audio only (soundcloud)
2213 # or video only (imgur)) best/worst will fallback to
2214 # best/worst {video,audio}-only format
2215 matches = formats
2216 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2217 try:
2218 yield matches[format_idx - 1]
2219 except IndexError:
2220 return
2221
2222 filters = [self._build_format_filter(f) for f in selector.filters]
2223
2224 def final_selector(ctx):
2225 ctx_copy = dict(ctx)
2226 for _filter in filters:
2227 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2228 return selector_function(ctx_copy)
2229 return final_selector
2230
2231 stream = io.BytesIO(format_spec.encode('utf-8'))
2232 try:
2233 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
2234 except tokenize.TokenError:
2235 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2236
2237 class TokenIterator(object):
2238 def __init__(self, tokens):
2239 self.tokens = tokens
2240 self.counter = 0
2241
2242 def __iter__(self):
2243 return self
2244
2245 def __next__(self):
2246 if self.counter >= len(self.tokens):
2247 raise StopIteration()
2248 value = self.tokens[self.counter]
2249 self.counter += 1
2250 return value
2251
2252 next = __next__
2253
2254 def restore_last_token(self):
2255 self.counter -= 1
2256
2257 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2258 return _build_selector_function(parsed_selector)
2259
2260 def _calc_headers(self, info_dict):
2261 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2262
2263 cookies = self._calc_cookies(info_dict)
2264 if cookies:
2265 res['Cookie'] = cookies
2266
2267 if 'X-Forwarded-For' not in res:
2268 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2269 if x_forwarded_for_ip:
2270 res['X-Forwarded-For'] = x_forwarded_for_ip
2271
2272 return res
2273
2274 def _calc_cookies(self, info_dict):
2275 pr = sanitized_Request(info_dict['url'])
2276 self.cookiejar.add_cookie_header(pr)
2277 return pr.get_header('Cookie')
2278
2279 def _sort_thumbnails(self, thumbnails):
2280 thumbnails.sort(key=lambda t: (
2281 t.get('preference') if t.get('preference') is not None else -1,
2282 t.get('width') if t.get('width') is not None else -1,
2283 t.get('height') if t.get('height') is not None else -1,
2284 t.get('id') if t.get('id') is not None else '',
2285 t.get('url')))
2286
2287 def _sanitize_thumbnails(self, info_dict):
2288 thumbnails = info_dict.get('thumbnails')
2289 if thumbnails is None:
2290 thumbnail = info_dict.get('thumbnail')
2291 if thumbnail:
2292 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2293 if not thumbnails:
2294 return
2295
2296 def check_thumbnails(thumbnails):
2297 for t in thumbnails:
2298 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2299 try:
2300 self.urlopen(HEADRequest(t['url']))
2301 except network_exceptions as err:
2302 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2303 continue
2304 yield t
2305
2306 self._sort_thumbnails(thumbnails)
2307 for i, t in enumerate(thumbnails):
2308 if t.get('id') is None:
2309 t['id'] = '%d' % i
2310 if t.get('width') and t.get('height'):
2311 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2312 t['url'] = sanitize_url(t['url'])
2313
2314 if self.params.get('check_formats') is True:
2315 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2316 else:
2317 info_dict['thumbnails'] = thumbnails
2318
2319 def _fill_common_fields(self, info_dict, is_video=True):
2320 # TODO: move sanitization here
2321 if is_video:
2322 # playlists are allowed to lack "title"
2323 info_dict['fulltitle'] = info_dict.get('title')
2324 if 'title' not in info_dict:
2325 raise ExtractorError('Missing "title" field in extractor result',
2326 video_id=info_dict['id'], ie=info_dict['extractor'])
2327 elif not info_dict.get('title'):
2328 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2329 info_dict['title'] = f'{info_dict["extractor"]} video #{info_dict["id"]}'
2330
2331 if info_dict.get('duration') is not None:
2332 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2333
2334 for ts_key, date_key in (
2335 ('timestamp', 'upload_date'),
2336 ('release_timestamp', 'release_date'),
2337 ('modified_timestamp', 'modified_date'),
2338 ):
2339 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2340 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2341 # see http://bugs.python.org/issue1646728)
2342 try:
2343 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2344 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2345 except (ValueError, OverflowError, OSError):
2346 pass
2347
2348 live_keys = ('is_live', 'was_live')
2349 live_status = info_dict.get('live_status')
2350 if live_status is None:
2351 for key in live_keys:
2352 if info_dict.get(key) is False:
2353 continue
2354 if info_dict.get(key):
2355 live_status = key
2356 break
2357 if all(info_dict.get(key) is False for key in live_keys):
2358 live_status = 'not_live'
2359 if live_status:
2360 info_dict['live_status'] = live_status
2361 for key in live_keys:
2362 if info_dict.get(key) is None:
2363 info_dict[key] = (live_status == key)
2364
2365 # Auto generate title fields corresponding to the *_number fields when missing
2366 # in order to always have clean titles. This is very common for TV series.
2367 for field in ('chapter', 'season', 'episode'):
2368 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2369 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2370
2371 def process_video_result(self, info_dict, download=True):
2372 assert info_dict.get('_type', 'video') == 'video'
2373 self._num_videos += 1
2374
2375 if 'id' not in info_dict:
2376 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2377 elif not info_dict.get('id'):
2378 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2379
2380 def report_force_conversion(field, field_not, conversion):
2381 self.report_warning(
2382 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2383 % (field, field_not, conversion))
2384
2385 def sanitize_string_field(info, string_field):
2386 field = info.get(string_field)
2387 if field is None or isinstance(field, compat_str):
2388 return
2389 report_force_conversion(string_field, 'a string', 'string')
2390 info[string_field] = compat_str(field)
2391
2392 def sanitize_numeric_fields(info):
2393 for numeric_field in self._NUMERIC_FIELDS:
2394 field = info.get(numeric_field)
2395 if field is None or isinstance(field, compat_numeric_types):
2396 continue
2397 report_force_conversion(numeric_field, 'numeric', 'int')
2398 info[numeric_field] = int_or_none(field)
2399
2400 sanitize_string_field(info_dict, 'id')
2401 sanitize_numeric_fields(info_dict)
2402 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2403 self.report_warning('"duration" field is negative, there is an error in extractor')
2404
2405 if 'playlist' not in info_dict:
2406 # It isn't part of a playlist
2407 info_dict['playlist'] = None
2408 info_dict['playlist_index'] = None
2409
2410 self._sanitize_thumbnails(info_dict)
2411
2412 thumbnail = info_dict.get('thumbnail')
2413 thumbnails = info_dict.get('thumbnails')
2414 if thumbnail:
2415 info_dict['thumbnail'] = sanitize_url(thumbnail)
2416 elif thumbnails:
2417 info_dict['thumbnail'] = thumbnails[-1]['url']
2418
2419 if info_dict.get('display_id') is None and 'id' in info_dict:
2420 info_dict['display_id'] = info_dict['id']
2421
2422 self._fill_common_fields(info_dict)
2423
2424 for cc_kind in ('subtitles', 'automatic_captions'):
2425 cc = info_dict.get(cc_kind)
2426 if cc:
2427 for _, subtitle in cc.items():
2428 for subtitle_format in subtitle:
2429 if subtitle_format.get('url'):
2430 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2431 if subtitle_format.get('ext') is None:
2432 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2433
2434 automatic_captions = info_dict.get('automatic_captions')
2435 subtitles = info_dict.get('subtitles')
2436
2437 info_dict['requested_subtitles'] = self.process_subtitles(
2438 info_dict['id'], subtitles, automatic_captions)
