4 from __future__
import absolute_import
, unicode_literals
31 from string
import ascii_letters
35 compat_get_terminal_size
,
42 compat_tokenize_tokenize
,
44 compat_urllib_request
,
45 compat_urllib_request_DataHandler
,
46 windows_enable_vt_mode
,
48 from .cookies
import load_cookies
70 format_decimal_suffix
,
91 PerRequestProxyHandler
,
98 register_socks_protocols
,
100 remove_terminal_sequences
,
114 supports_terminal_sequences
,
119 UnavailableVideoError
,
125 YoutubeDLCookieProcessor
,
127 YoutubeDLRedirectHandler
,
129 from .cache
import Cache
130 from .minicurses
import format_text
131 from .extractor
import (
132 gen_extractor_classes
,
135 _PLUGIN_CLASSES
as plugin_extractors
137 from .extractor
.openload
import PhantomJSwrapper
138 from .downloader
import (
140 get_suitable_downloader
,
141 shorten_protocol_name
143 from .downloader
.rtmp
import rtmpdump_version
144 from .postprocessor
import (
147 FFmpegFixupDuplicateMoovPP
,
148 FFmpegFixupDurationPP
,
151 FFmpegFixupStretchedPP
,
152 FFmpegFixupTimestampPP
,
155 MoveFilesAfterDownloadPP
,
156 _PLUGIN_CLASSES
as plugin_postprocessors
158 from .update
import detect_variant
159 from .version
import __version__
, RELEASE_GIT_HEAD
161 if compat_os_name
== 'nt':
165 class YoutubeDL(object):
168 YoutubeDL objects are the ones responsible of downloading the
169 actual video file and writing it to disk if the user has requested
170 it, among some other tasks. In most cases there should be one per
171 program. As, given a video URL, the downloader doesn't know how to
172 extract all the needed information, task that InfoExtractors do, it
173 has to pass the URL to one of them.
175 For this, YoutubeDL objects have a method that allows
176 InfoExtractors to be registered in a given order. When it is passed
177 a URL, the YoutubeDL object handles it to the first InfoExtractor it
178 finds that reports being able to handle it. The InfoExtractor extracts
179 all the information about the video or videos the URL refers to, and
180 YoutubeDL process the extracted information, possibly using a File
181 Downloader to download the video.
183 YoutubeDL objects accept a lot of parameters. In order not to saturate
184 the object constructor with arguments, it receives a dictionary of
185 options instead. These options are available through the params
186 attribute for the InfoExtractors to use. The YoutubeDL also
187 registers itself as the downloader in charge for the InfoExtractors
188 that are added to it, so this is a "mutual registration".
192 username: Username for authentication purposes.
193 password: Password for authentication purposes.
194 videopassword: Password for accessing a video.
195 ap_mso: Adobe Pass multiple-system operator identifier.
196 ap_username: Multiple-system operator account username.
197 ap_password: Multiple-system operator account password.
198 usenetrc: Use netrc for authentication instead.
199 verbose: Print additional info to stdout.
200 quiet: Do not print messages to stdout.
201 no_warnings: Do not print out anything for warnings.
202 forceprint: A dict with keys video/playlist mapped to
203 a list of templates to force print to stdout
204 For compatibility, a single list is also accepted
205 forceurl: Force printing final URL. (Deprecated)
206 forcetitle: Force printing title. (Deprecated)
207 forceid: Force printing ID. (Deprecated)
208 forcethumbnail: Force printing thumbnail URL. (Deprecated)
209 forcedescription: Force printing description. (Deprecated)
210 forcefilename: Force printing final filename. (Deprecated)
211 forceduration: Force printing duration. (Deprecated)
212 forcejson: Force printing info_dict as JSON.
213 dump_single_json: Force printing the info_dict of the whole playlist
214 (or video) as a single JSON line.
215 force_write_download_archive: Force writing download archive regardless
216 of 'skip_download' or 'simulate'.
217 simulate: Do not download the video files. If unset (or None),
218 simulate only if listsubtitles, listformats or list_thumbnails is used
219 format: Video format code. see "FORMAT SELECTION" for more details.
220 You can also pass a function. The function takes 'ctx' as
221 argument and returns the formats to download.
222 See "build_format_selector" for an implementation
223 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
224 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
225 extracting metadata even if the video is not actually
226 available for download (experimental)
227 format_sort: A list of fields by which to sort the video formats.
228 See "Sorting Formats" for more details.
229 format_sort_force: Force the given format_sort. see "Sorting Formats"
231 allow_multiple_video_streams: Allow multiple video streams to be merged
233 allow_multiple_audio_streams: Allow multiple audio streams to be merged
235 check_formats Whether to test if the formats are downloadable.
236 Can be True (check all), False (check none),
237 'selected' (check selected formats),
238 or None (check only if requested by extractor)
239 paths: Dictionary of output paths. The allowed keys are 'home'
240 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
241 outtmpl: Dictionary of templates for output names. Allowed keys
242 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
243 For compatibility with youtube-dl, a single string can also be used
244 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
245 restrictfilenames: Do not allow "&" and spaces in file names
246 trim_file_name: Limit length of filename (extension excluded)
247 windowsfilenames: Force the filenames to be windows compatible
248 ignoreerrors: Do not stop on download/postprocessing errors.
249 Can be 'only_download' to ignore only download errors.
250 Default is 'only_download' for CLI, but False for API
251 skip_playlist_after_errors: Number of allowed failures until the rest of
252 the playlist is skipped
253 force_generic_extractor: Force downloader to use the generic extractor
254 overwrites: Overwrite all video and metadata files if True,
255 overwrite only non-video files if None
256 and don't overwrite any file if False
257 For compatibility with youtube-dl,
258 "nooverwrites" may also be used instead
259 playliststart: Playlist item to start at.
260 playlistend: Playlist item to end at.
261 playlist_items: Specific indices of playlist to download.
262 playlistreverse: Download playlist items in reverse order.
263 playlistrandom: Download playlist items in random order.
264 matchtitle: Download only matching titles.
265 rejecttitle: Reject downloads for matching titles.
266 logger: Log messages to a logging.Logger instance.
267 logtostderr: Log messages to stderr instead of stdout.
268 consoletitle: Display progress in console window's titlebar.
269 writedescription: Write the video description to a .description file
270 writeinfojson: Write the video description to a .info.json file
271 clean_infojson: Remove private fields from the infojson
272 getcomments: Extract video comments. This will not be written to disk
273 unless writeinfojson is also given
274 writeannotations: Write the video annotations to a .annotations.xml file
275 writethumbnail: Write the thumbnail image to a file
276 allow_playlist_files: Whether to write playlists' description, infojson etc
277 also to disk when using the 'write*' options
278 write_all_thumbnails: Write all thumbnail formats to files
279 writelink: Write an internet shortcut file, depending on the
280 current platform (.url/.webloc/.desktop)
281 writeurllink: Write a Windows internet shortcut file (.url)
282 writewebloclink: Write a macOS internet shortcut file (.webloc)
283 writedesktoplink: Write a Linux internet shortcut file (.desktop)
284 writesubtitles: Write the video subtitles to a file
285 writeautomaticsub: Write the automatically generated subtitles to a file
286 allsubtitles: Deprecated - Use subtitleslangs = ['all']
287 Downloads all the subtitles of the video
288 (requires writesubtitles or writeautomaticsub)
289 listsubtitles: Lists all available subtitles for the video
290 subtitlesformat: The format code for subtitles
291 subtitleslangs: List of languages of the subtitles to download (can be regex).
292 The list may contain "all" to refer to all the available
293 subtitles. The language can be prefixed with a "-" to
294 exclude it from the requested languages. Eg: ['all', '-live_chat']
295 keepvideo: Keep the video file after post-processing
296 daterange: A DateRange object, download only if the upload_date is in the range.
297 skip_download: Skip the actual download of the video file
298 cachedir: Location of the cache files in the filesystem.
299 False to disable filesystem cache.
300 noplaylist: Download single video instead of a playlist if in doubt.
301 age_limit: An integer representing the user's age in years.
302 Unsuitable videos for the given age are skipped.
303 min_views: An integer representing the minimum view count the video
304 must have in order to not be skipped.
305 Videos without view count information are always
306 downloaded. None for no limit.
307 max_views: An integer representing the maximum view count.
308 Videos that are more popular than that are not
310 Videos without view count information are always
311 downloaded. None for no limit.
312 download_archive: File name of a file where all downloads are recorded.
313 Videos already present in the file are not downloaded
315 break_on_existing: Stop the download process after attempting to download a
316 file that is in the archive.
317 break_on_reject: Stop the download process when encountering a video that
318 has been filtered out.
319 break_per_url: Whether break_on_reject and break_on_existing
320 should act on each input URL as opposed to for the entire queue
321 cookiefile: File name where cookies should be read from and dumped to
322 cookiesfrombrowser: A tuple containing the name of the browser, the profile
323 name/pathfrom where cookies are loaded, and the name of the
324 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
325 nocheckcertificate: Do not verify SSL certificates
326 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
327 At the moment, this is only supported by YouTube.
328 proxy: URL of the proxy server to use
329 geo_verification_proxy: URL of the proxy to use for IP address verification
330 on geo-restricted sites.
331 socket_timeout: Time to wait for unresponsive hosts, in seconds
332 bidi_workaround: Work around buggy terminals without bidirectional text
333 support, using fridibi
334 debug_printtraffic:Print out sent and received HTTP traffic
335 include_ads: Download ads as well (deprecated)
336 default_search: Prepend this string if an input url is not valid.
337 'auto' for elaborate guessing
338 encoding: Use this encoding instead of the system-specified.
339 extract_flat: Do not resolve URLs, return the immediate result.
340 Pass in 'in_playlist' to only show this behavior for
342 wait_for_video: If given, wait for scheduled streams to become available.
343 The value should be a tuple containing the range
344 (min_secs, max_secs) to wait between retries
345 postprocessors: A list of dictionaries, each with an entry
346 * key: The name of the postprocessor. See
347 yt_dlp/postprocessor/__init__.py for a list.
348 * when: When to run the postprocessor. Can be one of
349 pre_process|before_dl|post_process|after_move.
350 Assumed to be 'post_process' if not given
351 post_hooks: Deprecated - Register a custom postprocessor instead
352 A list of functions that get called as the final step
353 for each video file, after all postprocessors have been
354 called. The filename will be passed as the only argument.
355 progress_hooks: A list of functions that get called on download
356 progress, with a dictionary with the entries
357 * status: One of "downloading", "error", or "finished".
358 Check this first and ignore unknown values.
359 * info_dict: The extracted info_dict
361 If status is one of "downloading", or "finished", the
362 following properties may also be present:
363 * filename: The final filename (always present)
364 * tmpfilename: The filename we're currently writing to
365 * downloaded_bytes: Bytes on disk
366 * total_bytes: Size of the whole file, None if unknown
367 * total_bytes_estimate: Guess of the eventual file size,
369 * elapsed: The number of seconds since download started.
370 * eta: The estimated time in seconds, None if unknown
371 * speed: The download speed in bytes/second, None if
373 * fragment_index: The counter of the currently
374 downloaded video fragment.
375 * fragment_count: The number of fragments (= individual
376 files that will be merged)
378 Progress hooks are guaranteed to be called at least once
379 (with status "finished") if the download is successful.
380 postprocessor_hooks: A list of functions that get called on postprocessing
381 progress, with a dictionary with the entries
382 * status: One of "started", "processing", or "finished".
383 Check this first and ignore unknown values.
384 * postprocessor: Name of the postprocessor
385 * info_dict: The extracted info_dict
387 Progress hooks are guaranteed to be called at least twice
388 (with status "started" and "finished") if the processing is successful.
389 merge_output_format: Extension to use when merging formats.
390 final_ext: Expected final extension; used to detect when the file was
391 already downloaded and converted
392 fixup: Automatically correct known faults of the file.
394 - "never": do nothing
395 - "warn": only emit a warning
396 - "detect_or_warn": check whether we can do anything
397 about it, warn otherwise (default)
398 source_address: Client-side IP address to bind to.
399 call_home: Boolean, true iff we are allowed to contact the
400 yt-dlp servers for debugging. (BROKEN)
401 sleep_interval_requests: Number of seconds to sleep between requests
403 sleep_interval: Number of seconds to sleep before each download when
404 used alone or a lower bound of a range for randomized
405 sleep before each download (minimum possible number
406 of seconds to sleep) when used along with
408 max_sleep_interval:Upper bound of a range for randomized sleep before each
409 download (maximum possible number of seconds to sleep).
410 Must only be used along with sleep_interval.
411 Actual sleep time will be a random float from range
412 [sleep_interval; max_sleep_interval].
413 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
414 listformats: Print an overview of available video formats and exit.
415 list_thumbnails: Print a table of all thumbnails and exit.
416 match_filter: A function that gets called with the info_dict of
418 If it returns a message, the video is ignored.
419 If it returns None, the video is downloaded.
420 match_filter_func in utils.py is one example for this.
421 no_color: Do not emit color codes in output.
422 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
425 Two-letter ISO 3166-2 country code that will be used for
426 explicit geographic restriction bypassing via faking
427 X-Forwarded-For HTTP header
429 IP range in CIDR notation that will be used similarly to
432 The following options determine which downloader is picked:
433 external_downloader: A dictionary of protocol keys and the executable of the
434 external downloader to use for it. The allowed protocols
435 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
436 Set the value to 'native' to use the native downloader
437 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
438 or {'m3u8': 'ffmpeg'} instead.
439 Use the native HLS downloader instead of ffmpeg/avconv
440 if True, otherwise use ffmpeg/avconv if False, otherwise
441 use downloader suggested by extractor if None.
442 compat_opts: Compatibility options. See "Differences in default behavior".
443 The following options do not work when used through the API:
444 filename, abort-on-error, multistreams, no-live-chat, format-sort
445 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
446 Refer __init__.py for their implementation
447 progress_template: Dictionary of templates for progress outputs.
448 Allowed keys are 'download', 'postprocess',
449 'download-title' (console title) and 'postprocess-title'.
450 The template is mapped on a dictionary with keys 'progress' and 'info'
452 The following parameters are not used by YoutubeDL itself, they are used by
453 the downloader (see yt_dlp/downloader/common.py):
454 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
455 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
456 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
457 external_downloader_args, concurrent_fragment_downloads.
459 The following options are used by the post processors:
460 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
461 otherwise prefer ffmpeg. (avconv support is deprecated)
462 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
463 to the binary or its containing directory.
464 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
465 and a list of additional command-line arguments for the
466 postprocessor/executable. The dict can also have "PP+EXE" keys
467 which are used when the given exe is used by the given PP.
468 Use 'default' as the name for arguments to passed to all PP
469 For compatibility with youtube-dl, a single list of args
472 The following options are used by the extractors:
473 extractor_retries: Number of times to retry for known errors
474 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
475 hls_split_discontinuity: Split HLS playlists to different formats at
476 discontinuities such as ad breaks (default: False)
477 extractor_args: A dictionary of arguments to be passed to the extractors.
478 See "EXTRACTOR ARGUMENTS" for details.
479 Eg: {'youtube': {'skip': ['dash', 'hls']}}
480 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
481 If True (default), DASH manifests and related
482 data will be downloaded and processed by extractor.
483 You can reduce network I/O by disabling it if you don't
484 care about DASH. (only for youtube)
485 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
486 If True (default), HLS manifests and related
487 data will be downloaded and processed by extractor.
488 You can reduce network I/O by disabling it if you don't
489 care about HLS. (only for youtube)
492 _NUMERIC_FIELDS
= set((
493 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
494 'timestamp', 'release_timestamp',
495 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
496 'average_rating', 'comment_count', 'age_limit',
497 'start_time', 'end_time',
498 'chapter_number', 'season_number', 'episode_number',
499 'track_number', 'disc_number', 'release_year',
502 _format_selection_exts
= {
503 'audio': {'m4a', 'mp3', 'ogg', 'aac'}
,
504 'video': {'mp4', 'flv', 'webm', '3gp'}
,
505 'storyboards': {'mhtml'}
,
510 _pps
= {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
511 _printed_messages
= set()
512 _first_webpage_request
= True
513 _download_retcode
= None
514 _num_downloads
= None
516 _playlist_urls
= set()
519 def __init__(self
, params
=None, auto_init
=True):
520 """Create a FileDownloader object with the given options.
521 @param auto_init Whether to load the default extractors and print header (if verbose).
