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