2439
2440 if info_dict.get('formats') is None:
2441 # There's only one format available
2442 formats = [info_dict]
2443 else:
2444 formats = info_dict['formats']
2445
2446 info_dict['__has_drm'] = any(f.get('has_drm') for f in formats)
2447 if not self.params.get('allow_unplayable_formats'):
2448 formats = [f for f in formats if not f.get('has_drm')]
2449
2450 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2451 if not get_from_start:
2452 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2453 if info_dict.get('is_live') and formats:
2454 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2455 if get_from_start and not formats:
2456 self.raise_no_formats(info_dict, msg='--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2457 'If you want to download from the current time, pass --no-live-from-start')
2458
2459 if not formats:
2460 self.raise_no_formats(info_dict)
2461
2462 def is_wellformed(f):
2463 url = f.get('url')
2464 if not url:
2465 self.report_warning(
2466 '"url" field is missing or empty - skipping format, '
2467 'there is an error in extractor')
2468 return False
2469 if isinstance(url, bytes):
2470 sanitize_string_field(f, 'url')
2471 return True
2472
2473 # Filter out malformed formats for better extraction robustness
2474 formats = list(filter(is_wellformed, formats))
2475
2476 formats_dict = {}
2477
2478 # We check that all the formats have the format and format_id fields
2479 for i, format in enumerate(formats):
2480 sanitize_string_field(format, 'format_id')
2481 sanitize_numeric_fields(format)
2482 format['url'] = sanitize_url(format['url'])
2483 if not format.get('format_id'):
2484 format['format_id'] = compat_str(i)
2485 else:
2486 # Sanitize format_id from characters used in format selector expression
2487 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2488 format_id = format['format_id']
2489 if format_id not in formats_dict:
2490 formats_dict[format_id] = []
2491 formats_dict[format_id].append(format)
2492
2493 # Make sure all formats have unique format_id
2494 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2495 for format_id, ambiguous_formats in formats_dict.items():
2496 ambigious_id = len(ambiguous_formats) > 1
2497 for i, format in enumerate(ambiguous_formats):
2498 if ambigious_id:
2499 format['format_id'] = '%s-%d' % (format_id, i)
2500 if format.get('ext') is None:
2501 format['ext'] = determine_ext(format['url']).lower()
2502 # Ensure there is no conflict between id and ext in format selection
2503 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2504 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2505 format['format_id'] = 'f%s' % format['format_id']
2506
2507 for i, format in enumerate(formats):
2508 if format.get('format') is None:
2509 format['format'] = '{id} - {res}{note}'.format(
2510 id=format['format_id'],
2511 res=self.format_resolution(format),
2512 note=format_field(format, 'format_note', ' (%s)'),
2513 )
2514 if format.get('protocol') is None:
2515 format['protocol'] = determine_protocol(format)
2516 if format.get('resolution') is None:
2517 format['resolution'] = self.format_resolution(format, default=None)
2518 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2519 format['dynamic_range'] = 'SDR'
2520 if (info_dict.get('duration') and format.get('tbr')
2521 and not format.get('filesize') and not format.get('filesize_approx')):
2522 format['filesize_approx'] = info_dict['duration'] * format['tbr'] * (1024 / 8)
2523
2524 # Add HTTP headers, so that external programs can use them from the
2525 # json output
2526 full_format_info = info_dict.copy()
2527 full_format_info.update(format)
2528 format['http_headers'] = self._calc_headers(full_format_info)
2529 # Remove private housekeeping stuff
2530 if '__x_forwarded_for_ip' in info_dict:
2531 del info_dict['__x_forwarded_for_ip']
2532
2533 if self.params.get('check_formats') is True:
2534 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2535
2536 if not formats or formats[0] is not info_dict:
2537 # only set the 'formats' fields if the original info_dict list them
2538 # otherwise we end up with a circular reference, the first (and unique)
2539 # element in the 'formats' field in info_dict is info_dict itself,
2540 # which can't be exported to json
2541 info_dict['formats'] = formats
2542
2543 info_dict, _ = self.pre_process(info_dict)
2544
2545 if self._match_entry(info_dict) is not None:
2546 return info_dict
2547
2548 self.post_extract(info_dict)
2549 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2550
2551 # The pre-processors may have modified the formats
2552 formats = info_dict.get('formats', [info_dict])
2553
2554 list_only = self.params.get('simulate') is None and (
2555 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2556 interactive_format_selection = not list_only and self.format_selector == '-'
2557 if self.params.get('list_thumbnails'):
2558 self.list_thumbnails(info_dict)
2559 if self.params.get('listsubtitles'):
2560 if 'automatic_captions' in info_dict:
2561 self.list_subtitles(
2562 info_dict['id'], automatic_captions, 'automatic captions')
2563 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2564 if self.params.get('listformats') or interactive_format_selection:
2565 self.list_formats(info_dict)
2566 if list_only:
2567 # Without this printing, -F --print-json will not work
2568 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2569 return
2570
2571 format_selector = self.format_selector
2572 if format_selector is None:
2573 req_format = self._default_format_spec(info_dict, download=download)
2574 self.write_debug('Default format spec: %s' % req_format)
2575 format_selector = self.build_format_selector(req_format)
2576
2577 while True:
2578 if interactive_format_selection:
2579 req_format = input(
2580 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2581 try:
2582 format_selector = self.build_format_selector(req_format)
2583 except SyntaxError as err:
2584 self.report_error(err, tb=False, is_error=False)
2585 continue
2586
2587 # While in format selection we may need to have an access to the original
2588 # format set in order to calculate some metrics or do some processing.
2589 # For now we need to be able to guess whether original formats provided
2590 # by extractor are incomplete or not (i.e. whether extractor provides only
2591 # video-only or audio-only formats) for proper formats selection for
2592 # extractors with such incomplete formats (see
2593 # https://github.com/ytdl-org/youtube-dl/pull/5556).
2594 # Since formats may be filtered during format selection and may not match
2595 # the original formats the results may be incorrect. Thus original formats
2596 # or pre-calculated metrics should be passed to format selection routines
2597 # as well.
2598 # We will pass a context object containing all necessary additional data
2599 # instead of just formats.
2600 # This fixes incorrect format selection issue (see
2601 # https://github.com/ytdl-org/youtube-dl/issues/10083).
2602 incomplete_formats = (
2603 # All formats are video-only or
2604 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2605 # all formats are audio-only
2606 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
2607
2608 ctx = {
2609 'formats': formats,
2610 'incomplete_formats': incomplete_formats,
2611 }
2612
2613 formats_to_download = list(format_selector(ctx))
2614 if interactive_format_selection and not formats_to_download:
2615 self.report_error('Requested format is not available', tb=False, is_error=False)
2616 continue
2617 break
2618
2619 if not formats_to_download:
2620 if not self.params.get('ignore_no_formats_error'):
2621 raise ExtractorError('Requested format is not available', expected=True,
2622 video_id=info_dict['id'], ie=info_dict['extractor'])
2623 self.report_warning('Requested format is not available')
2624 # Process what we can, even without any available formats.