522 Set to 'no_verbose_header' to not print the header
527 self
._ies
_instances
= {}
528 self
._pps
= {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
529 self
._printed
_messages
= set()
530 self
._first
_webpage
_request
= True
531 self
._post
_hooks
= []
532 self
._progress
_hooks
= []
533 self
._postprocessor
_hooks
= []
534 self
._download
_retcode
= 0
535 self
._num
_downloads
= 0
536 self
._screen
_file
= [sys
.stdout
, sys
.stderr
][params
.get('logtostderr', False)]
537 self
._err
_file
= sys
.stderr
539 self
.cache
= Cache(self
)
541 windows_enable_vt_mode()
542 self
._allow
_colors
= {
543 'screen': not self
.params
.get('no_color') and supports_terminal_sequences(self
._screen
_file
),
544 'err': not self
.params
.get('no_color') and supports_terminal_sequences(self
._err
_file
),
547 if sys
.version_info
< (3, 6):
549 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys
.version_info
[:2])
551 if self
.params
.get('allow_unplayable_formats'):
553 f
'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
554 'This is a developer option intended for debugging. \n'
555 ' If you experience any issues while using this option, '
556 f
'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
558 def check_deprecated(param
, option
, suggestion
):
559 if self
.params
.get(param
) is not None:
560 self
.report_warning('%s is deprecated. Use %s instead' % (option
, suggestion
))
564 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
565 if self
.params
.get('geo_verification_proxy') is None:
566 self
.params
['geo_verification_proxy'] = self
.params
['cn_verification_proxy']
568 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
569 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
570 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
572 for msg
in self
.params
.get('_warnings', []):
573 self
.report_warning(msg
)
574 for msg
in self
.params
.get('_deprecation_warnings', []):
575 self
.deprecation_warning(msg
)
577 if 'list-formats' in self
.params
.get('compat_opts', []):
578 self
.params
['listformats_table'] = False
580 if 'overwrites' not in self
.params
and self
.params
.get('nooverwrites') is not None:
581 # nooverwrites was unnecessarily changed to overwrites
582 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
583 # This ensures compatibility with both keys
584 self
.params
['overwrites'] = not self
.params
['nooverwrites']
585 elif self
.params
.get('overwrites') is None:
586 self
.params
.pop('overwrites', None)
588 self
.params
['nooverwrites'] = not self
.params
['overwrites']
590 # Compatibility with older syntax
591 params
.setdefault('forceprint', {})
592 if not isinstance(params
['forceprint'], dict):
593 params
['forceprint'] = {'video': params['forceprint']}
595 if params
.get('bidi_workaround', False):
598 master
, slave
= pty
.openpty()
599 width
= compat_get_terminal_size().columns
603 width_args
= ['-w', str(width
)]
605 stdin
=subprocess
.PIPE
,
607 stderr
=self
._err
_file
)
609 self
._output
_process
= Popen(['bidiv'] + width_args
, **sp_kwargs
)
611 self
._output
_process
= Popen(['fribidi', '-c', 'UTF-8'] + width_args
, **sp_kwargs
)
612 self
._output
_channel
= os
.fdopen(master
, 'rb')
613 except OSError as ose
:
614 if ose
.errno
== errno
.ENOENT
:
616 'Could not find fribidi executable, ignoring --bidi-workaround. '
617 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
621 if (sys
.platform
!= 'win32'
622 and sys
.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
623 and not params
.get('restrictfilenames', False)):
624 # Unicode filesystem API will throw errors (#1474, #13027)
626 'Assuming --restrict-filenames since file system encoding '
627 'cannot encode all characters. '
628 'Set the LC_ALL environment variable to fix this.')
629 self
.params
['restrictfilenames'] = True
631 self
.outtmpl_dict
= self
.parse_outtmpl()
633 # Creating format selector here allows us to catch syntax errors before the extraction
634 self
.format_selector
= (
635 self
.params
.get('format') if self
.params
.get('format') in (None, '-')
636 else self
.params
['format'] if callable(self
.params
['format'])
637 else self
.build_format_selector(self
.params
['format']))
642 if auto_init
!= 'no_verbose_header':
643 self
.print_debug_header()
644 self
.add_default_info_extractors()
647 'post_hooks': self
.add_post_hook
,
648 'progress_hooks': self
.add_progress_hook
,
649 'postprocessor_hooks': self
.add_postprocessor_hook
,
651 for opt
, fn
in hooks
.items():
652 for ph
in self
.params
.get(opt
, []):
655 for pp_def_raw
in self
.params
.get('postprocessors', []):
656 pp_def
= dict(pp_def_raw
)
657 when
= pp_def
.pop('when', 'post_process')
658 self
.add_post_processor(
659 get_postprocessor(pp_def
.pop('key'))(self
, **compat_kwargs(pp_def
)),
662 register_socks_protocols()
664 def preload_download_archive(fn
):
665 """Preload the archive, if any is specified"""
668 self
.write_debug(f
'Loading archive file {fn!r}')
670 with locked_file(fn
, 'r', encoding
='utf-8') as archive_file
:
671 for line
in archive_file
:
672 self
.archive
.add(line
.strip())
673 except IOError as ioe
:
674 if ioe
.errno
!= errno
.ENOENT
:
680 preload_download_archive(self
.params
.get('download_archive'))
682 def warn_if_short_id(self
, argv
):
683 # short YouTube ID starting with dash?
685 i
for i
, a
in enumerate(argv
)
686 if re
.match(r
'^-[0-9A-Za-z_-]{10}$', a
)]
690 + [a
for i
, a
in enumerate(argv
) if i
not in idxs
]
691 + ['--'] + [argv
[i
] for i
in idxs
]
694 'Long argument string detected. '
695 'Use -- to separate parameters and URLs, like this:\n%s' %
696 args_to_str(correct_argv
))
698 def add_info_extractor(self
, ie
):
699 """Add an InfoExtractor object to the end of the list."""
701 self
._ies
[ie_key
] = ie
702 if not isinstance(ie
, type):
703 self
._ies
_instances
[ie_key
] = ie
704 ie
.set_downloader(self
)
706 def _get_info_extractor_class(self
, ie_key
):
707 ie
= self
._ies
.get(ie_key
)
709 ie
= get_info_extractor(ie_key
)
710 self
.add_info_extractor(ie
)
713 def get_info_extractor(self
, ie_key
):
715 Get an instance of an IE with name ie_key, it will try to get one from
716 the _ies list, if there's no instance it will create a new one and add
717 it to the extractor list.
719 ie
= self
._ies
_instances
.get(ie_key
)
721 ie
= get_info_extractor(ie_key
)()
722 self
.add_info_extractor(ie
)
725 def add_default_info_extractors(self
):
727 Add the InfoExtractors returned by gen_extractors to the end of the list
729 for ie
in gen_extractor_classes():
730 self
.add_info_extractor(ie
)
732 def add_post_processor(self
, pp
, when
='post_process'):
733 """Add a PostProcessor object to the end of the chain."""
734 self
._pps
[when
].append(pp
)
735 pp
.set_downloader(self
)
737 def add_post_hook(self
, ph
):
738 """Add the post hook"""
739 self
._post
_hooks
.append(ph
)
741 def add_progress_hook(self
, ph
):
742 """Add the download progress hook"""
743 self
._progress
_hooks
.append(ph
)
745 def add_postprocessor_hook(self
, ph
):
746 """Add the postprocessing progress hook"""
747 self
._postprocessor
_hooks
.append(ph
)
748 for pps
in self
._pps
.values():
750 pp
.add_progress_hook(ph
)
752 def _bidi_workaround(self
, message
):
753 if not hasattr(self
, '_output_channel'):
756 assert hasattr(self
, '_output_process')
757 assert isinstance(message
, compat_str
)
758 line_count
= message
.count('\n') + 1
759 self
._output
_process
.stdin
.write((message
+ '\n').encode('utf-8'))
760 self
._output
_process
.stdin
.flush()
761 res
= ''.join(self
._output
_channel
.readline().decode('utf-8')
762 for _
in range(line_count
))
763 return res
[:-len('\n')]
765 def _write_string(self
, message
, out
=None, only_once
=False):
767 if message
in self
._printed
_messages
:
769 self
._printed
_messages
.add(message
)
770 write_string(message
, out
=out
, encoding
=self
.params
.get('encoding'))
772 def to_stdout(self
, message
, skip_eol
=False, quiet
=False):
773 """Print message to stdout"""
774 if self
.params
.get('logger'):
775 self
.params
['logger'].debug(message
)
776 elif not quiet
or self
.params
.get('verbose'):
778 '%s%s' % (self
._bidi
_workaround
(message
), ('' if skip_eol
else '\n')),
779 self
._err
_file
if quiet
else self
._screen
_file
)
781 def to_stderr(self
, message
, only_once
=False):
782 """Print message to stderr"""
783 assert isinstance(message
, compat_str
)
784 if self
.params
.get('logger'):
785 self
.params
['logger'].error(message
)
787 self
._write
_string
('%s\n' % self
._bidi
_workaround
(message
), self
._err
_file
, only_once
=only_once
)
789 def to_console_title(self
, message
):
790 if not self
.params
.get('consoletitle', False):
792 message
= remove_terminal_sequences(message
)
793 if compat_os_name
== 'nt':
794 if ctypes
.windll
.kernel32
.GetConsoleWindow():
795 # c_wchar_p() might not be necessary if `message` is
796 # already of type unicode()
797 ctypes
.windll
.kernel32
.SetConsoleTitleW(ctypes
.c_wchar_p(message
))
798 elif 'TERM' in os
.environ
:
799 self
._write
_string
('\033]0;%s\007' % message
, self
._screen
_file
)
801 def save_console_title(self
):
802 if not self
.params
.get('consoletitle', False):
804 if self
.params
.get('simulate'):
806 if compat_os_name
!= 'nt' and 'TERM' in os
.environ
:
807 # Save the title on stack
808 self
._write
_string
('\033[22;0t', self
._screen
_file
)
810 def restore_console_title(self
):
811 if not self
.params
.get('consoletitle', False):
813 if self
.params
.get('simulate'):
815 if compat_os_name
!= 'nt' and 'TERM' in os
.environ
:
816 # Restore the title from stack
817 self
._write
_string
('\033[23;0t', self
._screen
_file
)
820 self
.save_console_title()
823 def __exit__(self
, *args
):
824 self
.restore_console_title()
826 if self
.params
.get('cookiefile') is not None:
827 self
.cookiejar
.save(ignore_discard
=True, ignore_expires
=True)
829 def trouble(self
, message
=None, tb
=None, is_error
=True):
830 """Determine action to take when a download problem appears.
832 Depending on if the downloader has been configured to ignore
833 download errors or not, this method may throw an exception or
834 not when errors are found, after printing the message.
836 @param tb If given, is additional traceback information
837 @param is_error Whether to raise error according to ignorerrors
839 if message
is not None:
840 self
.to_stderr(message
)
841 if self
.params
.get('verbose'):
843 if sys
.exc_info()[0]: # if .trouble has been called from an except block
845 if hasattr(sys
.exc_info()[1], 'exc_info') and sys
.exc_info()[1].exc_info
[0]:
846 tb
+= ''.join(traceback
.format_exception(*sys
.exc_info()[1].exc_info
))
847 tb
+= encode_compat_str(traceback
.format_exc())
849 tb_data
= traceback
.format_list(traceback
.extract_stack())
850 tb
= ''.join(tb_data
)
855 if not self
.params
.get('ignoreerrors'):
856 if sys
.exc_info()[0] and hasattr(sys
.exc_info()[1], 'exc_info') and sys
.exc_info()[1].exc_info
[0]:
857 exc_info
= sys
.exc_info()[1].exc_info
859 exc_info
= sys
.exc_info()
860 raise DownloadError(message
, exc_info
)
861 self
._download
_retcode
= 1
863 def to_screen(self
, message
, skip_eol
=False):
864 """Print message to stdout if not in quiet mode"""
866 message
, skip_eol
, quiet
=self
.params
.get('quiet', False))
870 EMPHASIS
= 'light blue'
875 SUPPRESS
= 'light black'
877 def _format_text(self
, handle
, allow_colors
, text
, f
, fallback
=None, *, test_encoding
=False):
880 encoding
= self
.params
.get('encoding') or getattr(handle
, 'encoding', 'ascii')
881 text
= text
.encode(encoding
, 'ignore').decode(encoding
)
882 if fallback
is not None and text
!= original_text
:
884 if isinstance(f
, self
.Styles
):
886 return format_text(text
, f
) if allow_colors
else text
if fallback
is None else fallback
888 def _format_screen(self
, *args
, **kwargs
):
889 return self
._format
_text
(
890 self
._screen
_file
, self
._allow
_colors
['screen'], *args
, **kwargs
)
892 def _format_err(self
, *args
, **kwargs
):
893 return self
._format
_text
(
894 self
._err
_file
, self
._allow
_colors
['err'], *args
, **kwargs
)
896 def report_warning(self
, message
, only_once
=False):
898 Print the message to stderr, it will be prefixed with 'WARNING:'
899 If stderr is a tty file the 'WARNING:' will be colored
901 if self
.params
.get('logger') is not None:
902 self
.params
['logger'].warning(message
)
904 if self
.params
.get('no_warnings'):
906 self
.to_stderr(f
'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once
)
908 def deprecation_warning(self
, message
):
909 if self
.params
.get('logger') is not None:
910 self
.params
['logger'].warning('DeprecationWarning: {message}')
912 self
.to_stderr(f
'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
914 def report_error(self
, message
, *args
, **kwargs
):
916 Do the same as trouble, but prefixes the message with 'ERROR:', colored
917 in red if stderr is a tty file.
919 self
.trouble(f
'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args
, **kwargs
)
921 def write_debug(self
, message
, only_once
=False):
922 '''Log debug message or Print message to stderr'''
923 if not self
.params
.get('verbose', False):
925 message
= '[debug] %s' % message
926 if self
.params
.get('logger'):
927 self
.params
['logger'].debug(message
)
929 self
.to_stderr(message
, only_once
)
931 def report_file_already_downloaded(self
, file_name
):
932 """Report file has already been fully downloaded."""
934 self
.to_screen('[download] %s has already been downloaded' % file_name
)
935 except UnicodeEncodeError:
936 self
.to_screen('[download] The file has already been downloaded')
938 def report_file_delete(self
, file_name
):
939 """Report that existing file will be deleted."""
941 self
.to_screen('Deleting existing file %s' % file_name
)
942 except UnicodeEncodeError:
943 self
.to_screen('Deleting existing file')
945 def raise_no_formats(self
, info
, forced
=False):
946 has_drm
= info
.get('__has_drm')
947 msg
= 'This video is DRM protected' if has_drm
else 'No video formats found!'
948 expected
= self
.params
.get('ignore_no_formats_error')
949 if forced
or not expected
:
950 raise ExtractorError(msg
, video_id
=info
['id'], ie
=info
['extractor'],
951 expected
=has_drm
or expected
)
953 self
.report_warning(msg
)
955 def parse_outtmpl(self
):
956 outtmpl_dict
= self
.params
.get('outtmpl', {})
957 if not isinstance(outtmpl_dict
, dict):
958 outtmpl_dict
= {'default': outtmpl_dict}
959 # Remove spaces in the default template
960 if self
.params
.get('restrictfilenames'):
961 sanitize
= lambda x
: x
.replace(' - ', ' ').replace(' ', '-')
963 sanitize
= lambda x
: x
964 outtmpl_dict
.update({
965 k
: sanitize(v
) for k
, v
in DEFAULT_OUTTMPL
.items()
966 if outtmpl_dict
.get(k
) is None})
967 for key
, val
in outtmpl_dict
.items():
968 if isinstance(val
, bytes):
970 'Parameter outtmpl is bytes, but should be a unicode string. '
971 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
974 def get_output_path(self
, dir_type
='', filename
=None):
975 paths
= self
.params
.get('paths', {})
976 assert isinstance(paths
, dict)
978 expand_path(paths
.get('home', '').strip()),
979 expand_path(paths
.get(dir_type
, '').strip()) if dir_type
else '',
982 # Temporary fix for #4787
983 # 'Treat' all problem characters by passing filename through preferredencoding
984 # to workaround encoding issues with subprocess on python2 @ Windows
985 if sys
.version_info
< (3, 0) and sys
.platform
== 'win32':
986 path
= encodeFilename(path
, True).decode(preferredencoding())
987 return sanitize_path(path
, force
=self
.params
.get('windowsfilenames'))
990 def _outtmpl_expandpath(outtmpl
):
991 # expand_path translates '%%' into '%' and '$$' into '$'
992 # correspondingly that is not what we want since we need to keep
993 # '%%' intact for template dict substitution step. Working around
994 # with boundary-alike separator hack.
995 sep
= ''.join([random
.choice(ascii_letters
) for _
in range(32)])
996 outtmpl
= outtmpl
.replace('%%', '%{0}%'.format(sep
)).replace('$$', '${0}$'.format(sep
))
998 # outtmpl should be expand_path'ed before template dict substitution
999 # because meta fields may contain env variables we don't want to
1000 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1001 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1002 return expand_path(outtmpl
).replace(sep
, '')
1005 def escape_outtmpl(outtmpl
):
1006 ''' Escape any remaining strings like %s, %abc% etc. '''
1008 STR_FORMAT_RE_TMPL
.format('', '(?![%(\0])'),
1009 lambda mobj
: ('' if mobj
.group('has_key') else '%') + mobj
.group(0),
1013 def validate_outtmpl(cls
, outtmpl
):
1014 ''' @return None or Exception object '''
1016 STR_FORMAT_RE_TMPL
.format('[^)]*', '[ljqBUDS]'),
1017 lambda mobj
: f
'{mobj.group(0)[:-1]}s',
1018 cls
._outtmpl
_expandpath
(outtmpl
))
1020 cls
.escape_outtmpl(outtmpl
) % collections
.defaultdict(int)
1022 except ValueError as err
:
1026 def _copy_infodict(info_dict
):
1027 info_dict
= dict(info_dict
)
1028 for key
in ('__original_infodict', '__postprocessors'):
1029 info_dict
.pop(key
, None)
1032 def prepare_outtmpl(self
, outtmpl
, info_dict
, sanitize
=False):
1033 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1034 @param sanitize Whether to sanitize the output as a filename.
1035 For backward compatibility, a function can also be passed
1038 info_dict
.setdefault('epoch', int(time
.time())) # keep epoch consistent once set
1040 info_dict
= self
._copy
_infodict
(info_dict
)
1041 info_dict
['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1042 formatSeconds(info_dict
['duration'], '-' if sanitize
else ':')
1043 if info_dict
.get('duration', None) is not None
1045 info_dict
['autonumber'] = self
.params
.get('autonumber_start', 1) - 1 + self
._num
_downloads
1046 if info_dict
.get('resolution') is None:
1047 info_dict
['resolution'] = self
.format_resolution(info_dict
, default
=None)
1049 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1050 # of %(field)s to %(field)0Nd for backward compatibility
1051 field_size_compat_map
= {
1052 'playlist_index': number_of_digits(info_dict
.get('_last_playlist_index') or 0),
1053 'playlist_autonumber': number_of_digits(info_dict
.get('n_entries') or 0),
1054 'autonumber': self
.params
.get('autonumber_size') or 5,
1058 EXTERNAL_FORMAT_RE
= re
.compile(STR_FORMAT_RE_TMPL
.format('[^)]*', f
'[{STR_FORMAT_TYPES}ljqBUDS]'))
1063 # Field is of the form key1.key2...
1064 # where keys (except first) can be string, int or slice
1065 FIELD_RE
= r
'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num
=r
'(?:-?\d+)')
1066 MATH_FIELD_RE
= r
'''(?:{field}|{num})'''.format(field
=FIELD_RE
, num
=r
'-?\d+(?:.\d+)?')
1067 MATH_OPERATORS_RE
= r
'(?:%s)' % '|'.join(map(re
.escape
, MATH_FUNCTIONS
.keys()))
1068 INTERNAL_FORMAT_RE
= re
.compile(r
'''(?x)
1071 (?P<maths>(?:{math_op}{math_field})*)
1072 (?:>(?P<strf_format>.+?))?
1073 (?P<alternate>(?<!\\),[^|&)]+)?
1074 (?:&(?P<replacement>.*?))?
1075 (?:\|(?P<default>.*?))?
1076 $'''.format(field
=FIELD_RE
, math_op
=MATH_OPERATORS_RE
, math_field
=MATH_FIELD_RE
))
1078 def _traverse_infodict(k
):
1082 return traverse_obj(info_dict
, k
, is_user_input
=True, traverse_string
=True)
1084 def get_value(mdict
):
1086 value
= _traverse_infodict(mdict
['fields'])
1089 value
= float_or_none(value
)
1090 if value
is not None:
1093 offset_key
= mdict
['maths']
1095 value
= float_or_none(value
)
1099 MATH_FIELD_RE
if operator
else MATH_OPERATORS_RE
,
1100 offset_key
).group(0)
1101 offset_key
= offset_key
[len(item
):]
1102 if operator
is None:
1103 operator
= MATH_FUNCTIONS
[item
]
1105 item
, multiplier
= (item
[1:], -1) if item
[0] == '-' else (item
, 1)
1106 offset
= float_or_none(item
)
1108 offset
= float_or_none(_traverse_infodict(item
))
1110 value
= operator(value
, multiplier
* offset
)
1111 except (TypeError, ZeroDivisionError):
1114 # Datetime formatting
1115 if mdict
['strf_format']:
1116 value
= strftime_or_none(value
, mdict
['strf_format'].replace('\\,', ','))
1120 na
= self
.params
.get('outtmpl_na_placeholder', 'NA')
1122 def filename_sanitizer(key
, value
, restricted
=self
.params
.get('restrictfilenames')):
1123 return sanitize_filename(str(value
), restricted
=restricted
,
1124 is_id
=re
.search(r
'(^|[_.])id(\.|$)', key
))
1126 sanitizer
= sanitize
if callable(sanitize
) else filename_sanitizer
1127 sanitize
= bool(sanitize
)
1129 def _dumpjson_default(obj
):
1130 if isinstance(obj
, (set, LazyList
)):
1134 def create_key(outer_mobj
):
1135 if not outer_mobj
.group('has_key'):
1136 return outer_mobj
.group(0)
1137 key
= outer_mobj
.group('key')
1138 mobj
= re
.match(INTERNAL_FORMAT_RE
, key
)
1139 initial_field
= mobj
.group('fields') if mobj
else ''
1140 value
, replacement
, default
= None, None, na
1142 mobj
= mobj
.groupdict()
1143 default
= mobj
['default'] if mobj
['default'] is not None else default
1144 value
= get_value(mobj
)
1145 replacement
= mobj
['replacement']
1146 if value
is None and mobj
['alternate']:
1147 mobj
= re
.match(INTERNAL_FORMAT_RE
, mobj
['alternate'][1:])
1151 fmt
= outer_mobj
.group('format')
1152 if fmt
== 's' and value
is not None and key
in field_size_compat_map
.keys():
1153 fmt
= '0{:d}d'.format(field_size_compat_map
[key
])
1155 value
= default
if value
is None else value
if replacement
is None else replacement
1157 flags
= outer_mobj
.group('conversion') or ''
1158 str_fmt
= f
'{fmt[:-1]}s'
1159 if fmt
[-1] == 'l': # list
1160 delim
= '\n' if '#' in flags
else ', '
1161 value
, fmt
= delim
.join(variadic(value
, allowed_types
=(str, bytes))), str_fmt
1162 elif fmt
[-1] == 'j': # json
1163 value
, fmt
= json
.dumps(value
, default
=_dumpjson_default
, indent
=4 if '#' in flags
else None), str_fmt
1164 elif fmt
[-1] == 'q': # quoted
1165 value
= map(str, variadic(value
) if '#' in flags
else [value
])
1166 value
, fmt
= ' '.join(map(compat_shlex_quote
, value
)), str_fmt
1167 elif fmt
[-1] == 'B': # bytes
1168 value
= f
'%{str_fmt}'.encode('utf-8') % str(value
).encode('utf-8')
1169 value
, fmt
= value
.decode('utf-8', 'ignore'), 's'
1170 elif fmt
[-1] == 'U': # unicode normalized
1171 value
, fmt
= unicodedata
.normalize(
1172 # "+" = compatibility equivalence, "#" = NFD
1173 'NF%s%s' % ('K' if '+' in flags
else '', 'D' if '#' in flags
else 'C'),
1175 elif fmt
[-1] == 'D': # decimal suffix
1176 num_fmt
, fmt
= fmt
[:-1].replace('#', ''), 's'
1177 value
= format_decimal_suffix(value
, f
'%{num_fmt}f%s' if num_fmt
else '%d%s',
1178 factor
=1024 if '#' in flags
else 1000)
1179 elif fmt
[-1] == 'S': # filename sanitization
1180 value
, fmt
= filename_sanitizer(initial_field
, value
, restricted
='#' in flags
), str_fmt
1181 elif fmt
[-1] == 'c':
1183 value
= str(value
)[0]
1186 elif fmt
[-1] not in 'rs': # numeric
1187 value
= float_or_none(value
)
1189 value
, fmt
= default
, 's'
1193 # If value is an object, sanitize might convert it to a string
1194 # So we convert it to repr first
1195 value
, fmt
= repr(value
), str_fmt
1196 if fmt
[-1] in 'csr':
1197 value
= sanitizer(initial_field
, value
)
1199 key
= '%s\0%s' % (key
.replace('%', '%\0'), outer_mobj
.group('format'))
1200 TMPL_DICT
[key
] = value
1201 return '{prefix}%({key}){fmt}'.format(key
=key
, fmt
=fmt
, prefix
=outer_mobj
.group('prefix'))
1203 return EXTERNAL_FORMAT_RE
.sub(create_key
, outtmpl
), TMPL_DICT
1205 def evaluate_outtmpl(self
, outtmpl
, info_dict
, *args
, **kwargs
):
1206 outtmpl
, info_dict
= self
.prepare_outtmpl(outtmpl
, info_dict
, *args
, **kwargs
)
1207 return self
.escape_outtmpl(outtmpl
) % info_dict
1209 def _prepare_filename(self
, info_dict
, tmpl_type
='default'):
1211 outtmpl
= self
._outtmpl
_expandpath
(self
.outtmpl_dict
.get(tmpl_type
, self
.outtmpl_dict
['default']))
1212 filename
= self
.evaluate_outtmpl(outtmpl
, info_dict
, True)
1214 force_ext
= OUTTMPL_TYPES
.get(tmpl_type
)
1215 if filename
and force_ext
is not None:
1216 filename
= replace_extension(filename
, force_ext
, info_dict
.get('ext'))
1218 # https://github.com/blackjack4494/youtube-dlc/issues/85
1219 trim_file_name
= self
.params
.get('trim_file_name', False)
1221 no_ext
, *ext
= filename
.rsplit('.', 2)
1222 filename
= join_nonempty(no_ext
[:trim_file_name
], *ext
, delim
='.')
1225 except ValueError as err
:
1226 self
.report_error('Error in output template: ' + str(err
) + ' (encoding: ' + repr(preferredencoding()) + ')')
1229 def prepare_filename(self
, info_dict
, dir_type
='', warn
=False):
1230 """Generate the output filename."""
1232 filename
= self
._prepare
_filename
(info_dict
, dir_type
or 'default')
1233 if not filename
and dir_type
not in ('', 'temp'):
1237 if not self
.params
.get('paths'):
1239 elif filename
== '-':
1240 self
.report_warning('--paths is ignored when an outputting to stdout', only_once
=True)
1241 elif os
.path
.isabs(filename
):
1242 self
.report_warning('--paths is ignored since an absolute path is given in output template', only_once
=True)
1243 if filename
== '-' or not filename
:
1246 return self
.get_output_path(dir_type
, filename
)
1248 def _match_entry(self
, info_dict
, incomplete
=False, silent
=False):
1249 """ Returns None if the file should be downloaded """
1251 video_title
= info_dict
.get('title', info_dict
.get('id', 'video'))
1254 if 'title' in info_dict
:
1255 # This can happen when we're just evaluating the playlist
1256 title
= info_dict
['title']
1257 matchtitle
= self
.params
.get('matchtitle', False)
1259 if not re
.search(matchtitle
, title
, re
.IGNORECASE
):
1260 return '"' + title
+ '" title did not match pattern "' + matchtitle
+ '"'
1261 rejecttitle
= self
.params
.get('rejecttitle', False)
1263 if re
.search(rejecttitle
, title
, re
.IGNORECASE
):
1264 return '"' + title
+ '" title matched reject pattern "' + rejecttitle
+ '"'
1265 date
= info_dict
.get('upload_date')
1266 if date
is not None:
1267 dateRange
= self
.params
.get('daterange', DateRange())
1268 if date
not in dateRange
:
1269 return '%s upload date is not in range %s' % (date_from_str(date
).isoformat(), dateRange
)
1270 view_count
= info_dict
.get('view_count')
1271 if view_count
is not None:
1272 min_views
= self
.params
.get('min_views')
1273 if min_views
is not None and view_count
< min_views
:
1274 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title
, view_count
, min_views
)
1275 max_views
= self
.params
.get('max_views')
1276 if max_views
is not None and view_count
> max_views
:
1277 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title
, view_count
, max_views
)
1278 if age_restricted(info_dict
.get('age_limit'), self
.params
.get('age_limit')):
1279 return 'Skipping "%s" because it is age restricted' % video_title
1281 match_filter
= self
.params
.get('match_filter')
1282 if match_filter
is not None:
1284 ret
= match_filter(info_dict
, incomplete
=incomplete
)
1286 # For backward compatibility
1287 ret
= None if incomplete
else match_filter(info_dict
)
1292 if self
.in_download_archive(info_dict
):
1293 reason
= '%s has already been recorded in the archive' % video_title
1294 break_opt
, break_err
= 'break_on_existing', ExistingVideoReached
1296 reason
= check_filter()
1297 break_opt
, break_err
= 'break_on_reject', RejectedVideoReached
1298 if reason
is not None:
1300 self
.to_screen('[download] ' + reason
)
1301 if self
.params
.get(break_opt
, False):
1306 def add_extra_info(info_dict
, extra_info
):
1307 '''Set the keys from extra_info in info dict if they are missing'''
1308 for key
, value
in extra_info
.items():
1309 info_dict
.setdefault(key
, value
)
1311 def extract_info(self
, url
, download
=True, ie_key
=None, extra_info
=None,
1312 process
=True, force_generic_extractor
=False):
1314 Return a list with a dictionary for each video extracted.
1317 url -- URL to extract
1320 download -- whether to download videos during extraction
1321 ie_key -- extractor key hint
1322 extra_info -- dictionary containing the extra values to add to each result
1323 process -- whether to resolve all unresolved references (URLs, playlist items),
1324 must be True for download to work.
1325 force_generic_extractor -- force using the generic extractor
1328 if extra_info
is None:
1331 if not ie_key
and force_generic_extractor
:
1335 ies
= {ie_key: self._get_info_extractor_class(ie_key)}
1339 for ie_key
, ie
in ies
.items():
1340 if not ie
.suitable(url
):
1343 if not ie
.working():
1344 self
.report_warning('The program functionality for this site has been marked as broken, '
1345 'and will probably not work.')
1347 temp_id
= ie
.get_temp_id(url
)
1348 if temp_id
is not None and self
.in_download_archive({'id': temp_id, 'ie_key': ie_key}
):
1349 self
.to_screen(f
'[{ie_key}] {temp_id}: has already been recorded in the archive')
1350 if self
.params
.get('break_on_existing', False):
1351 raise ExistingVideoReached()
1353 return self
.__extract
_info
(url
, self
.get_info_extractor(ie_key
), download
, extra_info
, process
)
1355 self
.report_error('no suitable InfoExtractor for URL %s' % url
)
1357 def __handle_extraction_exceptions(func
):
1358 @functools.wraps(func
)
1359 def wrapper(self
, *args
, **kwargs
):
1362 return func(self
, *args
, **kwargs
)
1363 except (DownloadCancelled
, LazyList
.IndexError, PagedList
.IndexError):
1365 except ReExtractInfo
as e
:
1367 self
.to_screen(f
'{e}; Re-extracting data')
1369 self
.to_stderr('\r')
1370 self
.report_warning(f
'{e}; Re-extracting data')
1372 except GeoRestrictedError
as e
:
1375 msg
+= '\nThis video is available in %s.' % ', '.join(
1376 map(ISO3166Utils
.short2full
, e
.countries
))
1377 msg
+= '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1378 self
.report_error(msg
)
1379 except ExtractorError
as e
: # An error we somewhat expected
1380 self
.report_error(str(e
), e
.format_traceback())
1381 except Exception as e
:
1382 if self
.params
.get('ignoreerrors'):
1383 self
.report_error(str(e
), tb
=encode_compat_str(traceback
.format_exc()))
1389 def _wait_for_video(self
, ie_result
):
1390 if (not self
.params
.get('wait_for_video')
1391 or ie_result
.get('_type', 'video') != 'video'
1392 or ie_result
.get('formats') or ie_result
.get('url')):
1395 format_dur
= lambda dur
: '%02d:%02d:%02d' % timetuple_from_msec(dur
* 1000)[:-1]
1400 self
.to_screen(msg
+ ' ' * (len(last_msg
) - len(msg
)) + '\r', skip_eol
=True)
1403 min_wait
, max_wait
= self
.params
.get('wait_for_video')
1404 diff
= try_get(ie_result
, lambda x
: x
['release_timestamp'] - time
.time())
1405 if diff
is None and ie_result
.get('live_status') == 'is_upcoming':
1406 diff
= random
.randrange(min_wait
, max_wait
) if (max_wait
and min_wait
) else (max_wait
or min_wait
)
1407 self
.report_warning('Release time of video is not known')
1408 elif (diff
or 0) <= 0:
1409 self
.report_warning('Video should already be available according to extracted info')
1410 diff
= min(max(diff
or 0, min_wait
or 0), max_wait
or float('inf'))
1411 self
.to_screen(f
'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1413 wait_till
= time
.time() + diff
1416 diff
= wait_till
- time
.time()
1419 raise ReExtractInfo('[wait] Wait period ended', expected
=True)
1420 progress(f
'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1422 except KeyboardInterrupt:
1424 raise ReExtractInfo('[wait] Interrupted by user', expected
=True)
1425 except BaseException
as e
:
1426 if not isinstance(e
, ReExtractInfo
):
1430 @__handle_extraction_exceptions
1431 def __extract_info(self
, url
, ie
, download
, extra_info
, process
):
1432 ie_result
= ie
.extract(url
)
1433 if ie_result
is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1435 if isinstance(ie_result
, list):
1436 # Backwards compatibility: old IE result format
1438 '_type': 'compat_list',
1439 'entries': ie_result
,
1441 if extra_info
.get('original_url'):
1442 ie_result
.setdefault('original_url', extra_info
['original_url'])
1443 self
.add_default_extra_info(ie_result
, ie
, url
)
1445 self
._wait
_for
_video
(ie_result
)
1446 return self
.process_ie_result(ie_result
, download
, extra_info
)
1450 def add_default_extra_info(self
, ie_result
, ie
, url
):
1452 self
.add_extra_info(ie_result
, {
1454 'original_url': url
,
1455 'webpage_url_basename': url_basename(url
),
1456 'webpage_url_domain': get_domain(url
),
1459 self
.add_extra_info(ie_result
, {
1460 'extractor': ie
.IE_NAME
,
1461 'extractor_key': ie
.ie_key(),
1464 def process_ie_result(self
, ie_result
, download
=True, extra_info
=None):
1466 Take the result of the ie(may be modified) and resolve all unresolved
1467 references (URLs, playlist items).
1469 It will also download the videos if 'download'.
1470 Returns the resolved ie_result.
1472 if extra_info
is None:
1474 result_type
= ie_result
.get('_type', 'video')
1476 if result_type
in ('url', 'url_transparent'):
1477 ie_result
['url'] = sanitize_url(ie_result
['url'])
1478 if ie_result
.get('original_url'):
1479 extra_info
.setdefault('original_url', ie_result
['original_url'])
1481 extract_flat
= self
.params
.get('extract_flat', False)
1482 if ((extract_flat
== 'in_playlist' and 'playlist' in extra_info
)
1483 or extract_flat
is True):
1484 info_copy
= ie_result
.copy()
1485 ie
= try_get(ie_result
.get('ie_key'), self
.get_info_extractor
)
1486 if ie
and not ie_result
.get('id'):
1487 info_copy
['id'] = ie
.get_temp_id(ie_result
['url'])
1488 self
.add_default_extra_info(info_copy
, ie
, ie_result
['url'])
1489 self
.add_extra_info(info_copy
, extra_info
)
1490 info_copy
, _
= self
.pre_process(info_copy
)
1491 self
.__forced
_printings
(info_copy
, self
.prepare_filename(info_copy
), incomplete
=True)
1492 if self
.params
.get('force_write_download_archive', False):
1493 self
.record_download_archive(info_copy
)
1496 if result_type
== 'video':
1497 self
.add_extra_info(ie_result
, extra_info
)
1498 ie_result
= self
.process_video_result(ie_result
, download
=download
)
1499 additional_urls
= (ie_result
or {}).get('additional_urls')
1501 # TODO: Improve MetadataParserPP to allow setting a list
1502 if isinstance(additional_urls
, compat_str
):
1503 additional_urls
= [additional_urls
]
1505 '[info] %s: %d additional URL(s) requested' % (ie_result
['id'], len(additional_urls
)))
1506 self
.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls
))
1507 ie_result
['additional_entries'] = [
1509 url
, download
, extra_info
=extra_info
,
1510 force_generic_extractor
=self
.params
.get('force_generic_extractor'))
1511 for url
in additional_urls
1514 elif result_type
== 'url':
1515 # We have to add extra_info to the results because it may be
1516 # contained in a playlist
1517 return self
.extract_info(
1518 ie_result
['url'], download
,
1519 ie_key
=ie_result
.get('ie_key'),
1520 extra_info
=extra_info
)
1521 elif result_type
== 'url_transparent':
1522 # Use the information from the embedding page
1523 info
= self
.extract_info(
1524 ie_result
['url'], ie_key
=ie_result
.get('ie_key'),
1525 extra_info
=extra_info
, download
=False, process
=False)
1527 # extract_info may return None when ignoreerrors is enabled and
1528 # extraction failed with an error, don't crash and return early
1533 force_properties
= dict(
1534 (k
, v
) for k
, v
in ie_result
.items() if v
is not None)
1535 for f
in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1536 if f
in force_properties
:
1537 del force_properties
[f
]
1538 new_result
= info
.copy()
1539 new_result
.update(force_properties
)
1541 # Extracted info may not be a video result (i.e.