2625 formats_to_download = [{}]
2626
2627 best_format = formats_to_download[-1]
2628 if download:
2629 if best_format:
2630 self.to_screen(
2631 f'[info] {info_dict["id"]}: Downloading {len(formats_to_download)} format(s): '
2632 + ', '.join([f['format_id'] for f in formats_to_download]))
2633 max_downloads_reached = False
2634 for i, fmt in enumerate(formats_to_download):
2635 formats_to_download[i] = new_info = self._copy_infodict(info_dict)
2636 new_info.update(fmt)
2637 try:
2638 self.process_info(new_info)
2639 except MaxDownloadsReached:
2640 max_downloads_reached = True
2641 # Remove copied info
2642 for key, val in tuple(new_info.items()):
2643 if info_dict.get(key) == val:
2644 new_info.pop(key)
2645 if max_downloads_reached:
2646 break
2647
2648 write_archive = set(f.get('__write_download_archive', False) for f in formats_to_download)
2649 assert write_archive.issubset({True, False, 'ignore'})
2650 if True in write_archive and False not in write_archive:
2651 self.record_download_archive(info_dict)
2652
2653 info_dict['requested_downloads'] = formats_to_download
2654 info_dict = self.run_all_pps('after_video', info_dict)
2655 if max_downloads_reached:
2656 raise MaxDownloadsReached()
2657
2658 # We update the info dict with the selected best quality format (backwards compatibility)
2659 info_dict.update(best_format)
2660 return info_dict
2661
2662 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2663 """Select the requested subtitles and their format"""
2664 available_subs = {}
2665 if normal_subtitles and self.params.get('writesubtitles'):
2666 available_subs.update(normal_subtitles)
2667 if automatic_captions and self.params.get('writeautomaticsub'):
2668 for lang, cap_info in automatic_captions.items():
2669 if lang not in available_subs:
2670 available_subs[lang] = cap_info
2671
2672 if (not self.params.get('writesubtitles') and not
2673 self.params.get('writeautomaticsub') or not
2674 available_subs):
2675 return None
2676
2677 all_sub_langs = available_subs.keys()
2678 if self.params.get('allsubtitles', False):
2679 requested_langs = all_sub_langs
2680 elif self.params.get('subtitleslangs', False):
2681 # A list is used so that the order of languages will be the same as
2682 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2683 requested_langs = []
2684 for lang_re in self.params.get('subtitleslangs'):
2685 discard = lang_re[0] == '-'
2686 if discard:
2687 lang_re = lang_re[1:]
2688 if lang_re == 'all':
2689 if discard:
2690 requested_langs = []
2691 else:
2692 requested_langs.extend(all_sub_langs)
2693 continue
2694 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2695 if discard:
2696 for lang in current_langs:
2697 while lang in requested_langs:
2698 requested_langs.remove(lang)
2699 else:
2700 requested_langs.extend(current_langs)
2701 requested_langs = orderedSet(requested_langs)
2702 elif 'en' in available_subs:
2703 requested_langs = ['en']
2704 else:
2705 requested_langs = [list(all_sub_langs)[0]]
2706 if requested_langs:
2707 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2708
2709 formats_query = self.params.get('subtitlesformat', 'best')
2710 formats_preference = formats_query.split('/') if formats_query else []
2711 subs = {}
2712 for lang in requested_langs:
2713 formats = available_subs.get(lang)
2714 if formats is None:
2715 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
2716 continue
2717 for ext in formats_preference:
2718 if ext == 'best':
2719 f = formats[-1]
2720 break
2721 matches = list(filter(lambda f: f['ext'] == ext, formats))
2722 if matches:
2723 f = matches[-1]
2724 break
2725 else:
2726 f = formats[-1]
2727 self.report_warning(
2728 'No subtitle format found matching "%s" for language %s, '
2729 'using %s' % (formats_query, lang, f['ext']))
2730 subs[lang] = f
2731 return subs
2732
2733 def _forceprint(self, key, info_dict):
2734 if info_dict is None:
2735 return
2736 info_copy = info_dict.copy()
2737 info_copy['formats_table'] = self.render_formats_table(info_dict)
2738 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2739 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2740 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2741
2742 def format_tmpl(tmpl):
2743 mobj = re.match(r'\w+(=?)$', tmpl)
2744 if mobj and mobj.group(1):
2745 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2746 elif mobj:
2747 return f'%({tmpl})s'
2748 return tmpl
2749
2750 for tmpl in self.params['forceprint'].get(key, []):
2751 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2752
2753 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2754 filename = self.evaluate_outtmpl(file_tmpl, info_dict)
2755 tmpl = format_tmpl(tmpl)
2756 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2757 if self._ensure_dir_exists(filename):
2758 with io.open(filename, 'a', encoding='utf-8') as f:
2759 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2760
2761 def __forced_printings(self, info_dict, filename, incomplete):
2762 def print_mandatory(field, actual_field=None):
2763 if actual_field is None:
2764 actual_field = field
2765 if (self.params.get('force%s' % field, False)
2766 and (not incomplete or info_dict.get(actual_field) is not None)):
2767 self.to_stdout(info_dict[actual_field])
2768
2769 def print_optional(field):
2770 if (self.params.get('force%s' % field, False)
2771 and info_dict.get(field) is not None):
2772 self.to_stdout(info_dict[field])
2773
2774 info_dict = info_dict.copy()
2775 if filename is not None:
2776 info_dict['filename'] = filename
2777 if info_dict.get('requested_formats') is not None:
2778 # For RTMP URLs, also include the playpath
2779 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2780 elif 'url' in info_dict:
2781 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2782
2783 if (self.params.get('forcejson')
2784 or self.params['forceprint'].get('video')
2785 or self.params['print_to_file'].get('video')):
2786 self.post_extract(info_dict)
2787 self._forceprint('video', info_dict)
2788
2789 print_mandatory('title')
2790 print_mandatory('id')
2791 print_mandatory('url', 'urls')
2792 print_optional('thumbnail')
2793 print_optional('description')
2794 print_optional('filename')
2795 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2796 self.to_stdout(formatSeconds(info_dict['duration']))
2797 print_mandatory('format')
2798
2799 if self.params.get('forcejson'):
2800 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2801
2802 def dl(self, name, info, subtitle=False, test=False):
2803 if not info.get('url'):
2804 self.raise_no_formats(info, True)
2805
2806 if test:
2807 verbose = self.params.get('verbose')
2808 params = {
2809 'test': True,
2810 'quiet': self.params.get('quiet') or not verbose,
2811 'verbose': verbose,
2812 'noprogress': not verbose,
2813 'nopart': True,
2814 'skip_unavailable_fragments': False,
2815 'keep_fragments': False,
2816 'overwrites': True,
2817 '_no_ytdl_file': True,
2818 }
2819 else:
2820 params = self.params
2821 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2822 if not test:
2823 for ph in self._progress_hooks:
2824 fd.add_progress_hook(ph)
2825 urls = '", "'.join(
2826 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2827 for f in info.get('requested_formats', []) or [info])
2828 self.write_debug('Invoking downloader on "%s"' % urls)
2829
2830 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2831 # But it may contain objects that are not deep-copyable
2832 new_info = self._copy_infodict(info)
2833 if new_info.get('http_headers') is None:
2834 new_info['http_headers'] = self._calc_headers(new_info)
2835 return fd.download(name, new_info, subtitle)
2836
2837 def existing_file(self, filepaths, *, default_overwrite=True):
2838 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2839 if existing_files and not self.params.get('overwrites', default_overwrite):
2840 return existing_files[0]
2841
2842 for file in existing_files:
2843 self.report_file_delete(file)
2844 os.remove(file)
2845 return None
2846
2847 def process_info(self, info_dict):
2848 """Process a single resolved IE result. (Modifies it in-place)"""
2849
2850 assert info_dict.get('_type', 'video') == 'video'
2851 original_infodict = info_dict
2852
2853 if 'format' not in info_dict and 'ext' in info_dict:
2854 info_dict['format'] = info_dict['ext']
2855
2856 # This is mostly just for backward compatibility of process_info