1542 # info.get('_type', 'video') != video) but rather an url or
1543 # url_transparent. In such cases outer metadata (from ie_result)
1544 # should be propagated to inner one (info). For this to happen
1545 # _type of info should be overridden with url_transparent. This
1546 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1547 if new_result
.get('_type') == 'url':
1548 new_result
['_type'] = 'url_transparent'
1550 return self
.process_ie_result(
1551 new_result
, download
=download
, extra_info
=extra_info
)
1552 elif result_type
in ('playlist', 'multi_video'):
1553 # Protect from infinite recursion due to recursively nested playlists
1554 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1555 webpage_url
= ie_result
['webpage_url']
1556 if webpage_url
in self
._playlist
_urls
:
1558 '[download] Skipping already downloaded playlist: %s'
1559 % ie_result
.get('title') or ie_result
.get('id'))
1562 self
._playlist
_level
+= 1
1563 self
._playlist
_urls
.add(webpage_url
)
1564 self
._sanitize
_thumbnails
(ie_result
)
1566 return self
.__process
_playlist
(ie_result
, download
)
1568 self
._playlist
_level
-= 1
1569 if not self
._playlist
_level
:
1570 self
._playlist
_urls
.clear()
1571 elif result_type
== 'compat_list':
1572 self
.report_warning(
1573 'Extractor %s returned a compat_list result. '
1574 'It needs to be updated.' % ie_result
.get('extractor'))
1577 self
.add_extra_info(r
, {
1578 'extractor': ie_result
['extractor'],
1579 'webpage_url': ie_result
['webpage_url'],
1580 'webpage_url_basename': url_basename(ie_result
['webpage_url']),
1581 'webpage_url_domain': get_domain(ie_result
['webpage_url']),
1582 'extractor_key': ie_result
['extractor_key'],
1585 ie_result
['entries'] = [
1586 self
.process_ie_result(_fixup(r
), download
, extra_info
)
1587 for r
in ie_result
['entries']
1591 raise Exception('Invalid result type: %s' % result_type
)
1593 def _ensure_dir_exists(self
, path
):
1594 return make_dir(path
, self
.report_error
)
1596 def __process_playlist(self
, ie_result
, download
):
1597 # We process each entry in the playlist
1598 playlist
= ie_result
.get('title') or ie_result
.get('id')
1599 self
.to_screen('[download] Downloading playlist: %s' % playlist
)
1601 if 'entries' not in ie_result
:
1602 raise EntryNotInPlaylist('There are no entries')
1604 MissingEntry
= object()
1605 incomplete_entries
= bool(ie_result
.get('requested_entries'))
1606 if incomplete_entries
:
1607 def fill_missing_entries(entries
, indices
):
1608 ret
= [MissingEntry
] * max(indices
)
1609 for i
, entry
in zip(indices
, entries
):
1612 ie_result
['entries'] = fill_missing_entries(ie_result
['entries'], ie_result
['requested_entries'])
1614 playlist_results
= []
1616 playliststart
= self
.params
.get('playliststart', 1)
1617 playlistend
= self
.params
.get('playlistend')
1618 # For backwards compatibility, interpret -1 as whole list
1619 if playlistend
== -1:
1622 playlistitems_str
= self
.params
.get('playlist_items')
1623 playlistitems
= None
1624 if playlistitems_str
is not None:
1625 def iter_playlistitems(format
):
1626 for string_segment
in format
.split(','):
1627 if '-' in string_segment
:
1628 start
, end
= string_segment
.split('-')
1629 for item
in range(int(start
), int(end
) + 1):
1632 yield int(string_segment
)
1633 playlistitems
= orderedSet(iter_playlistitems(playlistitems_str
))
1635 ie_entries
= ie_result
['entries']
1637 'Downloading %d videos' if not isinstance(ie_entries
, list)
1638 else 'Collected %d videos; downloading %%d of them' % len(ie_entries
))
1640 if isinstance(ie_entries
, list):
1642 return ie_entries
[i
- 1]
1644 if not isinstance(ie_entries
, (PagedList
, LazyList
)):
1645 ie_entries
= LazyList(ie_entries
)
1648 return YoutubeDL
.__handle
_extraction
_exceptions
(
1649 lambda self
, i
: ie_entries
[i
- 1]
1653 items
= playlistitems
if playlistitems
is not None else itertools
.count(playliststart
)
1657 if playlistitems
is None and playlistend
is not None and playlistend
< i
:
1661 entry
= get_entry(i
)
1662 if entry
is MissingEntry
:
1663 raise EntryNotInPlaylist()
1664 except (IndexError, EntryNotInPlaylist
):
1665 if incomplete_entries
:
1666 raise EntryNotInPlaylist(f
'Entry {i} cannot be found')
1667 elif not playlistitems
:
1669 entries
.append(entry
)
1671 if entry
is not None:
1672 self
._match
_entry
(entry
, incomplete
=True, silent
=True)
1673 except (ExistingVideoReached
, RejectedVideoReached
):
1675 ie_result
['entries'] = entries
1677 # Save playlist_index before re-ordering
1679 ((playlistitems
[i
- 1] if playlistitems
else i
+ playliststart
- 1), entry
)
1680 for i
, entry
in enumerate(entries
, 1)
1681 if entry
is not None]
1682 n_entries
= len(entries
)
1684 if not playlistitems
and (playliststart
!= 1 or playlistend
):
1685 playlistitems
= list(range(playliststart
, playliststart
+ n_entries
))
1686 ie_result
['requested_entries'] = playlistitems
1688 _infojson_written
= False
1689 if not self
.params
.get('simulate') and self
.params
.get('allow_playlist_files', True):
1691 'playlist': playlist
,
1692 'playlist_id': ie_result
.get('id'),
1693 'playlist_title': ie_result
.get('title'),
1694 'playlist_uploader': ie_result
.get('uploader'),
1695 'playlist_uploader_id': ie_result
.get('uploader_id'),
1696 'playlist_index': 0,
1697 'n_entries': n_entries
,
1699 ie_copy
.update(dict(ie_result
))
1701 _infojson_written
= self
._write
_info
_json
(
1702 'playlist', ie_result
, self
.prepare_filename(ie_copy
, 'pl_infojson'))
1703 if _infojson_written
is None:
1705 if self
._write
_description
('playlist', ie_result
,
1706 self
.prepare_filename(ie_copy
, 'pl_description')) is None:
1708 # TODO: This should be passed to ThumbnailsConvertor if necessary
1709 self
._write
_thumbnails
('playlist', ie_copy
, self
.prepare_filename(ie_copy
, 'pl_thumbnail'))
1711 if self
.params
.get('playlistreverse', False):
1712 entries
= entries
[::-1]
1713 if self
.params
.get('playlistrandom', False):
1714 random
.shuffle(entries
)
1716 x_forwarded_for
= ie_result
.get('__x_forwarded_for_ip')
1718 self
.to_screen('[%s] playlist %s: %s' % (ie_result
['extractor'], playlist
, msg
% n_entries
))
1720 max_failures
= self
.params
.get('skip_playlist_after_errors') or float('inf')
1721 for i
, entry_tuple
in enumerate(entries
, 1):
1722 playlist_index
, entry
= entry_tuple
1723 if 'playlist-index' in self
.params
.get('compat_opts', []):
1724 playlist_index
= playlistitems
[i
- 1] if playlistitems
else i
+ playliststart
- 1
1725 self
.to_screen('[download] Downloading video %s of %s' % (i
, n_entries
))
1726 # This __x_forwarded_for_ip thing is a bit ugly but requires
1729 entry
['__x_forwarded_for_ip'] = x_forwarded_for
1731 'n_entries': n_entries
,
1732 '_last_playlist_index': max(playlistitems
) if playlistitems
else (playlistend
or n_entries
),
1733 'playlist_index': playlist_index
,
1734 'playlist_autonumber': i
,
1735 'playlist': playlist
,
1736 'playlist_id': ie_result
.get('id'),
1737 'playlist_title': ie_result
.get('title'),
1738 'playlist_uploader': ie_result
.get('uploader'),
1739 'playlist_uploader_id': ie_result
.get('uploader_id'),
1740 'extractor': ie_result
['extractor'],
1741 'webpage_url': ie_result
['webpage_url'],
1742 'webpage_url_basename': url_basename(ie_result
['webpage_url']),
1743 'webpage_url_domain': get_domain(ie_result
['webpage_url']),
1744 'extractor_key': ie_result
['extractor_key'],
1747 if self
._match
_entry
(entry
, incomplete
=True) is not None:
1750 entry_result
= self
.__process
_iterable
_entry
(entry
, download
, extra
)
1751 if not entry_result
:
1753 if failures
>= max_failures
:
1755 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist
, failures
))
1757 playlist_results
.append(entry_result
)
1758 ie_result
['entries'] = playlist_results
1760 # Write the updated info to json
1761 if _infojson_written
and self
._write
_info
_json
(
1762 'updated playlist', ie_result
,
1763 self
.prepare_filename(ie_copy
, 'pl_infojson'), overwrite
=True) is None:
1766 for tmpl
in self
.params
['forceprint'].get('playlist', []):
1767 self
._forceprint
(tmpl
, ie_result
)
1768 self
.to_screen('[download] Finished downloading playlist: %s' % playlist
)
1771 @__handle_extraction_exceptions
1772 def __process_iterable_entry(self
, entry
, download
, extra_info
):
1773 return self
.process_ie_result(
1774 entry
, download
=download
, extra_info
=extra_info
)
1776 def _build_format_filter(self
, filter_spec
):
1777 " Returns a function to filter the formats according to the filter_spec "
1787 operator_rex
= re
.compile(r
'''(?x)\s*
1788 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1789 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1790 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1791 ''' % '|'.join(map(re
.escape
, OPERATORS
.keys())))
1792 m
= operator_rex
.fullmatch(filter_spec
)
1795 comparison_value
= int(m
.group('value'))
1797 comparison_value
= parse_filesize(m
.group('value'))
1798 if comparison_value
is None:
1799 comparison_value
= parse_filesize(m
.group('value') + 'B')
1800 if comparison_value
is None:
1802 'Invalid value %r in format specification %r' % (
1803 m
.group('value'), filter_spec
))
1804 op
= OPERATORS
[m
.group('op')]
1809 '^=': lambda attr
, value
: attr
.startswith(value
),
1810 '$=': lambda attr
, value
: attr
.endswith(value
),
1811 '*=': lambda attr
, value
: value
in attr
,
1813 str_operator_rex
= re
.compile(r
'''(?x)\s*
1814 (?P<key>[a-zA-Z0-9._-]+)\s*
1815 (?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1816 (?P<value>[a-zA-Z0-9._-]+)\s*
1817 ''' % '|'.join(map(re
.escape
, STR_OPERATORS
.keys())))
1818 m
= str_operator_rex
.fullmatch(filter_spec
)
1820 comparison_value
= m
.group('value')
1821 str_op
= STR_OPERATORS
[m
.group('op')]
1822 if m
.group('negation'):
1823 op
= lambda attr
, value
: not str_op(attr
, value
)
1828 raise SyntaxError('Invalid filter specification %r' % filter_spec
)
1831 actual_value
= f
.get(m
.group('key'))
1832 if actual_value
is None:
1833 return m
.group('none_inclusive')
1834 return op(actual_value
, comparison_value
)
1837 def _check_formats(self
, formats
):
1839 self
.to_screen('[info] Testing format %s' % f
['format_id'])
1840 path
= self
.get_output_path('temp')
1841 if not self
._ensure
_dir
_exists
(f
'{path}/'):
1843 temp_file
= tempfile
.NamedTemporaryFile(suffix
='.tmp', delete
=False, dir=path
or None)
1846 success
, _
= self
.dl(temp_file
.name
, f
, test
=True)
1847 except (DownloadError
, IOError, OSError, ValueError) + network_exceptions
:
1850 if os
.path
.exists(temp_file
.name
):
1852 os
.remove(temp_file
.name
)
1854 self
.report_warning('Unable to delete temporary file "%s"' % temp_file
.name
)
1858 self
.to_screen('[info] Unable to download format %s. Skipping...' % f
['format_id'])
1860 def _default_format_spec(self
, info_dict
, download
=True):
1863 merger
= FFmpegMergerPP(self
)
1864 return merger
.available
and merger
.can_merge()
1867 not self
.params
.get('simulate')
1871 or info_dict
.get('is_live', False)
1872 or self
.outtmpl_dict
['default'] == '-'))
1875 or self
.params
.get('allow_multiple_audio_streams', False)
1876 or 'format-spec' in self
.params
.get('compat_opts', []))
1879 'best/bestvideo+bestaudio' if prefer_best
1880 else 'bestvideo*+bestaudio/best' if not compat
1881 else 'bestvideo+bestaudio/best')
1883 def build_format_selector(self
, format_spec
):
1884 def syntax_error(note
, start
):
1886 'Invalid format specification: '
1887 '{0}\n\t{1}\n\t{2}^'.format(note
, format_spec
, ' ' * start
[1]))
1888 return SyntaxError(message
)
1890 PICKFIRST
= 'PICKFIRST'
1894 FormatSelector
= collections
.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1896 allow_multiple_streams
= {'audio': self
.params
.get('allow_multiple_audio_streams', False),
1897 'video': self
.params
.get('allow_multiple_video_streams', False)}
1899 check_formats
= self
.params
.get('check_formats') == 'selected'
1901 def _parse_filter(tokens
):
1903 for type, string
, start
, _
, _
in tokens
:
1904 if type == tokenize
.OP
and string
== ']':
1905 return ''.join(filter_parts
)
1907 filter_parts
.append(string
)
1909 def _remove_unused_ops(tokens
):
1910 # Remove operators that we don't use and join them with the surrounding strings
1911 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1912 ALLOWED_OPS
= ('/', '+', ',', '(', ')')
1913 last_string
, last_start
, last_end
, last_line
= None, None, None, None
1914 for type, string
, start
, end
, line
in tokens
:
1915 if type == tokenize
.OP
and string
== '[':
1917 yield tokenize
.NAME
, last_string
, last_start
, last_end
, last_line
1919 yield type, string
, start
, end
, line
1920 # everything inside brackets will be handled by _parse_filter
1921 for type, string
, start
, end
, line
in tokens
:
1922 yield type, string
, start
, end
, line
1923 if type == tokenize
.OP
and string
== ']':
1925 elif type == tokenize
.OP
and string
in ALLOWED_OPS
:
1927 yield tokenize
.NAME
, last_string
, last_start
, last_end
, last_line
1929 yield type, string
, start
, end
, line
1930 elif type in [tokenize
.NAME
, tokenize
.NUMBER
, tokenize
.OP
]:
1932 last_string
= string
1936 last_string
+= string
1938 yield tokenize
.NAME
, last_string
, last_start
, last_end
, last_line
1940 def _parse_format_selection(tokens
, inside_merge
=False, inside_choice
=False, inside_group
=False):
1942 current_selector
= None
1943 for type, string
, start
, _
, _
in tokens
:
1944 # ENCODING is only defined in python 3.x
1945 if type == getattr(tokenize
, 'ENCODING', None):
1947 elif type in [tokenize
.NAME
, tokenize
.NUMBER
]:
1948 current_selector
= FormatSelector(SINGLE
, string
, [])
1949 elif type == tokenize
.OP
:
1951 if not inside_group
:
1952 # ')' will be handled by the parentheses group
1953 tokens
.restore_last_token()
1955 elif inside_merge
and string
in ['/', ',']:
1956 tokens
.restore_last_token()
1958 elif inside_choice
and string
== ',':
1959 tokens
.restore_last_token()
1962 if not current_selector
:
1963 raise syntax_error('"," must follow a format selector', start
)
1964 selectors
.append(current_selector
)
1965 current_selector
= None
1967 if not current_selector
:
1968 raise syntax_error('"/" must follow a format selector', start
)
1969 first_choice
= current_selector
1970 second_choice
= _parse_format_selection(tokens
, inside_choice
=True)
1971 current_selector
= FormatSelector(PICKFIRST
, (first_choice
, second_choice
), [])
1973 if not current_selector
:
1974 current_selector
= FormatSelector(SINGLE
, 'best', [])
1975 format_filter
= _parse_filter(tokens
)
1976 current_selector
.filters
.append(format_filter
)
1978 if current_selector
:
1979 raise syntax_error('Unexpected "("', start
)
1980 group
= _parse_format_selection(tokens
, inside_group
=True)
1981 current_selector
= FormatSelector(GROUP
, group
, [])
1983 if not current_selector
:
1984 raise syntax_error('Unexpected "+"', start
)
1985 selector_1
= current_selector
1986 selector_2
= _parse_format_selection(tokens
, inside_merge
=True)
1988 raise syntax_error('Expected a selector', start
)
1989 current_selector
= FormatSelector(MERGE
, (selector_1
, selector_2
), [])
1991 raise syntax_error('Operator not recognized: "{0}"'.format(string
), start
)
1992 elif type == tokenize
.ENDMARKER
:
1994 if current_selector
:
1995 selectors
.append(current_selector
)
1998 def _merge(formats_pair
):
1999 format_1
, format_2
= formats_pair
2002 formats_info
.extend(format_1
.get('requested_formats', (format_1
,)))
2003 formats_info
.extend(format_2
.get('requested_formats', (format_2
,)))
2005 if not allow_multiple_streams
['video'] or not allow_multiple_streams
['audio']:
2006 get_no_more
= {'video': False, 'audio': False}
2007 for (i
, fmt_info
) in enumerate(formats_info
):
2008 if fmt_info
.get('acodec') == fmt_info
.get('vcodec') == 'none':
2011 for aud_vid
in ['audio', 'video']:
2012 if not allow_multiple_streams
[aud_vid
] and fmt_info
.get(aud_vid
[0] + 'codec') != 'none':
2013 if get_no_more
[aud_vid
]:
2016 get_no_more
[aud_vid
] = True
2018 if len(formats_info
) == 1:
2019 return formats_info
[0]
2021 video_fmts
= [fmt_info
for fmt_info
in formats_info
if fmt_info
.get('vcodec') != 'none']
2022 audio_fmts
= [fmt_info
for fmt_info
in formats_info
if fmt_info
.get('acodec') != 'none']
2024 the_only_video
= video_fmts
[0] if len(video_fmts
) == 1 else None
2025 the_only_audio
= audio_fmts
[0] if len(audio_fmts
) == 1 else None
2027 output_ext
= self
.params
.get('merge_output_format')
2030 output_ext
= the_only_video
['ext']
2031 elif the_only_audio
and not video_fmts
:
2032 output_ext
= the_only_audio
['ext']
2036 filtered
= lambda *keys
: filter(None, (traverse_obj(fmt
, *keys
) for fmt
in formats_info
))
2039 'requested_formats': formats_info
,
2040 'format': '+'.join(filtered('format')),
2041 'format_id': '+'.join(filtered('format_id')),
2043 'protocol': '+'.join(map(determine_protocol
, formats_info