2857 # As a side-effect, this allows for format-specific filters
2858 if self._match_entry(info_dict) is not None:
2859 info_dict['__write_download_archive'] = 'ignore'
2860 return
2861
2862 # Does nothing under normal operation - for backward compatibility of process_info
2863 self.post_extract(info_dict)
2864
2865 # info_dict['_filename'] needs to be set for backward compatibility
2866 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2867 temp_filename = self.prepare_filename(info_dict, 'temp')
2868 files_to_move = {}
2869
2870 self._num_downloads += 1
2871
2872 # Forced printings
2873 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2874
2875 if self.params.get('simulate'):
2876 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2877 return
2878
2879 if full_filename is None:
2880 return
2881 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2882 return
2883 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2884 return
2885
2886 if self._write_description('video', info_dict,
2887 self.prepare_filename(info_dict, 'description')) is None:
2888 return
2889
2890 sub_files = self._write_subtitles(info_dict, temp_filename)
2891 if sub_files is None:
2892 return
2893 files_to_move.update(dict(sub_files))
2894
2895 thumb_files = self._write_thumbnails(
2896 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2897 if thumb_files is None:
2898 return
2899 files_to_move.update(dict(thumb_files))
2900
2901 infofn = self.prepare_filename(info_dict, 'infojson')
2902 _infojson_written = self._write_info_json('video', info_dict, infofn)
2903 if _infojson_written:
2904 info_dict['infojson_filename'] = infofn
2905 # For backward compatibility, even though it was a private field
2906 info_dict['__infojson_filename'] = infofn
2907 elif _infojson_written is None:
2908 return
2909
2910 # Note: Annotations are deprecated
2911 annofn = None
2912 if self.params.get('writeannotations', False):
2913 annofn = self.prepare_filename(info_dict, 'annotation')
2914 if annofn:
2915 if not self._ensure_dir_exists(encodeFilename(annofn)):
2916 return
2917 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2918 self.to_screen('[info] Video annotations are already present')
2919 elif not info_dict.get('annotations'):
2920 self.report_warning('There are no annotations to write.')
2921 else:
2922 try:
2923 self.to_screen('[info] Writing video annotations to: ' + annofn)
2924 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2925 annofile.write(info_dict['annotations'])
2926 except (KeyError, TypeError):
2927 self.report_warning('There are no annotations to write.')
2928 except (OSError, IOError):
2929 self.report_error('Cannot write annotations file: ' + annofn)
2930 return
2931
2932 # Write internet shortcut files
2933 def _write_link_file(link_type):
2934 url = try_get(info_dict['webpage_url'], iri_to_uri)
2935 if not url:
2936 self.report_warning(
2937 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2938 return True
2939 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2940 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2941 return False
2942 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2943 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2944 return True
2945 try:
2946 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2947 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2948 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2949 template_vars = {'url': url}
2950 if link_type == 'desktop':
2951 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2952 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2953 except (OSError, IOError):
2954 self.report_error(f'Cannot write internet shortcut {linkfn}')
2955 return False
2956 return True
2957
2958 write_links = {
2959 'url': self.params.get('writeurllink'),
2960 'webloc': self.params.get('writewebloclink'),
2961 'desktop': self.params.get('writedesktoplink'),
2962 }
2963 if self.params.get('writelink'):
2964 link_type = ('webloc' if sys.platform == 'darwin'
2965 else 'desktop' if sys.platform.startswith('linux')
2966 else 'url')
2967 write_links[link_type] = True
2968
2969 if any(should_write and not _write_link_file(link_type)
2970 for link_type, should_write in write_links.items()):
2971 return
2972
2973 def replace_info_dict(new_info):
2974 nonlocal info_dict
2975 if new_info == info_dict:
2976 return
2977 info_dict.clear()
2978 info_dict.update(new_info)
2979
2980 try:
2981 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2982 replace_info_dict(new_info)
2983 except PostProcessingError as err:
2984 self.report_error('Preprocessing: %s' % str(err))
2985 return
2986
2987 if self.params.get('skip_download'):
2988 info_dict['filepath'] = temp_filename
2989 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2990 info_dict['__files_to_move'] = files_to_move
2991 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2992 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2993 else:
2994 # Download
2995 info_dict.setdefault('__postprocessors', [])
2996 try:
2997
2998 def existing_video_file(*filepaths):
2999 ext = info_dict.get('ext')
3000 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3001 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3002 default_overwrite=False)
3003 if file:
3004 info_dict['ext'] = os.path.splitext(file)[1][1:]
3005 return file
3006
3007 success = True
3008 if info_dict.get('requested_formats') is not None:
3009
3010 def compatible_formats(formats):
3011 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3012 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3013 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3014 if len(video_formats) > 2 or len(audio_formats) > 2:
3015 return False
3016
3017 # Check extension
3018 exts = set(format.get('ext') for format in formats)
3019 COMPATIBLE_EXTS = (
3020 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
3021 set(('webm',)),
3022 )
3023 for ext_sets in COMPATIBLE_EXTS:
3024 if ext_sets.issuperset(exts):
3025 return True
3026 # TODO: Check acodec/vcodec
3027 return False
3028
3029 requested_formats = info_dict['requested_formats']
3030 old_ext = info_dict['ext']
3031 if self.params.get('merge_output_format') is None:
3032 if not compatible_formats(requested_formats):
3033 info_dict['ext'] = 'mkv'
3034 self.report_warning(
3035 'Requested formats are incompatible for merge and will be merged into mkv')
3036 if (info_dict['ext'] == 'webm'
3037 and info_dict.get('thumbnails')
3038 # check with type instead of pp_key, __name__, or isinstance
3039 # since we dont want any custom PPs to trigger this
3040 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])):
3041 info_dict['ext'] = 'mkv'
3042 self.report_warning(
3043 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3044 new_ext = info_dict['ext']
3045
3046 def correct_ext(filename, ext=new_ext):
3047 if filename == '-':
3048 return filename
3049 filename_real_ext = os.path.splitext(filename)[1][1:]
3050 filename_wo_ext = (
3051 os.path.splitext(filename)[0]
3052 if filename_real_ext in (old_ext, new_ext)
3053 else filename)
3054 return '%s.%s' % (filename_wo_ext, ext)
3055
3056 # Ensure filename always has a correct extension for successful merge
3057 full_filename = correct_ext(full_filename)
3058 temp_filename = correct_ext(temp_filename)
3059 dl_filename = existing_video_file(full_filename, temp_filename)
3060 info_dict['__real_download'] = False
3061
3062 downloaded = []
3063 merger = FFmpegMergerPP(self)
3064
3065 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3066 if dl_filename is not None:
3067 self.report_file_already_downloaded(dl_filename)
3068 elif fd:
3069 for f in requested_formats if fd != FFmpegFD else []:
3070 f['filepath'] = fname = prepend_extension(
3071 correct_ext(temp_filename, info_dict['ext']),
3072 'f%s' % f['format_id'], info_dict['ext'])
3073 downloaded.append(fname)
3074 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3075 success, real_download = self.dl(temp_filename, info_dict)
3076 info_dict['__real_download'] = real_download
3077 else:
3078 if self.params.get('allow_unplayable_formats'):
3079 self.report_warning(
3080 'You have requested merging of multiple formats '
3081 'while also allowing unplayable formats to be downloaded. '
3082 'The formats won\'t be merged to prevent data corruption.')
3083 elif not merger.available:
3084 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3085 if not self.params.get('ignoreerrors'):
3086 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3087 return
3088 self.report_warning(f'{msg}. The formats won\'t be merged')
3089
3090 if temp_filename == '-':
3091 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3092 else 'but the formats are incompatible for simultaneous download' if merger.available
3093 else 'but ffmpeg is not installed')
3094 self.report_warning(
3095 f'You have requested downloading multiple formats to stdout {reason}. '
3096 'The formats will be streamed one after the other')
3097 fname = temp_filename
3098 for f in requested_formats:
3099 new_info = dict(info_dict)
3100 del new_info['requested_formats']
3101 new_info.update(f)
3102 if temp_filename != '-':
3103 fname = prepend_extension(
3104 correct_ext(temp_filename, new_info['ext']),
3105 'f%s' % f['format_id'], new_info['ext'])
3106 if not self._ensure_dir_exists(fname):
3107 return
3108 f['filepath'] = fname
3109 downloaded.append(fname)
3110 partial_success, real_download = self.dl(fname, new_info)
3111 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3112 success = success and partial_success
3113
3114 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3115 info_dict['__postprocessors'].append(merger)
3116 info_dict['__files_to_merge'] = downloaded
3117 # Even if there were no downloads, it is being merged only now
3118 info_dict['__real_download'] = True
3119 else:
3120 for file in downloaded:
3121 files_to_move[file] = None
3122 else:
3123 # Just a single file
3124 dl_filename = existing_video_file(full_filename, temp_filename)
3125 if dl_filename is None or dl_filename == temp_filename:
3126 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3127 # So we should try to resume the download
3128 success, real_download = self.dl(temp_filename, info_dict)
3129 info_dict['__real_download'] = real_download
3130 else:
3131 self.report_file_already_downloaded(dl_filename)
3132
3133 dl_filename = dl_filename or temp_filename
3134 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3135
3136 except network_exceptions as err:
3137 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3138 return
3139 except (OSError, IOError) as err:
3140 raise UnavailableVideoError(err)
3141 except (ContentTooShortError, ) as err:
3142 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
3143 return
3144
3145 if success and full_filename != '-':
3146
3147 def fixup():
3148 do_fixup = True
3149 fixup_policy = self.params.get('fixup')
3150 vid = info_dict['id']
3151
3152 if fixup_policy in ('ignore', 'never'):
3153 return
3154 elif fixup_policy == 'warn':
3155 do_fixup = False
3156 elif fixup_policy != 'force':
3157 assert fixup_policy in ('detect_or_warn', None)
3158 if not info_dict.get('__real_download'):
3159 do_fixup = False
3160
3161 def ffmpeg_fixup(cndn, msg, cls):
3162 if not cndn:
3163 return
3164 if not do_fixup:
3165 self.report_warning(f'{vid}: {msg}')
3166 return
3167 pp = cls(self)
3168 if pp.available:
3169 info_dict['__postprocessors'].append(pp)
3170 else:
3171 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3172
3173 stretched_ratio = info_dict.get('stretched_ratio')
3174 ffmpeg_fixup(
3175 stretched_ratio not in (1, None),
3176 f'Non-uniform pixel ratio {stretched_ratio}',
3177 FFmpegFixupStretchedPP)
3178
3179 ffmpeg_fixup(
3180 (info_dict.get('requested_formats') is None
3181 and info_dict.get('container') == 'm4a_dash'
3182 and info_dict.get('ext') == 'm4a'),
3183 'writing DASH m4a. Only some players support this container',
3184 FFmpegFixupM4aPP)
3185
3186 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3187 downloader = downloader.__name__ if downloader else None
3188
3189 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3190 ffmpeg_fixup(downloader == 'HlsFD',
3191 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3192 FFmpegFixupM3u8PP)
3193 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3194 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3195
3196 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3197 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed duration detected', FFmpegFixupDurationPP)
3198
3199 fixup()
3200 try:
3201 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3202 except PostProcessingError as err:
3203 self.report_error('Postprocessing: %s' % str(err))
3204 return
3205 try:
3206 for ph in self._post_hooks:
3207 ph(info_dict['filepath'])
3208 except Exception as err:
3209 self.report_error('post hooks: %s' % str(err))
3210 return
3211 info_dict['__write_download_archive'] = True
3212
3213 if self.params.get('force_write_download_archive'):
3214 info_dict['__write_download_archive'] = True
3215
3216 # Make sure the info_dict was modified in-place
3217 assert info_dict is original_infodict
3218
3219 max_downloads = self.params.get('max_downloads')
3220 if max_downloads is not None and self._num_downloads >= int(max_downloads):
3221 raise MaxDownloadsReached()
3222
3223 def __download_wrapper(self, func):
3224 @functools.wraps(func)
3225 def wrapper(*args, **kwargs):
3226 try:
3227 res = func(*args, **kwargs)
3228 except UnavailableVideoError as e:
3229 self.report_error(e)
3230 except MaxDownloadsReached as e:
3231 self.to_screen(f'[info] {e}')
3232 raise
3233 except DownloadCancelled as e:
3234 self.to_screen(f'[info] {e}')
3235 if not self.params.get('break_per_url'):
3236 raise
3237 else:
3238 if self.params.get('dump_single_json', False):
3239 self.post_extract(res)
3240 self.to_stdout(json.dumps(self.sanitize_info(res)))
3241 return wrapper
3242
3243 def download(self, url_list):
3244 """Download a given list of URLs."""