)),
2044 'language': '+'.join(orderedSet(filtered('language'))) or None,
2045 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2046 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2047 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2052 'width': the_only_video
.get('width'),
2053 'height': the_only_video
.get('height'),
2054 'resolution': the_only_video
.get('resolution') or self
.format_resolution(the_only_video
),
2055 'fps': the_only_video
.get('fps'),
2056 'dynamic_range': the_only_video
.get('dynamic_range'),
2057 'vcodec': the_only_video
.get('vcodec'),
2058 'vbr': the_only_video
.get('vbr'),
2059 'stretched_ratio': the_only_video
.get('stretched_ratio'),
2064 'acodec': the_only_audio
.get('acodec'),
2065 'abr': the_only_audio
.get('abr'),
2066 'asr': the_only_audio
.get('asr'),
2071 def _check_formats(formats
):
2072 if not check_formats
:
2075 yield from self
._check
_formats
(formats
)
2077 def _build_selector_function(selector
):
2078 if isinstance(selector
, list): # ,
2079 fs
= [_build_selector_function(s
) for s
in selector
]
2081 def selector_function(ctx
):
2084 return selector_function
2086 elif selector
.type == GROUP
: # ()
2087 selector_function
= _build_selector_function(selector
.selector
)
2089 elif selector
.type == PICKFIRST
: # /
2090 fs
= [_build_selector_function(s
) for s
in selector
.selector
]
2092 def selector_function(ctx
):
2094 picked_formats
= list(f(ctx
))
2096 return picked_formats
2099 elif selector
.type == MERGE
: # +
2100 selector_1
, selector_2
= map(_build_selector_function
, selector
.selector
)
2102 def selector_function(ctx
):
2103 for pair
in itertools
.product(selector_1(ctx
), selector_2(ctx
)):
2106 elif selector
.type == SINGLE
: # atom
2107 format_spec
= selector
.selector
or 'best'
2109 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2110 if format_spec
== 'all':
2111 def selector_function(ctx
):
2112 yield from _check_formats(ctx
['formats'][::-1])
2113 elif format_spec
== 'mergeall':
2114 def selector_function(ctx
):
2115 formats
= list(_check_formats(ctx
['formats']))
2118 merged_format
= formats
[-1]
2119 for f
in formats
[-2::-1]:
2120 merged_format
= _merge((merged_format
, f
))
2124 format_fallback
, format_reverse
, format_idx
= False, True, 1
2126 r
'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2128 if mobj
is not None:
2129 format_idx
= int_or_none(mobj
.group('n'), default
=1)
2130 format_reverse
= mobj
.group('bw')[0] == 'b'
2131 format_type
= (mobj
.group('type') or [None])[0]
2132 not_format_type
= {'v': 'a', 'a': 'v'}
.get(format_type
)
2133 format_modified
= mobj
.group('mod') is not None
2135 format_fallback
= not format_type
and not format_modified
# for b, w
2137 (lambda f
: f
.get('%scodec' % format_type
) != 'none')
2138 if format_type
and format_modified
# bv*, ba*, wv*, wa*
2139 else (lambda f
: f
.get('%scodec' % not_format_type
) == 'none')
2140 if format_type
# bv, ba, wv, wa
2141 else (lambda f
: f
.get('vcodec') != 'none' and f
.get('acodec') != 'none')
2142 if not format_modified
# b, w
2143 else lambda f
: True) # b*, w*
2144 filter_f
= lambda f
: _filter_f(f
) and (
2145 f
.get('vcodec') != 'none' or f
.get('acodec') != 'none')
2147 if format_spec
in self
._format
_selection
_exts
['audio']:
2148 filter_f
= lambda f
: f
.get('ext') == format_spec
and f
.get('acodec') != 'none'
2149 elif format_spec
in self
._format
_selection
_exts
['video']:
2150 filter_f
= lambda f
: f
.get('ext') == format_spec
and f
.get('acodec') != 'none' and f
.get('vcodec') != 'none'
2151 elif format_spec
in self
._format
_selection
_exts
['storyboards']:
2152 filter_f
= lambda f
: f
.get('ext') == format_spec
and f
.get('acodec') == 'none' and f
.get('vcodec') == 'none'
2154 filter_f
= lambda f
: f
.get('format_id') == format_spec
# id
2156 def selector_function(ctx
):
2157 formats
= list(ctx
['formats'])
2158 matches
= list(filter(filter_f
, formats
)) if filter_f
is not None else formats
2159 if format_fallback
and ctx
['incomplete_formats'] and not matches
:
2160 # for extractors with incomplete formats (audio only (soundcloud)
2161 # or video only (imgur)) best/worst will fallback to
2162 # best/worst {video,audio}-only format
2164 matches
= LazyList(_check_formats(matches
[::-1 if format_reverse
else 1]))
2166 yield matches
[format_idx
- 1]
2170 filters
= [self
._build
_format
_filter
(f
) for f
in selector
.filters
]
2172 def final_selector(ctx
):
2173 ctx_copy
= dict(ctx
)
2174 for _filter
in filters
:
2175 ctx_copy
['formats'] = list(filter(_filter
, ctx_copy
['formats']))
2176 return selector_function(ctx_copy
)
2177 return final_selector
2179 stream
= io
.BytesIO(format_spec
.encode('utf-8'))
2181 tokens
= list(_remove_unused_ops(compat_tokenize_tokenize(stream
.readline
)))
2182 except tokenize
.TokenError
:
2183 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec
)))
2185 class TokenIterator(object):
2186 def __init__(self
, tokens
):
2187 self
.tokens
= tokens
2194 if self
.counter
>= len(self
.tokens
):
2195 raise StopIteration()
2196 value
= self
.tokens
[self
.counter
]
2202 def restore_last_token(self
):
2205 parsed_selector
= _parse_format_selection(iter(TokenIterator(tokens
)))
2206 return _build_selector_function(parsed_selector
)
2208 def _calc_headers(self
, info_dict
):
2209 res
= std_headers
.copy()
2211 add_headers
= info_dict
.get('http_headers')
2213 res
.update(add_headers
)
2215 cookies
= self
._calc
_cookies
(info_dict
)
2217 res
['Cookie'] = cookies
2219 if 'X-Forwarded-For' not in res
:
2220 x_forwarded_for_ip
= info_dict
.get('__x_forwarded_for_ip')
2221 if x_forwarded_for_ip
:
2222 res
['X-Forwarded-For'] = x_forwarded_for_ip
2226 def _calc_cookies(self
, info_dict
):
2227 pr
= sanitized_Request(info_dict
['url'])
2228 self
.cookiejar
.add_cookie_header(pr
)
2229 return pr
.get_header('Cookie')
2231 def _sort_thumbnails(self
, thumbnails
):
2232 thumbnails
.sort(key
=lambda t
: (
2233 t
.get('preference') if t
.get('preference') is not None else -1,
2234 t
.get('width') if t
.get('width') is not None else -1,
2235 t
.get('height') if t
.get('height') is not None else -1,
2236 t
.get('id') if t
.get('id') is not None else '',
2239 def _sanitize_thumbnails(self
, info_dict
):
2240 thumbnails
= info_dict
.get('thumbnails')
2241 if thumbnails
is None:
2242 thumbnail
= info_dict
.get('thumbnail')
2244 info_dict
['thumbnails'] = thumbnails
= [{'url': thumbnail}
]
2248 def check_thumbnails(thumbnails
):
2249 for t
in thumbnails
:
2250 self
.to_screen(f
'[info] Testing thumbnail {t["id"]}')
2252 self
.urlopen(HEADRequest(t
['url']))
2253 except network_exceptions
as err
:
2254 self
.to_screen(f
'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2258 self
._sort
_thumbnails
(thumbnails
)
2259 for i
, t
in enumerate(thumbnails
):
2260 if t
.get('id') is None:
2262 if t
.get('width') and t
.get('height'):
2263 t
['resolution'] = '%dx%d' % (t
['width'], t
['height'])
2264 t
['url'] = sanitize_url(t
['url'])
2266 if self
.params
.get('check_formats') is True:
2267 info_dict
['thumbnails'] = LazyList(check_thumbnails(thumbnails
[::-1]), reverse
=True)
2269 info_dict
['thumbnails'] = thumbnails
2271 def process_video_result(self
, info_dict
, download
=True):
2272 assert info_dict
.get('_type', 'video') == 'video'
2274 if 'id' not in info_dict
:
2275 raise ExtractorError('Missing "id" field in extractor result')
2276 if 'title' not in info_dict
:
2277 raise ExtractorError('Missing "title" field in extractor result',
2278 video_id
=info_dict
['id'], ie
=info_dict
['extractor'])
2280 def report_force_conversion(field
, field_not
, conversion
):
2281 self
.report_warning(
2282 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2283 % (field
, field_not
, conversion
))
2285 def sanitize_string_field(info
, string_field
):
2286 field
= info
.get(string_field
)
2287 if field
is None or isinstance(field
, compat_str
):
2289 report_force_conversion(string_field
, 'a string', 'string')
2290 info
[string_field
] = compat_str(field
)
2292 def sanitize_numeric_fields(info
):
2293 for numeric_field
in self
._NUMERIC
_FIELDS
:
2294 field
= info
.get(numeric_field
)
2295 if field
is None or isinstance(field
, compat_numeric_types
):
2297 report_force_conversion(numeric_field
, 'numeric', 'int')
2298 info
[numeric_field
] = int_or_none(field
)
2300 sanitize_string_field(info_dict
, 'id')
2301 sanitize_numeric_fields(info_dict
)
2303 if 'playlist' not in info_dict
:
2304 # It isn't part of a playlist
2305 info_dict
['playlist'] = None
2306 info_dict
['playlist_index'] = None
2308 self
._sanitize
_thumbnails
(info_dict
)
2310 thumbnail
= info_dict
.get('thumbnail')
2311 thumbnails
= info_dict
.get('thumbnails')
2313 info_dict
['thumbnail'] = sanitize_url(thumbnail
)
2315 info_dict
['thumbnail'] = thumbnails
[-1]['url']
2317 if info_dict
.get('display_id') is None and 'id' in info_dict
:
2318 info_dict
['display_id'] = info_dict
['id']
2320 if info_dict
.get('duration') is not None:
2321 info_dict
['duration_string'] = formatSeconds(info_dict
['duration'])
2323 for ts_key
, date_key
in (
2324 ('timestamp', 'upload_date'),
2325 ('release_timestamp', 'release_date'),
2327 if info_dict
.get(date_key
) is None and info_dict
.get(ts_key
) is not None:
2328 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2329 # see http://bugs.python.org/issue1646728)
2331 upload_date
= datetime
.datetime
.utcfromtimestamp(info_dict
[ts_key
])
2332 info_dict
[date_key
] = upload_date
.strftime('%Y%m%d')
2333 except (ValueError, OverflowError, OSError):
2336 live_keys
= ('is_live', 'was_live')
2337 live_status
= info_dict
.get('live_status')
2338 if live_status
is None:
2339 for key
in live_keys
:
2340 if info_dict
.get(key
) is False:
2342 if info_dict
.get(key
):
2345 if all(info_dict
.get(key
) is False for key
in live_keys
):
2346 live_status
= 'not_live'
2348 info_dict
['live_status'] = live_status
2349 for key
in live_keys
:
2350 if info_dict
.get(key
) is None:
2351 info_dict
[key
] = (live_status
== key
)
2353 # Auto generate title fields corresponding to the *_number fields when missing
2354 # in order to always have clean titles. This is very common for TV series.
2355 for field
in ('chapter', 'season', 'episode'):
2356 if info_dict
.get('%s_number' % field
) is not None and not info_dict
.get(field
):
2357 info_dict
[field
] = '%s %d' % (field
.capitalize(), info_dict
['%s_number' % field
])
2359 for cc_kind
in ('subtitles', 'automatic_captions'):
2360 cc
= info_dict
.get(cc_kind
)
2362 for _
, subtitle
in cc
.items():
2363 for subtitle_format
in subtitle
:
2364 if subtitle_format
.get('url'):
2365 subtitle_format
['url'] = sanitize_url(subtitle_format
['url'])
2366 if subtitle_format
.get('ext') is None:
2367 subtitle_format
['ext'] = determine_ext(subtitle_format
['url']).lower()
2369 automatic_captions
= info_dict
.get('automatic_captions')
2370 subtitles
= info_dict
.get('subtitles')
2372 info_dict
['requested_subtitles'] = self
.process_subtitles(
2373 info_dict
['id'], subtitles
, automatic_captions
)
2375 if info_dict
.get('formats') is None:
2376 # There's only one format available
2377 formats
= [info_dict
]
2379 formats
= info_dict
['formats']
2381 info_dict
['__has_drm'] = any(f
.get('has_drm') for f
in formats
)
2382 if not self
.params
.get('allow_unplayable_formats'):
2383 formats
= [f
for f
in formats
if not f
.get('has_drm')]
2385 if info_dict
.get('is_live'):
2386 get_from_start
= bool(self
.params
.get('live_from_start'))
2387 formats
= [f
for f
in formats
if bool(f
.get('is_from_start')) == get_from_start
]
2388 if not get_from_start
:
2389 info_dict
['title'] += ' ' + datetime
.datetime
.now().strftime('%Y-%m-%d %H:%M')
2392 self
.raise_no_formats(info_dict
)
2394 def is_wellformed(f
):
2397 self
.report_warning(
2398 '"url" field is missing or empty - skipping format, '
2399 'there is an error in extractor')
2401 if isinstance(url
, bytes):
2402 sanitize_string_field(f
, 'url')
2405 # Filter out malformed formats for better extraction robustness
2406 formats
= list(filter(is_wellformed
, formats
))
2410 # We check that all the formats have the format and format_id fields
2411 for i
, format
in enumerate(formats
):
2412 sanitize_string_field(format
, 'format_id')
2413 sanitize_numeric_fields(format
)
2414 format
['url'] = sanitize_url(format
['url'])
2415 if not format
.get('format_id'):
2416 format
['format_id'] = compat_str(i
)
2418 # Sanitize format_id from characters used in format selector expression
2419 format
['format_id'] = re
.sub(r
'[\s,/+\[\]()]', '_', format
['format_id'])
2420 format_id
= format
['format_id']
2421 if format_id
not in formats_dict
:
2422 formats_dict
[format_id
] = []
2423 formats_dict
[format_id
].append(format
)
2425 # Make sure all formats have unique format_id
2426 common_exts
= set(itertools
.chain(*self
._format
_selection
_exts
.values()))
2427 for format_id
, ambiguous_formats
in formats_dict
.items():
2428 ambigious_id
= len(ambiguous_formats
) > 1
2429 for i
, format
in enumerate(ambiguous_formats
):
2431 format
['format_id'] = '%s-%d' % (format_id
, i
)
2432 if format
.get('ext') is None:
2433 format
['ext'] = determine_ext(format
['url']).lower()
2434 # Ensure there is no conflict between id and ext in format selection
2435 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2436 if format
['format_id'] != format
['ext'] and format
['format_id'] in common_exts
:
2437 format
['format_id'] = 'f%s' % format
['format_id']
2439 for i
, format
in enumerate(formats
):
2440 if format
.get('format') is None:
2441 format
['format'] = '{id} - {res}{note}'.format(
2442 id=format
['format_id'],
2443 res
=self
.format_resolution(format
),
2444 note
=format_field(format
, 'format_note', ' (%s)'),
2446 if format
.get('protocol') is None:
2447 format
['protocol'] = determine_protocol(format
)
2448 if format
.get('resolution') is None:
2449 format
['resolution'] = self
.format_resolution(format
, default
=None)
2450 if format
.get('dynamic_range') is None and format
.get('vcodec') != 'none':
2451 format
['dynamic_range'] = 'SDR'
2452 if (info_dict
.get('duration') and format
.get('tbr')
2453 and not format
.get('filesize') and not format
.get('filesize_approx')):
2454 format
['filesize_approx'] = info_dict
['duration'] * format
['tbr'] * (1024 / 8)
2456 # Add HTTP headers, so that external programs can use them from the
2458 full_format_info
= info_dict
.copy()
2459 full_format_info
.update(format
)
2460 format
['http_headers'] = self
._calc
_headers
(full_format_info
)
2461 # Remove private housekeeping stuff
2462 if '__x_forwarded_for_ip' in info_dict
:
2463 del info_dict
['__x_forwarded_for_ip']
2465 # TODO Central sorting goes here
2467 if self
.params
.get('check_formats') is True:
2468 formats
= LazyList(self
._check
_formats
(formats
[::-1]), reverse
=True)
2470 if not formats
or formats
[0] is not info_dict
:
2471 # only set the 'formats' fields if the original info_dict list them
2472 # otherwise we end up with a circular reference, the first (and unique)
2473 # element in the 'formats' field in info_dict is info_dict itself,
2474 # which can't be exported to json
2475 info_dict
['formats'] = formats
2477 info_dict
, _
= self
.pre_process(info_dict
)
2479 # The pre-processors may have modified the formats
2480 formats
= info_dict
.get('formats', [info_dict
])
2482 list_only
= self
.params
.get('simulate') is None and (
2483 self
.params
.get('list_thumbnails') or self
.params
.get('listformats') or self
.params
.get('listsubtitles'))
2484 interactive_format_selection
= not list_only
and self
.format_selector
== '-'
2485 if self
.params
.get('list_thumbnails'):
2486 self
.list_thumbnails(info_dict
)
2487 if self
.params
.get('listsubtitles'):
2488 if 'automatic_captions' in info_dict
:
2489 self
.list_subtitles(
2490 info_dict
['id'], automatic_captions
, 'automatic captions')
2491 self
.list_subtitles(info_dict
['id'], subtitles
, 'subtitles')
2492 if self
.params
.get('listformats') or interactive_format_selection
:
2493 self
.list_formats(info_dict
)
2495 # Without this printing, -F --print-json will not work
2496 self
.__forced
_printings
(info_dict
, self
.prepare_filename(info_dict
), incomplete
=True)
2499 format_selector
= self
.format_selector
2500 if format_selector
is None:
2501 req_format
= self
._default
_format
_spec
(info_dict
, download
=download
)
2502 self
.write_debug('Default format spec: %s' % req_format
)
2503 format_selector
= self
.build_format_selector(req_format
)
2506 if interactive_format_selection
:
2508 self
._format
_screen
('\nEnter format selector: ', self
.Styles
.EMPHASIS
))
2510 format_selector
= self
.build_format_selector(req_format
)
2511 except SyntaxError as err
:
2512 self
.report_error(err
, tb
=False, is_error
=False)
2515 # While in format selection we may need to have an access to the original
2516 # format set in order to calculate some metrics or do some processing.