3245 url_list = variadic(url_list) # Passing a single URL is a common mistake
3246 outtmpl = self.outtmpl_dict['default']
3247 if (len(url_list) > 1
3248 and outtmpl != '-'
3249 and '%' not in outtmpl
3250 and self.params.get('max_downloads') != 1):
3251 raise SameFileError(outtmpl)
3252
3253 for url in url_list:
3254 self.__download_wrapper(self.extract_info)(
3255 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3256
3257 return self._download_retcode
3258
3259 def download_with_info_file(self, info_filename):
3260 with contextlib.closing(fileinput.FileInput(
3261 [info_filename], mode='r',
3262 openhook=fileinput.hook_encoded('utf-8'))) as f:
3263 # FileInput doesn't have a read method, we can't call json.load
3264 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3265 try:
3266 self.__download_wrapper(self.process_ie_result)(info, download=True)
3267 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3268 if not isinstance(e, EntryNotInPlaylist):
3269 self.to_stderr('\r')
3270 webpage_url = info.get('webpage_url')
3271 if webpage_url is not None:
3272 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3273 return self.download([webpage_url])
3274 else:
3275 raise
3276 return self._download_retcode
3277
3278 @staticmethod
3279 def sanitize_info(info_dict, remove_private_keys=False):
3280 ''' Sanitize the infodict for converting to json '''
3281 if info_dict is None:
3282 return info_dict
3283 info_dict.setdefault('epoch', int(time.time()))
3284 info_dict.setdefault('_type', 'video')
3285
3286 if remove_private_keys:
3287 reject = lambda k, v: v is None or (k.startswith('_') and k != '_type') or k in {
3288 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3289 'entries', 'filepath', 'infojson_filename', 'original_url', 'playlist_autonumber',
3290 }
3291 else:
3292 reject = lambda k, v: False
3293
3294 def filter_fn(obj):
3295 if isinstance(obj, dict):
3296 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3297 elif isinstance(obj, (list, tuple, set, LazyList)):
3298 return list(map(filter_fn, obj))
3299 elif obj is None or isinstance(obj, (str, int, float, bool)):
3300 return obj
3301 else:
3302 return repr(obj)
3303
3304 return filter_fn(info_dict)
3305
3306 @staticmethod
3307 def filter_requested_info(info_dict, actually_filter=True):
3308 ''' Alias of sanitize_info for backward compatibility '''
3309 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3310
3311 @staticmethod
3312 def post_extract(info_dict):
3313 def actual_post_extract(info_dict):
3314 if info_dict.get('_type') in ('playlist', 'multi_video'):
3315 for video_dict in info_dict.get('entries', {}):
3316 actual_post_extract(video_dict or {})
3317 return
3318
3319 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3320 info_dict.update(post_extractor())
3321
3322 actual_post_extract(info_dict or {})
3323
3324 def run_pp(self, pp, infodict):
3325 files_to_delete = []
3326 if '__files_to_move' not in infodict:
3327 infodict['__files_to_move'] = {}
3328 try:
3329 files_to_delete, infodict = pp.run(infodict)
3330 except PostProcessingError as e:
3331 # Must be True and not 'only_download'
3332 if self.params.get('ignoreerrors') is True:
3333 self.report_error(e)
3334 return infodict
3335 raise
3336
3337 if not files_to_delete:
3338 return infodict
3339 if self.params.get('keepvideo', False):
3340 for f in files_to_delete:
3341 infodict['__files_to_move'].setdefault(f, '')
3342 else:
3343 for old_filename in set(files_to_delete):
3344 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
3345 try:
3346 os.remove(encodeFilename(old_filename))
3347 except (IOError, OSError):
3348 self.report_warning('Unable to remove downloaded original file')
3349 if old_filename in infodict['__files_to_move']:
3350 del infodict['__files_to_move'][old_filename]
3351 return infodict
3352
3353 def run_all_pps(self, key, info, *, additional_pps=None):
3354 self._forceprint(key, info)
3355 for pp in (additional_pps or []) + self._pps[key]:
3356 info = self.run_pp(pp, info)
3357 return info
3358
3359 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3360 info = dict(ie_info)
3361 info['__files_to_move'] = files_to_move or {}
3362 info = self.run_all_pps(key, info)
3363 return info, info.pop('__files_to_move', None)
3364
3365 def post_process(self, filename, info, files_to_move=None):
3366 """Run all the postprocessors on the given file."""
3367 info['filepath'] = filename
3368 info['__files_to_move'] = files_to_move or {}
3369 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3370 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3371 del info['__files_to_move']
3372 return self.run_all_pps('after_move', info)
3373
3374 def _make_archive_id(self, info_dict):
3375 video_id = info_dict.get('id')
3376 if not video_id:
3377 return
3378 # Future-proof against any change in case
3379 # and backwards compatibility with prior versions
3380 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3381 if extractor is None:
3382 url = str_or_none(info_dict.get('url'))
3383 if not url:
3384 return
3385 # Try to find matching extractor for the URL and take its ie_key
3386 for ie_key, ie in self._ies.items():
3387 if ie.suitable(url):
3388 extractor = ie_key
3389 break
3390 else:
3391 return
3392 return '%s %s' % (extractor.lower(), video_id)
3393
3394 def in_download_archive(self, info_dict):
3395 fn = self.params.get('download_archive')
3396 if fn is None:
3397 return False
3398
3399 vid_id = self._make_archive_id(info_dict)
3400 if not vid_id:
3401 return False # Incomplete video information
3402
3403 return vid_id in self.archive
3404
3405 def record_download_archive(self, info_dict):
3406 fn = self.params.get('download_archive')
3407 if fn is None:
3408 return
3409 vid_id = self._make_archive_id(info_dict)
3410 assert vid_id
3411 self.write_debug(f'Adding to archive: {vid_id}')
3412 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3413 archive_file.write(vid_id + '\n')
3414 self.archive.add(vid_id)
3415
3416 @staticmethod
3417 def format_resolution(format, default='unknown'):
3418 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3419 return 'audio only'
3420 if format.get('resolution') is not None:
3421 return format['resolution']
3422 if format.get('width') and format.get('height'):
3423 return '%dx%d' % (format['width'], format['height'])
3424 elif format.get('height'):
3425 return '%sp' % format['height']
3426 elif format.get('width'):
3427 return '%dx?' % format['width']
3428 return default
3429
3430 def _list_format_headers(self, *headers):
3431 if self.params.get('listformats_table', True) is not False:
3432 return [self._format_screen(header, self.Styles.HEADERS) for header in headers]
3433 return headers
3434
3435 def _format_note(self, fdict):
3436 res = ''
3437 if fdict.get('ext') in ['f4f', 'f4m']:
3438 res += '(unsupported)'
3439 if fdict.get('language'):
3440 if res:
3441 res += ' '
3442 res += '[%s]' % fdict['language']
3443 if fdict.get('format_note') is not None:
3444 if res:
3445 res += ' '
3446 res += fdict['format_note']
3447 if fdict.get('tbr') is not None:
3448 if res:
3449 res += ', '
3450 res += '%4dk' % fdict['tbr']
3451 if fdict.get('container') is not None:
3452 if res:
3453 res += ', '
3454 res += '%s container' % fdict['container']
3455 if (fdict.get('vcodec') is not None
3456 and fdict.get('vcodec') != 'none'):
3457 if res:
3458 res += ', '
3459 res += fdict['vcodec']
3460 if fdict.get('vbr') is not None:
3461 res += '@'
3462 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3463 res += 'video@'
3464 if fdict.get('vbr') is not None:
3465 res += '%4dk' % fdict['vbr']
3466 if fdict.get('fps') is not None:
3467 if res:
3468 res += ', '
3469 res += '%sfps' % fdict['fps']
3470 if fdict.get('acodec') is not None:
3471 if res:
3472 res += ', '
3473 if fdict['acodec'] == 'none':
3474 res += 'video only'
3475 else:
3476 res += '%-5s' % fdict['acodec']
3477 elif fdict.get('abr') is not None:
3478 if res:
3479 res += ', '
3480 res += 'audio'
3481 if fdict.get('abr') is not None:
3482 res += '@%3dk' % fdict['abr']
3483 if fdict.get('asr') is not None:
3484 res += ' (%5dHz)' % fdict['asr']
3485 if fdict.get('filesize') is not None:
3486 if res:
3487 res += ', '
3488 res += format_bytes(fdict['filesize'])
3489 elif fdict.get('filesize_approx') is not None:
3490 if res:
3491 res += ', '
3492 res += '~' + format_bytes(fdict['filesize_approx'])
3493 return res
3494
3495 def render_formats_table(self, info_dict):
3496 if not info_dict.get('formats') and not info_dict.get('url'):
3497 return None
3498
3499 formats = info_dict.get('formats', [info_dict])
3500 if not self.params.get('listformats_table', True) is not False:
3501 table = [
3502 [
3503 format_field(f, 'format_id'),
3504 format_field(f, 'ext'),
3505 self.format_resolution(f),
3506 self._format_note(f)
3507 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3508 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3509
3510 delim = self._format_screen('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3511 table = [
3512 [
3513 self._format_screen(format_field(f, 'format_id'), self.Styles.ID),
3514 format_field(f, 'ext'),
3515 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3516 format_field(f, 'fps', '\t%d'),
3517 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3518 delim,
3519 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3520 format_field(f, 'tbr', '\t%dk'),
3521 shorten_protocol_name(f.get('protocol', '')),
3522 delim,
3523 format_field(f, 'vcodec', default='unknown').replace(
3524 'none', 'images' if f.get('acodec') == 'none'
3525 else self._format_screen('audio only', self.Styles.SUPPRESS)),
3526 format_field(f, 'vbr', '\t%dk'),
3527 format_field(f, 'acodec', default='unknown').replace(
3528 'none', '' if f.get('vcodec') == 'none'
3529 else self._format_screen('video only', self.Styles.SUPPRESS)),
3530 format_field(f, 'abr', '\t%dk'),
3531 format_field(f, 'asr', '\t%dHz'),
3532 join_nonempty(
3533 self._format_screen('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3534 format_field(f, 'language', '[%s]'),
3535 join_nonempty(format_field(f, 'format_note'),
3536 format_field(f, 'container', ignore=(None, f.get('ext'))),
3537 delim=', '),
3538 delim=' '),
3539 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3540 header_line = self._list_format_headers(
3541 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3542 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3543
3544 return render_table(
3545 header_line, table, hide_empty=True,
3546 delim=self._format_screen('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3547
3548 def render_thumbnails_table(self, info_dict):
3549 thumbnails = list(info_dict.get('thumbnails') or [])
3550 if not thumbnails:
3551 return None
3552 return render_table(
3553 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3554 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3555
3556 def render_subtitles_table(self, video_id, subtitles):
3557 def _row(lang, formats):
3558 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3559 if len(set(names)) == 1:
3560 names = [] if names[0] == 'unknown' else names[:1]
3561 return [lang, ', '.join(names), ', '.join(exts)]
3562
3563 if not subtitles:
3564 return None
3565 return render_table(
3566 self._list_format_headers('Language', 'Name', 'Formats'),
3567 [_row(lang, formats) for lang, formats in subtitles.items()],
3568 hide_empty=True)
3569
3570 def __list_table(self, video_id, name, func, *args):
3571 table = func(*args)
3572 if not table:
3573 self.to_screen(f'{video_id} has no {name}')
3574 return
3575 self.to_screen(f'[info] Available {name} for {video_id}:')
3576 self.to_stdout(table)
3577
3578 def list_formats(self, info_dict):
3579 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3580
3581 def list_thumbnails(self, info_dict):
3582 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3583
3584 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3585 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3586
3587 def urlopen(self, req):
3588 """ Start an HTTP download """
3589 if isinstance(req, compat_basestring):
3590 req = sanitized_Request(req)
3591 return self._opener.open(req, timeout=self._socket_timeout)
3592
3593 def print_debug_header(self):
3594 if not self.params.get('verbose'):
3595 return
3596
3597 def get_encoding(stream):
3598 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3599 if not supports_terminal_sequences(stream):
3600 from .compat import WINDOWS_VT_MODE
3601 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3602 return ret
3603
3604 encoding_str = 'Encodings: locale %s, fs %s, out %s, err %s, pref %s' % (
3605 locale.getpreferredencoding(),
3606 sys.getfilesystemencoding(),
3607 get_encoding(self._screen_file), get_encoding(self._err_file),
3608 self.get_encoding())
3609
3610 logger = self.params.get('logger')
3611 if logger:
3612 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3613 write_debug(encoding_str)
3614 else:
3615 write_string(f'[debug] {encoding_str}\n', encoding=None)
3616 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3617
3618 source = detect_variant()
3619 write_debug(join_nonempty(
3620 'yt-dlp version', __version__,
3621 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3622 '' if source == 'unknown' else f'({source})',
3623 delim=' '))
3624 if not _LAZY_LOADER:
3625 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3626 write_debug('Lazy loading extractors is forcibly disabled')
3627 else:
3628 write_debug('Lazy loading extractors is disabled')
3629 if plugin_extractors or plugin_postprocessors:
3630 write_debug('Plugins: %s' % [
3631 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3632 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3633 if self.params.get('compat_opts'):
3634 write_debug('Compatibility options: %s' % ', '.join(self.params.get('compat_opts')))
3635
3636 if source == 'source':
3637 try:
3638 sp = Popen(
3639 ['git', 'rev-parse', '--short', 'HEAD'],
3640 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3641 cwd=os.path.dirname(os.path.abspath(__file__)))
3642 out, err = sp.communicate_or_kill()
3643 out = out.decode().strip()
3644 if re.match('[0-9a-f]+', out):
3645 write_debug('Git HEAD: %s' % out)
3646 except Exception:
3647 try:
3648 sys.exc_clear()
3649 except Exception:
3650 pass
3651
3652 def python_implementation():
3653 impl_name = platform.python_implementation()
3654 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3655 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3656 return impl_name
3657
3658 write_debug('Python version %s (%s %s) - %s' % (
3659 platform.python_version(),
3660 python_implementation(),
3661 platform.architecture()[0],
3662 platform_name()))
3663
3664 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3665 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3666 if ffmpeg_features:
3667 exe_versions['ffmpeg'] += ' (%s)' % ','.join(ffmpeg_features)
3668
3669 exe_versions['rtmpdump'] = rtmpdump_version()
3670 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3671 exe_str = ', '.join(
3672 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3673 ) or 'none'
3674 write_debug('exe versions: %s' % exe_str)
3675
3676 from .downloader.websocket import has_websockets
3677 from .postprocessor.embedthumbnail import has_mutagen
3678 from .cookies import SQLITE_AVAILABLE, SECRETSTORAGE_AVAILABLE
3679
3680 lib_str = join_nonempty(
3681 compat_brotli and compat_brotli.__name__,
3682 compat_pycrypto_AES and compat_pycrypto_AES.__name__.split('.')[0],
3683 SECRETSTORAGE_AVAILABLE and 'secretstorage',
3684 has_mutagen and 'mutagen',
3685 SQLITE_AVAILABLE and 'sqlite',
3686 has_websockets and 'websockets',
3687 delim=', ') or 'none'
3688 write_debug('Optional libraries: %s' % lib_str)
3689
3690 proxy_map = {}
3691 for handler in self._opener.handlers:
3692 if hasattr(handler, 'proxies'):
3693 proxy_map.update(handler.proxies)
3694 write_debug(f'Proxy map: {proxy_map}')
3695
3696 # Not implemented
3697 if False and self.params.get('call_home'):
3698 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3699 write_debug('Public IP address: %s' % ipaddr)
3700 latest_version = self.urlopen(
3701 'https://yt-dl.org/latest/version').read().decode('utf-8')
3702 if version_tuple(latest_version) > version_tuple(__version__):
3703 self.report_warning(
3704 'You are using an outdated version (newest version: %s)! '