2517 # For now we need to be able to guess whether original formats provided
2518 # by extractor are incomplete or not (i.e. whether extractor provides only
2519 # video-only or audio-only formats) for proper formats selection for
2520 # extractors with such incomplete formats (see
2521 # https://github.com/ytdl-org/youtube-dl/pull/5556).
2522 # Since formats may be filtered during format selection and may not match
2523 # the original formats the results may be incorrect. Thus original formats
2524 # or pre-calculated metrics should be passed to format selection routines
2526 # We will pass a context object containing all necessary additional data
2527 # instead of just formats.
2528 # This fixes incorrect format selection issue (see
2529 # https://github.com/ytdl-org/youtube-dl/issues/10083).
2530 incomplete_formats
= (
2531 # All formats are video-only or
2532 all(f
.get('vcodec') != 'none' and f
.get('acodec') == 'none' for f
in formats
)
2533 # all formats are audio-only
2534 or all(f
.get('vcodec') == 'none' and f
.get('acodec') != 'none' for f
in formats
))
2538 'incomplete_formats': incomplete_formats
,
2541 formats_to_download
= list(format_selector(ctx
))
2542 if interactive_format_selection
and not formats_to_download
:
2543 self
.report_error('Requested format is not available', tb
=False, is_error
=False)
2547 if not formats_to_download
:
2548 if not self
.params
.get('ignore_no_formats_error'):
2549 raise ExtractorError('Requested format is not available', expected
=True,
2550 video_id
=info_dict
['id'], ie
=info_dict
['extractor'])
2552 self
.report_warning('Requested format is not available')
2553 # Process what we can, even without any available formats.
2554 self
.process_info(dict(info_dict
))
2557 '[info] %s: Downloading %d format(s): %s' % (
2558 info_dict
['id'], len(formats_to_download
),
2559 ", ".join([f
['format_id'] for f
in formats_to_download
])))
2560 for fmt
in formats_to_download
:
2561 new_info
= dict(info_dict
)
2562 # Save a reference to the original info_dict so that it can be modified in process_info if needed
2563 new_info
['__original_infodict'] = info_dict
2564 new_info
.update(fmt
)
2565 self
.process_info(new_info
)
2566 # We update the info dict with the selected best quality format (backwards compatibility)
2567 if formats_to_download
:
2568 info_dict
.update(formats_to_download
[-1])
2571 def process_subtitles(self
, video_id
, normal_subtitles
, automatic_captions
):
2572 """Select the requested subtitles and their format"""
2574 if normal_subtitles
and self
.params
.get('writesubtitles'):
2575 available_subs
.update(normal_subtitles
)
2576 if automatic_captions
and self
.params
.get('writeautomaticsub'):
2577 for lang
, cap_info
in automatic_captions
.items():
2578 if lang
not in available_subs
:
2579 available_subs
[lang
] = cap_info
2581 if (not self
.params
.get('writesubtitles') and not
2582 self
.params
.get('writeautomaticsub') or not
2586 all_sub_langs
= available_subs
.keys()
2587 if self
.params
.get('allsubtitles', False):
2588 requested_langs
= all_sub_langs
2589 elif self
.params
.get('subtitleslangs', False):
2590 # A list is used so that the order of languages will be the same as
2591 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2592 requested_langs
= []
2593 for lang_re
in self
.params
.get('subtitleslangs'):
2594 if lang_re
== 'all':
2595 requested_langs
.extend(all_sub_langs
)
2597 discard
= lang_re
[0] == '-'
2599 lang_re
= lang_re
[1:]
2600 current_langs
= filter(re
.compile(lang_re
+ '$').match
, all_sub_langs
)
2602 for lang
in current_langs
:
2603 while lang
in requested_langs
:
2604 requested_langs
.remove(lang
)
2606 requested_langs
.extend(current_langs
)
2607 requested_langs
= orderedSet(requested_langs
)
2608 elif 'en' in available_subs
:
2609 requested_langs
= ['en']
2611 requested_langs
= [list(all_sub_langs
)[0]]
2613 self
.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs
))
2615 formats_query
= self
.params
.get('subtitlesformat', 'best')
2616 formats_preference
= formats_query
.split('/') if formats_query
else []
2618 for lang
in requested_langs
:
2619 formats
= available_subs
.get(lang
)
2621 self
.report_warning('%s subtitles not available for %s' % (lang
, video_id
))
2623 for ext
in formats_preference
:
2627 matches
= list(filter(lambda f
: f
['ext'] == ext
, formats
))
2633 self
.report_warning(
2634 'No subtitle format found matching "%s" for language %s, '
2635 'using %s' % (formats_query
, lang
, f
['ext']))
2639 def _forceprint(self
, tmpl
, info_dict
):
2640 mobj
= re
.match(r
'\w+(=?)$', tmpl
)
2641 if mobj
and mobj
.group(1):
2642 tmpl
= f
'{tmpl[:-1]} = %({tmpl[:-1]})s'
2644 tmpl
= '%({})s'.format(tmpl
)
2645 self
.to_stdout(self
.evaluate_outtmpl(tmpl
, info_dict
))
2647 def __forced_printings(self
, info_dict
, filename
, incomplete
):
2648 def print_mandatory(field
, actual_field
=None):
2649 if actual_field
is None:
2650 actual_field
= field
2651 if (self
.params
.get('force%s' % field
, False)
2652 and (not incomplete
or info_dict
.get(actual_field
) is not None)):
2653 self
.to_stdout(info_dict
[actual_field
])
2655 def print_optional(field
):
2656 if (self
.params
.get('force%s' % field
, False)
2657 and info_dict
.get(field
) is not None):
2658 self
.to_stdout(info_dict
[field
])
2660 info_dict
= info_dict
.copy()
2661 if filename
is not None:
2662 info_dict
['filename'] = filename
2663 if info_dict
.get('requested_formats') is not None:
2664 # For RTMP URLs, also include the playpath
2665 info_dict
['urls'] = '\n'.join(f
['url'] + f
.get('play_path', '') for f
in info_dict
['requested_formats'])
2666 elif 'url' in info_dict
:
2667 info_dict
['urls'] = info_dict
['url'] + info_dict
.get('play_path', '')
2669 if self
.params
['forceprint'].get('video') or self
.params
.get('forcejson'):
2670 self
.post_extract(info_dict
)
2671 for tmpl
in self
.params
['forceprint'].get('video', []):
2672 self
._forceprint
(tmpl
, info_dict
)
2674 print_mandatory('title')
2675 print_mandatory('id')
2676 print_mandatory('url', 'urls')
2677 print_optional('thumbnail')
2678 print_optional('description')
2679 print_optional('filename')
2680 if self
.params
.get('forceduration') and info_dict
.get('duration') is not None:
2681 self
.to_stdout(formatSeconds(info_dict
['duration']))
2682 print_mandatory('format')
2684 if self
.params
.get('forcejson'):
2685 self
.to_stdout(json
.dumps(self
.sanitize_info(info_dict
)))
2687 def dl(self
, name
, info
, subtitle
=False, test
=False):
2688 if not info
.get('url'):
2689 self
.raise_no_formats(info
, True)
2692 verbose
= self
.params
.get('verbose')
2695 'quiet': self
.params
.get('quiet') or not verbose
,
2697 'noprogress': not verbose
,
2699 'skip_unavailable_fragments': False,
2700 'keep_fragments': False,
2702 '_no_ytdl_file': True,
2705 params
= self
.params
2706 fd
= get_suitable_downloader(info
, params
, to_stdout
=(name
== '-'))(self
, params
)
2708 for ph
in self
._progress
_hooks
:
2709 fd
.add_progress_hook(ph
)
2710 urls
= '", "'.join([f
['url'] for f
in info
.get('requested_formats', [])] or [info
['url']])
2711 self
.write_debug('Invoking downloader on "%s"' % urls
)
2713 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2714 # But it may contain objects that are not deep-copyable
2715 new_info
= self
._copy
_infodict
(info
)
2716 if new_info
.get('http_headers') is None:
2717 new_info
['http_headers'] = self
._calc
_headers
(new_info
)
2718 return fd
.download(name
, new_info
, subtitle
)
2720 def process_info(self
, info_dict
):
2721 """Process a single resolved IE result."""
2723 assert info_dict
.get('_type', 'video') == 'video'
2725 max_downloads
= self
.params
.get('max_downloads')
2726 if max_downloads
is not None:
2727 if self
._num
_downloads
>= int(max_downloads
):
2728 raise MaxDownloadsReached()
2730 # TODO: backward compatibility, to be removed
2731 info_dict
['fulltitle'] = info_dict
['title']
2733 if 'format' not in info_dict
and 'ext' in info_dict
:
2734 info_dict
['format'] = info_dict
['ext']
2736 if self
._match
_entry
(info_dict
) is not None:
2739 self
.post_extract(info_dict
)
2740 self
._num
_downloads
+= 1
2742 # info_dict['_filename'] needs to be set for backward compatibility
2743 info_dict
['_filename'] = full_filename
= self
.prepare_filename(info_dict
, warn
=True)
2744 temp_filename
= self
.prepare_filename(info_dict
, 'temp')
2748 self
.__forced
_printings
(info_dict
, full_filename
, incomplete
=('format' not in info_dict
))
2750 if self
.params
.get('simulate'):
2751 if self
.params
.get('force_write_download_archive', False):
2752 self
.record_download_archive(info_dict
)
2753 # Do nothing else if in simulate mode
2756 if full_filename
is None:
2758 if not self
._ensure
_dir
_exists
(encodeFilename(full_filename
)):
2760 if not self
._ensure
_dir
_exists
(encodeFilename(temp_filename
)):
2763 if self
._write
_description
('video', info_dict
,
2764 self
.prepare_filename(info_dict
, 'description')) is None:
2767 sub_files
= self
._write
_subtitles
(info_dict
, temp_filename
)
2768 if sub_files
is None:
2770 files_to_move
.update(dict(sub_files
))
2772 thumb_files
= self
._write
_thumbnails
(
2773 'video', info_dict
, temp_filename
, self
.prepare_filename(info_dict
, 'thumbnail'))
2774 if thumb_files
is None:
2776 files_to_move
.update(dict(thumb_files
))
2778 infofn
= self
.prepare_filename(info_dict
, 'infojson')
2779 _infojson_written
= self
._write
_info
_json
('video', info_dict
, infofn
)
2780 if _infojson_written
:
2781 info_dict
['infojson_filename'] = infofn
2782 # For backward compatibility, even though it was a private field
2783 info_dict
['__infojson_filename'] = infofn
2784 elif _infojson_written
is None:
2787 # Note: Annotations are deprecated
2789 if self
.params
.get('writeannotations', False):
2790 annofn
= self
.prepare_filename(info_dict
, 'annotation')
2792 if not self
._ensure
_dir
_exists
(encodeFilename(annofn
)):
2794 if not self
.params
.get('overwrites', True) and os
.path
.exists(encodeFilename(annofn
)):
2795 self
.to_screen('[info] Video annotations are already present')
2796 elif not info_dict
.get('annotations'):
2797 self
.report_warning('There are no annotations to write.')
2800 self
.to_screen('[info] Writing video annotations to: ' + annofn
)
2801 with io
.open(encodeFilename(annofn
), 'w', encoding
='utf-8') as annofile
:
2802 annofile
.write(info_dict
['annotations'])
2803 except (KeyError, TypeError):
2804 self
.report_warning('There are no annotations to write.')
2805 except (OSError, IOError):
2806 self
.report_error('Cannot write annotations file: ' + annofn
)
2809 # Write internet shortcut files
2810 def _write_link_file(link_type
):
2811 if 'webpage_url' not in info_dict
:
2812 self
.report_error('Cannot write internet shortcut file because the "webpage_url" field is missing in the media information')
2814 linkfn
= replace_extension(self
.prepare_filename(info_dict
, 'link'), link_type
, info_dict
.get('ext'))
2815 if not self
._ensure
_dir
_exists
(encodeFilename(linkfn
)):
2817 if self
.params
.get('overwrites', True) and os
.path
.exists(encodeFilename(linkfn
)):
2818 self
.to_screen(f
'[info] Internet shortcut (.{link_type}) is already present')
2821 self
.to_screen(f
'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2822 with io
.open(encodeFilename(to_high_limit_path(linkfn
)), 'w', encoding
='utf-8',
2823 newline
='\r\n' if link_type
== 'url' else '\n') as linkfile
:
2824 template_vars
= {'url': iri_to_uri(info_dict['webpage_url'])}
2825 if link_type
== 'desktop':
2826 template_vars
['filename'] = linkfn
[:-(len(link_type
) + 1)]
2827 linkfile
.write(LINK_TEMPLATES
[link_type
] % template_vars
)
2828 except (OSError, IOError):
2829 self
.report_error(f
'Cannot write internet shortcut {linkfn}')
2834 'url': self
.params
.get('writeurllink'),
2835 'webloc': self
.params
.get('writewebloclink'),
2836 'desktop': self
.params
.get('writedesktoplink'),
2838 if self
.params
.get('writelink'):
2839 link_type
= ('webloc' if sys
.platform
== 'darwin'
2840 else 'desktop' if sys
.platform
.startswith('linux')
2842 write_links
[link_type
] = True
2844 if any(should_write
and not _write_link_file(link_type
)
2845 for link_type
, should_write
in write_links
.items()):
2849 info_dict
, files_to_move
= self
.pre_process(info_dict
, 'before_dl', files_to_move
)
2850 except PostProcessingError
as err
:
2851 self
.report_error('Preprocessing: %s' % str(err
))
2854 must_record_download_archive
= False
2855 if self
.params
.get('skip_download', False):
2856 info_dict
['filepath'] = temp_filename
2857 info_dict
['__finaldir'] = os
.path
.dirname(os
.path
.abspath(encodeFilename(full_filename
)))
2858 info_dict
['__files_to_move'] = files_to_move
2859 info_dict
= self
.run_pp(MoveFilesAfterDownloadPP(self
, False), info_dict
)
2862 info_dict
.setdefault('__postprocessors', [])
2865 def existing_file(*filepaths
):
2866 ext
= info_dict
.get('ext')
2867 final_ext
= self
.params
.get('final_ext', ext
)
2869 for file in orderedSet(filepaths
):
2870 if final_ext
!= ext
:
2871 converted
= replace_extension(file, final_ext
, ext
)
2872 if os
.path
.exists(encodeFilename(converted
)):
2873 existing_files
.append(converted
)
2874 if os
.path
.exists(encodeFilename(file)):
2875 existing_files
.append(file)
2877 if not existing_files
or self
.params
.get('overwrites', False):
2878 for file in orderedSet(existing_files
):
2879 self
.report_file_delete(file)
2880 os
.remove(encodeFilename(file))
2883 info_dict
['ext'] = os
.path
.splitext(existing_files
[0])[1][1:]
2884 return existing_files
[0]
2887 if info_dict
.get('requested_formats') is not None:
2889 def compatible_formats(formats
):
2890 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2891 video_formats
= [format
for format
in formats
if format
.get('vcodec') != 'none']
2892 audio_formats
= [format
for format
in formats
if format
.get('acodec') != 'none']
2893 if len(video_formats
) > 2 or len(audio_formats
) > 2:
2897 exts
= set(format
.get('ext') for format
in formats
)
2899 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
2902 for ext_sets
in COMPATIBLE_EXTS
:
2903 if ext_sets
.issuperset(exts
):
2905 # TODO: Check acodec/vcodec
2908 requested_formats
= info_dict
['requested_formats']
2909 old_ext
= info_dict
['ext']
2910 if self
.params
.get('merge_output_format') is None:
2911 if not compatible_formats(requested_formats
):
2912 info_dict
['ext'] = 'mkv'
2913 self
.report_warning(
2914 'Requested formats are incompatible for merge and will be merged into mkv')
2915 if (info_dict
['ext'] == 'webm'
2916 and info_dict
.get('thumbnails')
2917 # check with type instead of pp_key, __name__, or isinstance
2918 # since we dont want any custom PPs to trigger this
2919 and any(type(pp
) == EmbedThumbnailPP
for pp
in self
._pps
['post_process'])):
2920 info_dict
['ext'] = 'mkv'
2921 self
.report_warning(
2922 'webm doesn\'t support embedding a thumbnail, mkv will be used')
2923 new_ext
= info_dict
['ext']
2925 def correct_ext(filename
, ext
=new_ext
):
2928 filename_real_ext
= os
.path
.splitext(filename
)[1][1:]
2930 os
.path
.splitext(filename
)[0]
2931 if filename_real_ext
in (old_ext
, new_ext
)
2933 return '%s.%s' % (filename_wo_ext
, ext
)
2935 # Ensure filename always has a correct extension for successful merge
2936 full_filename
= correct_ext(full_filename
)
2937 temp_filename
= correct_ext(temp_filename
)
2938 dl_filename
= existing_file(full_filename
, temp_filename
)
2939 info_dict
['__real_download'] = False
2942 merger
= FFmpegMergerPP(self
)
2944 fd
= get_suitable_downloader(info_dict
, self
.params
, to_stdout
=temp_filename
== '-')
2945 if dl_filename
is not None:
2946 self
.report_file_already_downloaded(dl_filename
)
2948 for f
in requested_formats
if fd
!= FFmpegFD
else []:
2949 f
['filepath'] = fname
= prepend_extension(
2950 correct_ext(temp_filename
, info_dict
['ext']),
2951 'f%s' % f
['format_id'], info_dict
['ext'])
2952 downloaded
.append(fname
)
2953 info_dict
['url'] = '\n'.join(f
['url'] for f
in requested_formats
)
2954 success
, real_download
= self
.dl(temp_filename
, info_dict
)
2955 info_dict
['__real_download'] = real_download
2957 if self
.params
.get('allow_unplayable_formats'):
2958 self
.report_warning(
2959 'You have requested merging of multiple formats '
2960 'while also allowing unplayable formats to be downloaded. '
2961 'The formats won\'t be merged to prevent data corruption.')