3705 'See https://yt-dl.org/update if you need help updating.' %
3706 latest_version)
3707
3708 def _setup_opener(self):
3709 timeout_val = self.params.get('socket_timeout')
3710 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3711
3712 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3713 opts_cookiefile = self.params.get('cookiefile')
3714 opts_proxy = self.params.get('proxy')
3715
3716 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3717
3718 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3719 if opts_proxy is not None:
3720 if opts_proxy == '':
3721 proxies = {}
3722 else:
3723 proxies = {'http': opts_proxy, 'https': opts_proxy}
3724 else:
3725 proxies = compat_urllib_request.getproxies()
3726 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3727 if 'http' in proxies and 'https' not in proxies:
3728 proxies['https'] = proxies['http']
3729 proxy_handler = PerRequestProxyHandler(proxies)
3730
3731 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3732 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3733 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3734 redirect_handler = YoutubeDLRedirectHandler()
3735 data_handler = compat_urllib_request_DataHandler()
3736
3737 # When passing our own FileHandler instance, build_opener won't add the
3738 # default FileHandler and allows us to disable the file protocol, which
3739 # can be used for malicious purposes (see
3740 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3741 file_handler = compat_urllib_request.FileHandler()
3742
3743 def file_open(*args, **kwargs):
3744 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3745 file_handler.file_open = file_open
3746
3747 opener = compat_urllib_request.build_opener(
3748 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3749
3750 # Delete the default user-agent header, which would otherwise apply in
3751 # cases where our custom HTTP handler doesn't come into play
3752 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3753 opener.addheaders = []
3754 self._opener = opener
3755
3756 def encode(self, s):
3757 if isinstance(s, bytes):
3758 return s # Already encoded
3759
3760 try:
3761 return s.encode(self.get_encoding())
3762 except UnicodeEncodeError as err:
3763 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3764 raise
3765
3766 def get_encoding(self):
3767 encoding = self.params.get('encoding')
3768 if encoding is None:
3769 encoding = preferredencoding()
3770 return encoding
3771
3772 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3773 ''' Write infojson and returns True = written, False = skip, None = error '''
3774 if overwrite is None:
3775 overwrite = self.params.get('overwrites', True)
3776 if not self.params.get('writeinfojson'):
3777 return False
3778 elif not infofn:
3779 self.write_debug(f'Skipping writing {label} infojson')
3780 return False
3781 elif not self._ensure_dir_exists(infofn):
3782 return None
3783 elif not overwrite and os.path.exists(infofn):
3784 self.to_screen(f'[info] {label.title()} metadata is already present')
3785 else:
3786 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3787 try:
3788 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3789 except (OSError, IOError):
3790 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3791 return None
3792 return True
3793
3794 def _write_description(self, label, ie_result, descfn):
3795 ''' Write description and returns True = written, False = skip, None = error '''
3796 if not self.params.get('writedescription'):
3797 return False
3798 elif not descfn:
3799 self.write_debug(f'Skipping writing {label} description')
3800 return False
3801 elif not self._ensure_dir_exists(descfn):
3802 return None
3803 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3804 self.to_screen(f'[info] {label.title()} description is already present')
3805 elif ie_result.get('description') is None:
3806 self.report_warning(f'There\'s no {label} description to write')
3807 return False
3808 else:
3809 try:
3810 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3811 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3812 descfile.write(ie_result['description'])
3813 except (OSError, IOError):
3814 self.report_error(f'Cannot write {label} description file {descfn}')
3815 return None
3816 return True
3817
3818 def _write_subtitles(self, info_dict, filename):
3819 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3820 ret = []
3821 subtitles = info_dict.get('requested_subtitles')
3822 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3823 # subtitles download errors are already managed as troubles in relevant IE
3824 # that way it will silently go on when used with unsupporting IE
3825 return ret
3826
3827 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3828 if not sub_filename_base:
3829 self.to_screen('[info] Skipping writing video subtitles')
3830 return ret
3831 for sub_lang, sub_info in subtitles.items():
3832 sub_format = sub_info['ext']
3833 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3834 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3835 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3836 if existing_sub:
3837 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3838 sub_info['filepath'] = existing_sub
3839 ret.append((existing_sub, sub_filename_final))
3840 continue
3841
3842 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3843 if sub_info.get('data') is not None:
3844 try:
3845 # Use newline='' to prevent conversion of newline characters
3846 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3847 with io.open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3848 subfile.write(sub_info['data'])
3849 sub_info['filepath'] = sub_filename
3850 ret.append((sub_filename, sub_filename_final))
3851 continue
3852 except (OSError, IOError):
3853 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3854 return None
3855
3856 try:
3857 sub_copy = sub_info.copy()
3858 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3859 self.dl(sub_filename, sub_copy, subtitle=True)
3860 sub_info['filepath'] = sub_filename
3861 ret.append((sub_filename, sub_filename_final))
3862 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3863 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3864 raise DownloadError(f'Unable to download video subtitles for {sub_lang!r}: {err}', err)
3865 self.report_warning(f'Unable to download video subtitles for {sub_lang!r}: {err}')
3866 return ret
3867
3868 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3869 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3870 write_all = self.params.get('write_all_thumbnails', False)
3871 thumbnails, ret = [], []
3872 if write_all or self.params.get('writethumbnail', False):
3873 thumbnails = info_dict.get('thumbnails') or []
3874 multiple = write_all and len(thumbnails) > 1
3875
3876 if thumb_filename_base is None:
3877 thumb_filename_base = filename
3878 if thumbnails and not thumb_filename_base:
3879 self.write_debug(f'Skipping writing {label} thumbnail')
3880 return ret
3881
3882 for idx, t in list(enumerate(thumbnails))[::-1]:
3883 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3884 thumb_display_id = f'{label} thumbnail {t["id"]}'
3885 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3886 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3887
3888 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3889 if existing_thumb:
3890 self.to_screen('[info] %s is already present' % (
3891 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3892 t['filepath'] = existing_thumb
3893 ret.append((existing_thumb, thumb_filename_final))
3894 else:
3895 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3896 try:
3897 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3898 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3899 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3900 shutil.copyfileobj(uf, thumbf)
3901 ret.append((thumb_filename, thumb_filename_final))
3902 t['filepath'] = thumb_filename
3903 except network_exceptions as err:
3904 thumbnails.pop(idx)
3905 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3906 if ret and not write_all:
3907 break
3908 return ret