2962 elif not merger
.available
:
2963 self
.report_warning(
2964 'You have requested merging of multiple formats but ffmpeg is not installed. '
2965 'The formats won\'t be merged.')
2967 if temp_filename
== '-':
2968 reason
= ('using a downloader other than ffmpeg' if FFmpegFD
.can_merge_formats(info_dict
, self
.params
)
2969 else 'but the formats are incompatible for simultaneous download' if merger
.available
2970 else 'but ffmpeg is not installed')
2971 self
.report_warning(
2972 f
'You have requested downloading multiple formats to stdout {reason}. '
2973 'The formats will be streamed one after the other')
2974 fname
= temp_filename
2975 for f
in requested_formats
:
2976 new_info
= dict(info_dict
)
2977 del new_info
['requested_formats']
2979 if temp_filename
!= '-':
2980 fname
= prepend_extension(
2981 correct_ext(temp_filename
, new_info
['ext']),
2982 'f%s' % f
['format_id'], new_info
['ext'])
2983 if not self
._ensure
_dir
_exists
(fname
):
2985 f
['filepath'] = fname
2986 downloaded
.append(fname
)
2987 partial_success
, real_download
= self
.dl(fname
, new_info
)
2988 info_dict
['__real_download'] = info_dict
['__real_download'] or real_download
2989 success
= success
and partial_success
2991 if downloaded
and merger
.available
and not self
.params
.get('allow_unplayable_formats'):
2992 info_dict
['__postprocessors'].append(merger
)
2993 info_dict
['__files_to_merge'] = downloaded
2994 # Even if there were no downloads, it is being merged only now
2995 info_dict
['__real_download'] = True
2997 for file in downloaded
:
2998 files_to_move
[file] = None
3000 # Just a single file
3001 dl_filename
= existing_file(full_filename
, temp_filename
)
3002 if dl_filename
is None or dl_filename
== temp_filename
:
3003 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3004 # So we should try to resume the download
3005 success
, real_download
= self
.dl(temp_filename
, info_dict
)
3006 info_dict
['__real_download'] = real_download
3008 self
.report_file_already_downloaded(dl_filename
)
3010 dl_filename
= dl_filename
or temp_filename
3011 info_dict
['__finaldir'] = os
.path
.dirname(os
.path
.abspath(encodeFilename(full_filename
)))
3013 except network_exceptions
as err
:
3014 self
.report_error('unable to download video data: %s' % error_to_compat_str(err
))
3016 except (OSError, IOError) as err
:
3017 raise UnavailableVideoError(err
)
3018 except (ContentTooShortError
, ) as err
:
3019 self
.report_error('content too short (expected %s bytes and served %s)' % (err
.expected
, err
.downloaded
))
3022 if success
and full_filename
!= '-':
3026 fixup_policy
= self
.params
.get('fixup')
3027 vid
= info_dict
['id']
3029 if fixup_policy
in ('ignore', 'never'):
3031 elif fixup_policy
== 'warn':
3033 elif fixup_policy
!= 'force':
3034 assert fixup_policy
in ('detect_or_warn', None)
3035 if not info_dict
.get('__real_download'):
3038 def ffmpeg_fixup(cndn
, msg
, cls
):
3042 self
.report_warning(f
'{vid}: {msg}')
3046 info_dict
['__postprocessors'].append(pp
)
3048 self
.report_warning(f
'{vid}: {msg}. Install ffmpeg to fix this automatically')
3050 stretched_ratio
= info_dict
.get('stretched_ratio')
3052 stretched_ratio
not in (1, None),
3053 f
'Non-uniform pixel ratio {stretched_ratio}',
3054 FFmpegFixupStretchedPP
)
3057 (info_dict
.get('requested_formats') is None
3058 and info_dict
.get('container') == 'm4a_dash'
3059 and info_dict
.get('ext') == 'm4a'),
3060 'writing DASH m4a. Only some players support this container',
3063 downloader
= get_suitable_downloader(info_dict
, self
.params
) if 'protocol' in info_dict
else None
3064 downloader
= downloader
.__name
__ if downloader
else None
3066 if info_dict
.get('requested_formats') is None: # Not necessary if doing merger
3067 ffmpeg_fixup(downloader
== 'HlsFD',
3068 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3070 ffmpeg_fixup(info_dict
.get('is_live') and downloader
== 'DashSegmentsFD',
3071 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP
)
3073 ffmpeg_fixup(downloader
== 'WebSocketFragmentFD', 'Malformed timestamps detected', FFmpegFixupTimestampPP
)
3074 ffmpeg_fixup(downloader
== 'WebSocketFragmentFD', 'Malformed duration detected', FFmpegFixupDurationPP
)
3078 info_dict
= self
.post_process(dl_filename
, info_dict
, files_to_move
)
3079 except PostProcessingError
as err
:
3080 self
.report_error('Postprocessing: %s' % str(err
))
3083 for ph
in self
._post
_hooks
:
3084 ph(info_dict
['filepath'])
3085 except Exception as err
:
3086 self
.report_error('post hooks: %s' % str(err
))
3088 must_record_download_archive
= True
3090 if must_record_download_archive
or self
.params
.get('force_write_download_archive', False):
3091 self
.record_download_archive(info_dict
)
3092 max_downloads
= self
.params
.get('max_downloads')
3093 if max_downloads
is not None and self
._num
_downloads
>= int(max_downloads
):
3094 raise MaxDownloadsReached()
3096 def __download_wrapper(self
, func
):
3097 @functools.wraps(func
)
3098 def wrapper(*args
, **kwargs
):
3100 res
= func(*args
, **kwargs
)
3101 except UnavailableVideoError
as e
:
3102 self
.report_error(e
)
3103 except MaxDownloadsReached
as e
:
3104 self
.to_screen(f
'[info] {e}')
3106 except DownloadCancelled
as e
:
3107 self
.to_screen(f
'[info] {e}')
3108 if not self
.params
.get('break_per_url'):
3111 if self
.params
.get('dump_single_json', False):
3112 self
.post_extract(res
)
3113 self
.to_stdout(json
.dumps(self
.sanitize_info(res
)))
3116 def download(self
, url_list
):
3117 """Download a given list of URLs."""
3118 url_list
= variadic(url_list
) # Passing a single URL is a common mistake
3119 outtmpl
= self
.outtmpl_dict
['default']
3120 if (len(url_list
) > 1
3122 and '%' not in outtmpl
3123 and self
.params
.get('max_downloads') != 1):
3124 raise SameFileError(outtmpl
)
3126 for url
in url_list
:
3127 self
.__download
_wrapper
(self
.extract_info
)(
3128 url
, force_generic_extractor
=self
.params
.get('force_generic_extractor', False))
3130 return self
._download
_retcode
3132 def download_with_info_file(self
, info_filename
):
3133 with contextlib
.closing(fileinput
.FileInput(
3134 [info_filename
], mode
='r',
3135 openhook
=fileinput
.hook_encoded('utf-8'))) as f
:
3136 # FileInput doesn't have a read method, we can't call json.load
3137 info
= self
.sanitize_info(json
.loads('\n'.join(f
)), self
.params
.get('clean_infojson', True))
3139 self
.__download
_wrapper
(self
.process_ie_result
)(info
, download
=True)
3140 except (DownloadError
, EntryNotInPlaylist
, ReExtractInfo
) as e
:
3141 if not isinstance(e
, EntryNotInPlaylist
):
3142 self
.to_stderr('\r')
3143 webpage_url
= info
.get('webpage_url')
3144 if webpage_url
is not None:
3145 self
.report_warning(f
'The info failed to download: {e}; trying with URL {webpage_url}')
3146 return self
.download([webpage_url
])
3149 return self
._download
_retcode
3152 def sanitize_info(info_dict
, remove_private_keys
=False):
3153 ''' Sanitize the infodict for converting to json '''
3154 if info_dict
is None:
3156 info_dict
.setdefault('epoch', int(time
.time()))
3157 remove_keys
= {'__original_infodict'}
# Always remove this since this may contain a copy of the entire dict
3158 keep_keys
= ['_type'] # Always keep this to facilitate load-info-json
3159 if remove_private_keys
:
3161 'requested_formats', 'requested_subtitles', 'requested_entries', 'entries',
3162 'filepath', 'infojson_filename', 'original_url', 'playlist_autonumber',
3164 reject
= lambda k
, v
: k
not in keep_keys
and (
3165 k
.startswith('_') or k
in remove_keys
or v
is None)
3167 reject
= lambda k
, v
: k
in remove_keys
3170 if isinstance(obj
, dict):
3171 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3172 elif isinstance(obj
, (list, tuple, set, LazyList
)):
3173 return list(map(filter_fn
, obj
))
3174 elif obj
is None or isinstance(obj
, (str, int, float, bool)):
3179 return filter_fn(info_dict
)
3182 def filter_requested_info(info_dict
, actually_filter
=True):
3183 ''' Alias of sanitize_info for backward compatibility '''
3184 return YoutubeDL
.sanitize_info(info_dict
, actually_filter
)
3186 def run_pp(self
, pp
, infodict
):
3187 files_to_delete
= []
3188 if '__files_to_move' not in infodict
:
3189 infodict
['__files_to_move'] = {}
3191 files_to_delete
, infodict
= pp
.run(infodict
)
3192 except PostProcessingError
as e
:
3193 # Must be True and not 'only_download'
3194 if self
.params
.get('ignoreerrors') is True:
3195 self
.report_error(e
)
3199 if not files_to_delete
:
3201 if self
.params
.get('keepvideo', False):
3202 for f
in files_to_delete
:
3203 infodict
['__files_to_move'].setdefault(f
, '')
3205 for old_filename
in set(files_to_delete
):
3206 self
.to_screen('Deleting original file %s (pass -k to keep)' % old_filename
)
3208 os
.remove(encodeFilename(old_filename
))
3209 except (IOError, OSError):
3210 self
.report_warning('Unable to remove downloaded original file')
3211 if old_filename
in infodict
['__files_to_move']:
3212 del infodict
['__files_to_move'][old_filename
]
3216 def post_extract(info_dict
):
3217 def actual_post_extract(info_dict
):
3218 if info_dict
.get('_type') in ('playlist', 'multi_video'):
3219 for video_dict
in info_dict
.get('entries', {}):
3220 actual_post_extract(video_dict
or {})
3223 post_extractor
= info_dict
.get('__post_extractor') or (lambda: {})
3224 extra
= post_extractor().items()
3225 info_dict
.update(extra
)
3226 info_dict
.pop('__post_extractor', None)
3228 original_infodict
= info_dict
.get('__original_infodict') or {}
3229 original_infodict
.update(extra
)
3230 original_infodict
.pop('__post_extractor', None)
3232 actual_post_extract(info_dict
or {})
3234 def pre_process(self
, ie_info
, key
='pre_process', files_to_move
=None):
3235 info
= dict(ie_info
)
3236 info
['__files_to_move'] = files_to_move
or {}
3237 for pp
in self
._pps
[key
]:
3238 info
= self
.run_pp(pp
, info
)
3239 return info
, info
.pop('__files_to_move', None)
3241 def post_process(self
, filename
, ie_info
, files_to_move
=None):
3242 """Run all the postprocessors on the given file."""
3243 info
= dict(ie_info
)
3244 info
['filepath'] = filename
3245 info
['__files_to_move'] = files_to_move
or {}
3247 for pp
in ie_info
.get('__postprocessors', []) + self
._pps
['post_process']:
3248 info
= self
.run_pp(pp
, info
)
3249 info
= self
.run_pp(MoveFilesAfterDownloadPP(self
), info
)
3250 del info
['__files_to_move']
3251 for pp
in self
._pps
['after_move']:
3252 info
= self
.run_pp(pp
, info
)
3255 def _make_archive_id(self
, info_dict
):
3256 video_id
= info_dict
.get('id')
3259 # Future-proof against any change in case
3260 # and backwards compatibility with prior versions
3261 extractor
= info_dict
.get('extractor_key') or info_dict
.get('ie_key') # key in a playlist
3262 if extractor
is None:
3263 url
= str_or_none(info_dict
.get('url'))
3266 # Try to find matching extractor for the URL and take its ie_key
3267 for ie_key
, ie
in self
._ies
.items():
3268 if ie
.suitable(url
):
3273 return '%s %s' % (extractor
.lower(), video_id
)
3275 def in_download_archive(self
, info_dict
):
3276 fn
= self
.params
.get('download_archive')
3280 vid_id
= self
._make
_archive
_id
(info_dict
)
3282 return False # Incomplete video information
3284 return vid_id
in self
.archive
3286 def record_download_archive(self
, info_dict
):
3287 fn
= self
.params
.get('download_archive')
3290 vid_id
= self
._make
_archive
_id
(info_dict
)
3292 with locked_file(fn
, 'a', encoding
='utf-8') as archive_file
:
3293 archive_file
.write(vid_id
+ '\n')
3294 self
.archive
.add(vid_id
)
3297 def format_resolution(format
, default
='unknown'):
3298 if format
.get('vcodec') == 'none' and format
.get('acodec') != 'none':
3300 if format
.get('resolution') is not None:
3301 return format
['resolution']
3302 if format
.get('width') and format
.get('height'):
3303 return '%dx%d' % (format
['width'], format
['height'])
3304 elif format
.get('height'):
3305 return '%sp' % format
['height']
3306 elif format
.get('width'):
3307 return '%dx?' % format
['width']
3310 def _format_note(self
, fdict
):
3312 if fdict
.get('ext') in ['f4f', 'f4m']:
3313 res
+= '(unsupported)'
3314 if fdict
.get('language'):
3317 res
+= '[%s]' % fdict
['language']
3318 if fdict
.get('format_note') is not None:
3321 res
+= fdict
['format_note']
3322 if fdict
.get('tbr') is not None:
3325 res
+= '%4dk' % fdict
['tbr']
3326 if fdict
.get('container') is not None:
3329 res
+= '%s container' % fdict
['container']
3330 if (fdict
.get('vcodec') is not None
3331 and fdict
.get('vcodec') != 'none'):
3334 res
+= fdict
['vcodec']
3335 if fdict
.get('vbr') is not None:
3337 elif fdict
.get('vbr') is not None and fdict
.get('abr') is not None:
3339 if fdict
.get('vbr') is not None:
3340 res
+= '%4dk' % fdict
['vbr']
3341 if fdict
.get('fps') is not None:
3344 res
+= '%sfps' % fdict
['fps']
3345 if fdict
.get('acodec') is not None:
3348 if fdict
['acodec'] == 'none':
3351 res
+= '%-5s' % fdict
['acodec']
3352 elif fdict
.get('abr') is not None:
3356 if fdict
.get('abr') is not None:
3357 res
+= '@%3dk' % fdict
['abr']
3358 if fdict
.get('asr') is not None:
3359 res
+= ' (%5dHz)' % fdict
['asr']
3360 if fdict
.get('filesize') is not None:
3363 res
+= format_bytes(fdict
['filesize'])
3364 elif fdict
.get('filesize_approx') is not None:
3367 res
+= '~' + format_bytes(fdict
['filesize_approx'])
3370 def _list_format_headers(self
, *headers
):
3371 if self
.params
.get('listformats_table', True) is not False:
3372 return [self
._format
_screen
(header
, self
.Styles
.HEADERS
) for header
in headers
]
3375 def list_formats(self
, info_dict
):
3376 if not info_dict
.get('formats') and not info_dict
.get('url'):
3377 self
.to_screen('%s has no formats' % info_dict
['id'])
3379 self
.to_screen('[info] Available formats for %s:' % info_dict
['id'])
3381 formats
= info_dict
.get('formats', [info_dict
])
3382 new_format
= self
.params
.get('listformats_table', True) is not False
3384 delim
= self
._format
_screen
('\u2502', self
.Styles
.DELIM
, '|', test_encoding
=True)
3387 self
._format
_screen
(format_field(f
, 'format_id'), self
.Styles
.ID
),
3388 format_field(f
, 'ext'),
3389 format_field(f
, func
=self
.format_resolution
, ignore
=('audio only', 'images')),
3390 format_field(f
, 'fps', '\t%d'),
3391 format_field(f
, 'dynamic_range', '%s', ignore
=(None, 'SDR')).replace('HDR', ''),
3393 format_field(f
, 'filesize', ' \t%s', func
=format_bytes
) + format_field(f
, 'filesize_approx', '~\t%s', func
=format_bytes
),
3394 format_field(f
, 'tbr', '\t%dk'),
3395 shorten_protocol_name(f
.get('protocol', '')),
3397 format_field(f
, 'vcodec', default
='unknown').replace(
3399 'images' if f
.get('acodec') == 'none'
3400 else self
._format
_screen
('audio only', self
.Styles
.SUPPRESS
)),
3401 format_field(f
, 'vbr', '\t%dk'),
3402 format_field(f
, 'acodec', default
='unknown').replace(
3404 '' if f
.get('vcodec') == 'none'
3405 else self
._format
_screen
('video only', self
.Styles
.SUPPRESS
)),
3406 format_field(f
, 'abr', '\t%dk'),
3407 format_field(f
, 'asr', '\t%dHz'),
3409 self
._format
_screen
('UNSUPPORTED', 'light red') if f
.get('ext') in ('f4f', 'f4m') else None,
3410 format_field(f
, 'language', '[%s]'),
3412 format_field(f
, 'format_note'),
3413 format_field(f
, 'container', ignore
=(None, f
.get('ext'))),
3416 ] for f
in formats
if f
.get('preference') is None or f
['preference'] >= -1000]
3417 header_line
= self
._list
_format
_headers
(
3418 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim
, '\tFILESIZE', '\tTBR', 'PROTO',
3419 delim
, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3423 format_field(f
, 'format_id'),
3424 format_field(f
, 'ext'),
3425 self
.format_resolution(f
),
3426 self
._format
_note
(f
)]
3428 if f
.get('preference') is None or f
['preference'] >= -1000]
3429 header_line
= ['format code', 'extension', 'resolution', 'note']
3431 self
.to_stdout(render_table(
3433 extra_gap
=(0 if new_format
else 1),
3434 hide_empty
=new_format
,
3435 delim
=new_format
and self
._format
_screen
('\u2500', self
.Styles
.DELIM
, '-', test_encoding
=True)))
3437 def list_thumbnails(self
, info_dict
):
3438 thumbnails
= list(info_dict
.get('thumbnails'))
3440 self
.to_screen('[info] No thumbnails present for %s' % info_dict
['id'])
3444 '[info] Thumbnails for %s:' % info_dict
['id'])
3445 self
.to_stdout(render_table(
3446 self
._list
_format
_headers
('ID', 'Width', 'Height', 'URL'),
3447 [[t
['id'], t
.get('width', 'unknown'), t
.get('height', 'unknown'), t
['url']] for t
in thumbnails
]))
3449 def list_subtitles(self
, video_id
, subtitles
, name
='subtitles'):
3451 self
.to_screen('%s has no %s' % (video_id
, name
))
3454 'Available %s for %s:' % (name
, video_id
))
3456 def _row(lang
, formats
):
3457 exts
, names
= zip(*((f
['ext'], f
.get('name') or 'unknown') for f
in reversed(formats
)))
3458 if len(set(names
)) == 1:
3459 names
= [] if names
[0] == 'unknown' else names
[:1]
3460 return [lang
, ', '.join(names
), ', '.join(exts
)]
3462 self
.to_stdout(render_table(
3463 self
._list
_format
_headers
('Language', 'Name', 'Formats'),
3464 [_row(lang
, formats
) for lang
, formats
in subtitles
.items()],
3467 def urlopen(self
, req
):
3468 """ Start an HTTP download """
3469 if isinstance(req
, compat_basestring
):
3470 req
= sanitized_Request(req
)
3471 return self
._opener
.open(req
, timeout
=self
._socket
_timeout
)
3473 def print_debug_header(self
):
3474 if not self
.params
.get('verbose'):
3477 def get_encoding(stream
):
3478 ret
= getattr(stream
, 'encoding', 'missing (%s)' % type(stream
).__name
__)
3479 if not supports_terminal_sequences(stream
):
3480 from .compat
import WINDOWS_VT_MODE
3481 ret
+= ' (No VT)' if WINDOWS_VT_MODE
is False else ' (No ANSI)'
3484 encoding_str
= 'Encodings: locale %s, fs %s, out %s, err %s, pref %s' % (
3485 locale
.getpreferredencoding(),
3486 sys
.getfilesystemencoding(),
3487 get_encoding(self
._screen
_file
), get_encoding(self
._err
_file
),
3488 self
.get_encoding())
3490 logger
= self
.params
.get('logger')
3492 write_debug
= lambda msg
: logger
.debug(f
'[debug] {msg}')
3493 write_debug(encoding_str
)
3495 write_string(f
'[debug] {encoding_str}\n', encoding
=None)
3496 write_debug
= lambda msg
: self
._write
_string
(f
'[debug] {msg}\n')
3498 source
= detect_variant()
3499 write_debug(join_nonempty(
3500 'yt-dlp version', __version__
,
3501 f
'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD
else '',
3502 '' if source
== 'unknown' else f
'({source})',
3504 if not _LAZY_LOADER
:
3505 if os
.environ
.get('YTDLP_NO_LAZY_EXTRACTORS'):
3506 write_debug('Lazy loading extractors is forcibly disabled')
3508 write_debug('Lazy loading extractors is disabled')
3509 if plugin_extractors
or plugin_postprocessors
:
3510 write_debug('Plugins: %s' % [
3511 '%s%s' % (klass
.__name
__, '' if klass
.__name
__ == name
else f
' as {name}')
3512 for name
, klass
in itertools
.chain(plugin_extractors
.items(), plugin_postprocessors
.items())])
3513 if self
.params
.get('compat_opts'):
3514 write_debug('Compatibility options: %s' % ', '.join(self
.params
.get('compat_opts')))
3516 if source
== 'source':
3519 ['git', 'rev-parse', '--short', 'HEAD'],
3520 stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
,
3521 cwd
=os
.path
.dirname(os
.path
.abspath(__file__
)))
3522 out
, err
= sp
.communicate_or_kill()
3523 out
= out
.decode().strip()
3524 if re
.match('[0-9a-f]+', out
):
3525 write_debug('Git HEAD: %s' % out
)
3532 def python_implementation():
3533 impl_name
= platform
.python_implementation()
3534 if impl_name
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
3535 return impl_name
+ ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
3538 write_debug('Python version %s (%s %s) - %s' % (
3539 platform
.python_version(),
3540 python_implementation(),
3541 platform
.architecture()[0],
3544 exe_versions
, ffmpeg_features
= FFmpegPostProcessor
.get_versions_and_features(self
)
3545 ffmpeg_features
= {key for key, val in ffmpeg_features.items() if val}
3547 exe_versions
['ffmpeg'] += ' (%s)' % ','.join(ffmpeg_features
)
3549 exe_versions
['rtmpdump'] = rtmpdump_version()
3550 exe_versions
['phantomjs'] = PhantomJSwrapper
._version
()
3551 exe_str
= ', '.join(
3552 f
'{exe} {v}' for exe
, v
in sorted(exe_versions
.items()) if v
3554 write_debug('exe versions: %s' % exe_str
)
3556 from .downloader
.websocket
import has_websockets
3557 from .postprocessor
.embedthumbnail
import has_mutagen
3558 from .cookies
import SQLITE_AVAILABLE
, SECRETSTORAGE_AVAILABLE
3560 lib_str
= join_nonempty(
3561 compat_pycrypto_AES
and compat_pycrypto_AES
.__name
__.split('.')[0],
3562 SECRETSTORAGE_AVAILABLE
and 'secretstorage',
3563 has_mutagen
and 'mutagen',
3564 SQLITE_AVAILABLE
and 'sqlite',
3565 has_websockets
and 'websockets',
3566 delim
=', ') or 'none'
3567 write_debug('Optional libraries: %s' % lib_str
)
3570 for handler
in self
._opener
.handlers
:
3571 if hasattr(handler
, 'proxies'):
3572 proxy_map
.update(handler
.proxies
)
3573 write_debug(f
'Proxy map: {proxy_map}')
3576 if False and self
.params
.get('call_home'):
3577 ipaddr
= self
.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3578 write_debug('Public IP address: %s' % ipaddr
)
3579 latest_version
= self
.urlopen(
3580 'https://yt-dl.org/latest/version').read().decode('utf-8')
3581 if version_tuple(latest_version
) > version_tuple(__version__
):
3582 self
.report_warning(
3583 'You are using an outdated version (newest version: %s)! '
3584 'See https://yt-dl.org/update if you need help updating.' %
3587 def _setup_opener(self
):
3588 timeout_val
= self
.params
.get('socket_timeout')
3589 self
._socket
_timeout
= 20 if timeout_val
is None else float(timeout_val
)
3591 opts_cookiesfrombrowser
= self
.params
.get('cookiesfrombrowser')
3592 opts_cookiefile
= self
.params
.get('cookiefile')
3593 opts_proxy
= self
.params
.get('proxy')
3595 self
.cookiejar
= load_cookies(opts_cookiefile
, opts_cookiesfrombrowser
, self
)
3597 cookie_processor
= YoutubeDLCookieProcessor(self
.cookiejar
)
3598 if opts_proxy
is not None:
3599 if opts_proxy
== '':
3602 proxies
= {'http': opts_proxy, 'https': opts_proxy}
3604 proxies
= compat_urllib_request
.getproxies()
3605 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3606 if 'http' in proxies
and 'https' not in proxies
:
3607 proxies
['https'] = proxies
['http']
3608 proxy_handler
= PerRequestProxyHandler(proxies
)
3610 debuglevel
= 1 if self
.params
.get('debug_printtraffic') else 0
3611 https_handler
= make_HTTPS_handler(self
.params
, debuglevel
=debuglevel
)
3612 ydlh
= YoutubeDLHandler(self
.params
, debuglevel
=debuglevel
)
3613 redirect_handler
= YoutubeDLRedirectHandler()
3614 data_handler
= compat_urllib_request_DataHandler()
3616 # When passing our own FileHandler instance, build_opener won't add the
3617 # default FileHandler and allows us to disable the file protocol, which
3618 # can be used for malicious purposes (see
3619 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3620 file_handler
= compat_urllib_request
.FileHandler()
3622 def file_open(*args
, **kwargs
):
3623 raise compat_urllib_error
.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3624 file_handler
.file_open
= file_open
3626 opener
= compat_urllib_request
.build_opener(
3627 proxy_handler
, https_handler
, cookie_processor
, ydlh
, redirect_handler
, data_handler
, file_handler
)
3629 # Delete the default user-agent header, which would otherwise apply in
3630 # cases where our custom HTTP handler doesn't come into play
3631 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3632 opener
.addheaders
= []
3633 self
._opener
= opener
3635 def encode(self
, s
):
3636 if isinstance(s
, bytes):
3637 return s
# Already encoded
3640 return s
.encode(self
.get_encoding())
3641 except UnicodeEncodeError as err
:
3642 err
.reason
= err
.reason
+ '. Check your system encoding configuration or use the --encoding option.'
3645 def get_encoding(self
):
3646 encoding
= self
.params
.get('encoding')
3647 if encoding
is None:
3648 encoding
= preferredencoding()
3651 def _write_info_json(self
, label
, ie_result
, infofn
, overwrite
=None):
3652 ''' Write infojson and returns True = written, False = skip, None = error '''
3653 if overwrite
is None:
3654 overwrite
= self
.params
.get('overwrites', True)
3655 if not self
.params
.get('writeinfojson'):
3658 self
.write_debug(f
'Skipping writing {label} infojson')
3660 elif not self
._ensure
_dir
_exists
(infofn
):
3662 elif not overwrite
and os
.path
.exists(infofn
):
3663 self
.to_screen(f
'[info] {label.title()} metadata is already present')
3665 self
.to_screen(f
'[info] Writing {label} metadata as JSON to: {infofn}')
3667 write_json_file(self
.sanitize_info(ie_result
, self
.params
.get('clean_infojson', True)), infofn
)
3668 except (OSError, IOError):
3669 self
.report_error(f
'Cannot write {label} metadata to JSON file {infofn}')
3673 def _write_description(self
, label
, ie_result
, descfn
):
3674 ''' Write description and returns True = written, False = skip, None = error '''
3675 if not self
.params
.get('writedescription'):
3678 self
.write_debug(f
'Skipping writing {label} description')
3680 elif not self
._ensure
_dir
_exists
(descfn
):
3682 elif not self
.params
.get('overwrites', True) and os
.path
.exists(descfn
):
3683 self
.to_screen(f
'[info] {label.title()} description is already present')
3684 elif ie_result
.get('description') is None:
3685 self
.report_warning(f
'There\'s no {label} description to write')
3689 self
.to_screen(f
'[info] Writing {label} description to: {descfn}')
3690 with io
.open(encodeFilename(descfn
), 'w', encoding
='utf-8') as descfile
:
3691 descfile
.write(ie_result
['description'])
3692 except (OSError, IOError):
3693 self
.report_error(f
'Cannot write {label} description file {descfn}')
3697 def _write_subtitles(self
, info_dict
, filename
):
3698 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3700 subtitles
= info_dict
.get('requested_subtitles')
3701 if not subtitles
or not (self
.params
.get('writesubtitles') or self
.params
.get('writeautomaticsub')):
3702 # subtitles download errors are already managed as troubles in relevant IE
3703 # that way it will silently go on when used with unsupporting IE
3706 sub_filename_base
= self
.prepare_filename(info_dict
, 'subtitle')
3707 if not sub_filename_base
:
3708 self
.to_screen('[info] Skipping writing video subtitles')
3710 for sub_lang
, sub_info
in subtitles
.items():
3711 sub_format
= sub_info
['ext']
3712 sub_filename
= subtitles_filename(filename
, sub_lang
, sub_format
, info_dict
.get('ext'))
3713 sub_filename_final
= subtitles_filename(sub_filename_base
, sub_lang
, sub_format
, info_dict
.get('ext'))
3714 if not self
.params
.get('overwrites', True) and os
.path
.exists(sub_filename
):
3715 self
.to_screen(f
'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3716 sub_info
['filepath'] = sub_filename
3717 ret
.append((sub_filename
, sub_filename_final
))
3720 self
.to_screen(f
'[info] Writing video subtitles to: {sub_filename}')
3721 if sub_info
.get('data') is not None:
3723 # Use newline='' to prevent conversion of newline characters
3724 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3725 with io
.open(sub_filename
, 'w', encoding
='utf-8', newline
='') as subfile
:
3726 subfile
.write(sub_info
['data'])
3727 sub_info
['filepath'] = sub_filename
3728 ret
.append((sub_filename
, sub_filename_final
))
3730 except (OSError, IOError):
3731 self
.report_error(f
'Cannot write video subtitles file {sub_filename}')
3735 sub_copy
= sub_info
.copy()
3736 sub_copy
.setdefault('http_headers', info_dict
.get('http_headers'))
3737 self
.dl(sub_filename
, sub_copy
, subtitle
=True)
3738 sub_info
['filepath'] = sub_filename
3739 ret
.append((sub_filename
, sub_filename_final
))
3740 except (ExtractorError
, IOError, OSError, ValueError) + network_exceptions
as err
:
3741 self
.report_warning(f
'Unable to download video subtitles for {sub_lang!r}: {err}')
3745 def _write_thumbnails(self
, label
, info_dict
, filename
, thumb_filename_base
=None):
3746 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3747 write_all
= self
.params
.get('write_all_thumbnails', False)
3748 thumbnails
, ret
= [], []
3749 if write_all
or self
.params
.get('writethumbnail', False):
3750 thumbnails
= info_dict
.get('thumbnails') or []
3751 multiple
= write_all
and len(thumbnails
) > 1
3753 if thumb_filename_base
is None:
3754 thumb_filename_base
= filename
3755 if thumbnails
and not thumb_filename_base
:
3756 self
.write_debug(f
'Skipping writing {label} thumbnail')
3759 for idx
, t
in list(enumerate(thumbnails
))[::-1]:
3760 thumb_ext
= (f
'{t["id"]}.' if multiple
else '') + determine_ext(t
['url'], 'jpg')
3761 thumb_display_id
= f
'{label} thumbnail {t["id"]}'
3762 thumb_filename
= replace_extension(filename
, thumb_ext
, info_dict
.get('ext'))
3763 thumb_filename_final
= replace_extension(thumb_filename_base
, thumb_ext
, info_dict
.get('ext'))
3765 if not self
.params
.get('overwrites', True) and os
.path
.exists(thumb_filename
):
3766 ret
.append((thumb_filename
, thumb_filename_final
))
3767 t
['filepath'] = thumb_filename
3768 self
.to_screen('[info] %s is already present' % (
3769 thumb_display_id
if multiple
else f
'{label} thumbnail').capitalize())
3771 self
.to_screen(f
'[info] Downloading {thumb_display_id} ...')
3773 uf
= self
.urlopen(t
['url'])
3774 self
.to_screen(f
'[info] Writing {thumb_display_id} to: {thumb_filename}')
3775 with open(encodeFilename(thumb_filename
), 'wb') as thumbf
:
3776 shutil
.copyfileobj(uf
, thumbf
)
3777 ret
.append((thumb_filename
, thumb_filename_final
))
3778 t
['filepath'] = thumb_filename
3779 except network_exceptions
as err
:
3781 self
.report_warning(f
'Unable to download {thumb_display_id}: {err}')
3782 if ret
and not write_all
: