]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[cleanup] Mark some compat variables for removal (#2173)
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 #!/usr/bin/env python3
2 # coding: utf-8
3
4 from __future__ import absolute_import, unicode_literals
5
6 import collections
7 import contextlib
8 import datetime
9 import errno
10 import fileinput
11 import functools
12 import io
13 import itertools
14 import json
15 import locale
16 import operator
17 import os
18 import platform
19 import re
20 import shutil
21 import subprocess
22 import sys
23 import tempfile
24 import time
25 import tokenize
26 import traceback
27 import random
28 import unicodedata
29 import urllib.request
30
31 from enum import Enum
32 from string import ascii_letters
33
34 from .compat import (
35 compat_brotli,
36 compat_get_terminal_size,
37 compat_os_name,
38 compat_pycrypto_AES,
39 compat_shlex_quote,
40 compat_str,
41 compat_urllib_error,
42 compat_urllib_request,
43 windows_enable_vt_mode,
44 )
45 from .cookies import load_cookies
46 from .utils import (
47 age_restricted,
48 args_to_str,
49 ContentTooShortError,
50 date_from_str,
51 DateRange,
52 DEFAULT_OUTTMPL,
53 determine_ext,
54 determine_protocol,
55 DownloadCancelled,
56 DownloadError,
57 encode_compat_str,
58 encodeFilename,
59 EntryNotInPlaylist,
60 error_to_compat_str,
61 ExistingVideoReached,
62 expand_path,
63 ExtractorError,
64 filter_dict,
65 float_or_none,
66 format_bytes,
67 format_field,
68 format_decimal_suffix,
69 formatSeconds,
70 GeoRestrictedError,
71 get_domain,
72 has_certifi,
73 HEADRequest,
74 InAdvancePagedList,
75 int_or_none,
76 iri_to_uri,
77 ISO3166Utils,
78 join_nonempty,
79 LazyList,
80 LINK_TEMPLATES,
81 locked_file,
82 make_dir,
83 make_HTTPS_handler,
84 MaxDownloadsReached,
85 merge_headers,
86 network_exceptions,
87 NO_DEFAULT,
88 number_of_digits,
89 orderedSet,
90 OUTTMPL_TYPES,
91 PagedList,
92 parse_filesize,
93 PerRequestProxyHandler,
94 platform_name,
95 Popen,
96 POSTPROCESS_WHEN,
97 PostProcessingError,
98 preferredencoding,
99 prepend_extension,
100 ReExtractInfo,
101 register_socks_protocols,
102 RejectedVideoReached,
103 remove_terminal_sequences,
104 render_table,
105 replace_extension,
106 SameFileError,
107 sanitize_filename,
108 sanitize_path,
109 sanitize_url,
110 sanitized_Request,
111 std_headers,
112 STR_FORMAT_RE_TMPL,
113 STR_FORMAT_TYPES,
114 str_or_none,
115 strftime_or_none,
116 subtitles_filename,
117 supports_terminal_sequences,
118 timetuple_from_msec,
119 to_high_limit_path,
120 traverse_obj,
121 try_get,
122 UnavailableVideoError,
123 url_basename,
124 variadic,
125 version_tuple,
126 write_json_file,
127 write_string,
128 YoutubeDLCookieProcessor,
129 YoutubeDLHandler,
130 YoutubeDLRedirectHandler,
131 )
132 from .cache import Cache
133 from .minicurses import format_text
134 from .extractor import (
135 gen_extractor_classes,
136 get_info_extractor,
137 _LAZY_LOADER,
138 _PLUGIN_CLASSES as plugin_extractors
139 )
140 from .extractor.openload import PhantomJSwrapper
141 from .downloader import (
142 FFmpegFD,
143 get_suitable_downloader,
144 shorten_protocol_name
145 )
146 from .downloader.rtmp import rtmpdump_version
147 from .postprocessor import (
148 get_postprocessor,
149 EmbedThumbnailPP,
150 FFmpegFixupDuplicateMoovPP,
151 FFmpegFixupDurationPP,
152 FFmpegFixupM3u8PP,
153 FFmpegFixupM4aPP,
154 FFmpegFixupStretchedPP,
155 FFmpegFixupTimestampPP,
156 FFmpegMergerPP,
157 FFmpegPostProcessor,
158 MoveFilesAfterDownloadPP,
159 _PLUGIN_CLASSES as plugin_postprocessors
160 )
161 from .update import detect_variant
162 from .version import __version__, RELEASE_GIT_HEAD
163
164 if compat_os_name == 'nt':
165 import ctypes
166
167
168 class YoutubeDL(object):
169 """YoutubeDL class.
170
171 YoutubeDL objects are the ones responsible of downloading the
172 actual video file and writing it to disk if the user has requested
173 it, among some other tasks. In most cases there should be one per
174 program. As, given a video URL, the downloader doesn't know how to
175 extract all the needed information, task that InfoExtractors do, it
176 has to pass the URL to one of them.
177
178 For this, YoutubeDL objects have a method that allows
179 InfoExtractors to be registered in a given order. When it is passed
180 a URL, the YoutubeDL object handles it to the first InfoExtractor it
181 finds that reports being able to handle it. The InfoExtractor extracts
182 all the information about the video or videos the URL refers to, and
183 YoutubeDL process the extracted information, possibly using a File
184 Downloader to download the video.
185
186 YoutubeDL objects accept a lot of parameters. In order not to saturate
187 the object constructor with arguments, it receives a dictionary of
188 options instead. These options are available through the params
189 attribute for the InfoExtractors to use. The YoutubeDL also
190 registers itself as the downloader in charge for the InfoExtractors
191 that are added to it, so this is a "mutual registration".
192
193 Available options:
194
195 username: Username for authentication purposes.
196 password: Password for authentication purposes.
197 videopassword: Password for accessing a video.
198 ap_mso: Adobe Pass multiple-system operator identifier.
199 ap_username: Multiple-system operator account username.
200 ap_password: Multiple-system operator account password.
201 usenetrc: Use netrc for authentication instead.
202 verbose: Print additional info to stdout.
203 quiet: Do not print messages to stdout.
204 no_warnings: Do not print out anything for warnings.
205 forceprint: A dict with keys WHEN mapped to a list of templates to
206 print to stdout. The allowed keys are video or any of the
207 items in utils.POSTPROCESS_WHEN.
208 For compatibility, a single list is also accepted
209 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
210 a list of tuples with (template, filename)
211 forceurl: Force printing final URL. (Deprecated)
212 forcetitle: Force printing title. (Deprecated)
213 forceid: Force printing ID. (Deprecated)
214 forcethumbnail: Force printing thumbnail URL. (Deprecated)
215 forcedescription: Force printing description. (Deprecated)
216 forcefilename: Force printing final filename. (Deprecated)
217 forceduration: Force printing duration. (Deprecated)
218 forcejson: Force printing info_dict as JSON.
219 dump_single_json: Force printing the info_dict of the whole playlist
220 (or video) as a single JSON line.
221 force_write_download_archive: Force writing download archive regardless
222 of 'skip_download' or 'simulate'.
223 simulate: Do not download the video files. If unset (or None),
224 simulate only if listsubtitles, listformats or list_thumbnails is used
225 format: Video format code. see "FORMAT SELECTION" for more details.
226 You can also pass a function. The function takes 'ctx' as
227 argument and returns the formats to download.
228 See "build_format_selector" for an implementation
229 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
230 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
231 extracting metadata even if the video is not actually
232 available for download (experimental)
233 format_sort: A list of fields by which to sort the video formats.
234 See "Sorting Formats" for more details.
235 format_sort_force: Force the given format_sort. see "Sorting Formats"
236 for more details.
237 prefer_free_formats: Whether to prefer video formats with free containers
238 over non-free ones of same quality.
239 allow_multiple_video_streams: Allow multiple video streams to be merged
240 into a single file
241 allow_multiple_audio_streams: Allow multiple audio streams to be merged
242 into a single file
243 check_formats Whether to test if the formats are downloadable.
244 Can be True (check all), False (check none),
245 'selected' (check selected formats),
246 or None (check only if requested by extractor)
247 paths: Dictionary of output paths. The allowed keys are 'home'
248 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
249 outtmpl: Dictionary of templates for output names. Allowed keys
250 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
251 For compatibility with youtube-dl, a single string can also be used
252 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
253 restrictfilenames: Do not allow "&" and spaces in file names
254 trim_file_name: Limit length of filename (extension excluded)
255 windowsfilenames: Force the filenames to be windows compatible
256 ignoreerrors: Do not stop on download/postprocessing errors.
257 Can be 'only_download' to ignore only download errors.
258 Default is 'only_download' for CLI, but False for API
259 skip_playlist_after_errors: Number of allowed failures until the rest of
260 the playlist is skipped
261 force_generic_extractor: Force downloader to use the generic extractor
262 overwrites: Overwrite all video and metadata files if True,
263 overwrite only non-video files if None
264 and don't overwrite any file if False
265 For compatibility with youtube-dl,
266 "nooverwrites" may also be used instead
267 playliststart: Playlist item to start at.
268 playlistend: Playlist item to end at.
269 playlist_items: Specific indices of playlist to download.
270 playlistreverse: Download playlist items in reverse order.
271 playlistrandom: Download playlist items in random order.
272 matchtitle: Download only matching titles.
273 rejecttitle: Reject downloads for matching titles.
274 logger: Log messages to a logging.Logger instance.
275 logtostderr: Log messages to stderr instead of stdout.
276 consoletitle: Display progress in console window's titlebar.
277 writedescription: Write the video description to a .description file
278 writeinfojson: Write the video description to a .info.json file
279 clean_infojson: Remove private fields from the infojson
280 getcomments: Extract video comments. This will not be written to disk
281 unless writeinfojson is also given
282 writeannotations: Write the video annotations to a .annotations.xml file
283 writethumbnail: Write the thumbnail image to a file
284 allow_playlist_files: Whether to write playlists' description, infojson etc
285 also to disk when using the 'write*' options
286 write_all_thumbnails: Write all thumbnail formats to files
287 writelink: Write an internet shortcut file, depending on the
288 current platform (.url/.webloc/.desktop)
289 writeurllink: Write a Windows internet shortcut file (.url)
290 writewebloclink: Write a macOS internet shortcut file (.webloc)
291 writedesktoplink: Write a Linux internet shortcut file (.desktop)
292 writesubtitles: Write the video subtitles to a file
293 writeautomaticsub: Write the automatically generated subtitles to a file
294 allsubtitles: Deprecated - Use subtitleslangs = ['all']
295 Downloads all the subtitles of the video
296 (requires writesubtitles or writeautomaticsub)
297 listsubtitles: Lists all available subtitles for the video
298 subtitlesformat: The format code for subtitles
299 subtitleslangs: List of languages of the subtitles to download (can be regex).
300 The list may contain "all" to refer to all the available
301 subtitles. The language can be prefixed with a "-" to
302 exclude it from the requested languages. Eg: ['all', '-live_chat']
303 keepvideo: Keep the video file after post-processing
304 daterange: A DateRange object, download only if the upload_date is in the range.
305 skip_download: Skip the actual download of the video file
306 cachedir: Location of the cache files in the filesystem.
307 False to disable filesystem cache.
308 noplaylist: Download single video instead of a playlist if in doubt.
309 age_limit: An integer representing the user's age in years.
310 Unsuitable videos for the given age are skipped.
311 min_views: An integer representing the minimum view count the video
312 must have in order to not be skipped.
313 Videos without view count information are always
314 downloaded. None for no limit.
315 max_views: An integer representing the maximum view count.
316 Videos that are more popular than that are not
317 downloaded.
318 Videos without view count information are always
319 downloaded. None for no limit.
320 download_archive: File name of a file where all downloads are recorded.
321 Videos already present in the file are not downloaded
322 again.
323 break_on_existing: Stop the download process after attempting to download a
324 file that is in the archive.
325 break_on_reject: Stop the download process when encountering a video that
326 has been filtered out.
327 break_per_url: Whether break_on_reject and break_on_existing
328 should act on each input URL as opposed to for the entire queue
329 cookiefile: File name where cookies should be read from and dumped to
330 cookiesfrombrowser: A tuple containing the name of the browser, the profile
331 name/pathfrom where cookies are loaded, and the name of the
332 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
333 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
334 support RFC 5746 secure renegotiation
335 nocheckcertificate: Do not verify SSL certificates
336 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
337 At the moment, this is only supported by YouTube.
338 http_headers: A dictionary of custom headers to be used for all requests
339 proxy: URL of the proxy server to use
340 geo_verification_proxy: URL of the proxy to use for IP address verification
341 on geo-restricted sites.
342 socket_timeout: Time to wait for unresponsive hosts, in seconds
343 bidi_workaround: Work around buggy terminals without bidirectional text
344 support, using fridibi
345 debug_printtraffic:Print out sent and received HTTP traffic
346 include_ads: Download ads as well (deprecated)
347 default_search: Prepend this string if an input url is not valid.
348 'auto' for elaborate guessing
349 encoding: Use this encoding instead of the system-specified.
350 extract_flat: Do not resolve URLs, return the immediate result.
351 Pass in 'in_playlist' to only show this behavior for
352 playlist items.
353 wait_for_video: If given, wait for scheduled streams to become available.
354 The value should be a tuple containing the range
355 (min_secs, max_secs) to wait between retries
356 postprocessors: A list of dictionaries, each with an entry
357 * key: The name of the postprocessor. See
358 yt_dlp/postprocessor/__init__.py for a list.
359 * when: When to run the postprocessor. Allowed values are
360 the entries of utils.POSTPROCESS_WHEN
361 Assumed to be 'post_process' if not given
362 post_hooks: Deprecated - Register a custom postprocessor instead
363 A list of functions that get called as the final step
364 for each video file, after all postprocessors have been
365 called. The filename will be passed as the only argument.
366 progress_hooks: A list of functions that get called on download
367 progress, with a dictionary with the entries
368 * status: One of "downloading", "error", or "finished".
369 Check this first and ignore unknown values.
370 * info_dict: The extracted info_dict
371
372 If status is one of "downloading", or "finished", the
373 following properties may also be present:
374 * filename: The final filename (always present)
375 * tmpfilename: The filename we're currently writing to
376 * downloaded_bytes: Bytes on disk
377 * total_bytes: Size of the whole file, None if unknown
378 * total_bytes_estimate: Guess of the eventual file size,
379 None if unavailable.
380 * elapsed: The number of seconds since download started.
381 * eta: The estimated time in seconds, None if unknown
382 * speed: The download speed in bytes/second, None if
383 unknown
384 * fragment_index: The counter of the currently
385 downloaded video fragment.
386 * fragment_count: The number of fragments (= individual
387 files that will be merged)
388
389 Progress hooks are guaranteed to be called at least once
390 (with status "finished") if the download is successful.
391 postprocessor_hooks: A list of functions that get called on postprocessing
392 progress, with a dictionary with the entries
393 * status: One of "started", "processing", or "finished".
394 Check this first and ignore unknown values.
395 * postprocessor: Name of the postprocessor
396 * info_dict: The extracted info_dict
397
398 Progress hooks are guaranteed to be called at least twice
399 (with status "started" and "finished") if the processing is successful.
400 merge_output_format: Extension to use when merging formats.
401 final_ext: Expected final extension; used to detect when the file was
402 already downloaded and converted
403 fixup: Automatically correct known faults of the file.
404 One of:
405 - "never": do nothing
406 - "warn": only emit a warning
407 - "detect_or_warn": check whether we can do anything
408 about it, warn otherwise (default)
409 source_address: Client-side IP address to bind to.
410 call_home: Boolean, true iff we are allowed to contact the
411 yt-dlp servers for debugging. (BROKEN)
412 sleep_interval_requests: Number of seconds to sleep between requests
413 during extraction
414 sleep_interval: Number of seconds to sleep before each download when
415 used alone or a lower bound of a range for randomized
416 sleep before each download (minimum possible number
417 of seconds to sleep) when used along with
418 max_sleep_interval.
419 max_sleep_interval:Upper bound of a range for randomized sleep before each
420 download (maximum possible number of seconds to sleep).
421 Must only be used along with sleep_interval.
422 Actual sleep time will be a random float from range
423 [sleep_interval; max_sleep_interval].
424 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
425 listformats: Print an overview of available video formats and exit.
426 list_thumbnails: Print a table of all thumbnails and exit.
427 match_filter: A function that gets called with the info_dict of
428 every video.
429 If it returns a message, the video is ignored.
430 If it returns None, the video is downloaded.
431 match_filter_func in utils.py is one example for this.
432 no_color: Do not emit color codes in output.
433 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
434 HTTP header
435 geo_bypass_country:
436 Two-letter ISO 3166-2 country code that will be used for
437 explicit geographic restriction bypassing via faking
438 X-Forwarded-For HTTP header
439 geo_bypass_ip_block:
440 IP range in CIDR notation that will be used similarly to
441 geo_bypass_country
442
443 The following options determine which downloader is picked:
444 external_downloader: A dictionary of protocol keys and the executable of the
445 external downloader to use for it. The allowed protocols
446 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
447 Set the value to 'native' to use the native downloader
448 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
449 or {'m3u8': 'ffmpeg'} instead.
450 Use the native HLS downloader instead of ffmpeg/avconv
451 if True, otherwise use ffmpeg/avconv if False, otherwise
452 use downloader suggested by extractor if None.
453 compat_opts: Compatibility options. See "Differences in default behavior".
454 The following options do not work when used through the API:
455 filename, abort-on-error, multistreams, no-live-chat, format-sort
456 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
457 Refer __init__.py for their implementation
458 progress_template: Dictionary of templates for progress outputs.
459 Allowed keys are 'download', 'postprocess',
460 'download-title' (console title) and 'postprocess-title'.
461 The template is mapped on a dictionary with keys 'progress' and 'info'
462
463 The following parameters are not used by YoutubeDL itself, they are used by
464 the downloader (see yt_dlp/downloader/common.py):
465 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
466 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
467 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
468 external_downloader_args, concurrent_fragment_downloads.
469
470 The following options are used by the post processors:
471 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
472 otherwise prefer ffmpeg. (avconv support is deprecated)
473 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
474 to the binary or its containing directory.
475 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
476 and a list of additional command-line arguments for the
477 postprocessor/executable. The dict can also have "PP+EXE" keys
478 which are used when the given exe is used by the given PP.
479 Use 'default' as the name for arguments to passed to all PP
480 For compatibility with youtube-dl, a single list of args
481 can also be used
482
483 The following options are used by the extractors:
484 extractor_retries: Number of times to retry for known errors
485 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
486 hls_split_discontinuity: Split HLS playlists to different formats at
487 discontinuities such as ad breaks (default: False)
488 extractor_args: A dictionary of arguments to be passed to the extractors.
489 See "EXTRACTOR ARGUMENTS" for details.
490 Eg: {'youtube': {'skip': ['dash', 'hls']}}
491 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
492 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
493 If True (default), DASH manifests and related
494 data will be downloaded and processed by extractor.
495 You can reduce network I/O by disabling it if you don't
496 care about DASH. (only for youtube)
497 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
498 If True (default), HLS manifests and related
499 data will be downloaded and processed by extractor.
500 You can reduce network I/O by disabling it if you don't
501 care about HLS. (only for youtube)
502 """
503
504 _NUMERIC_FIELDS = set((
505 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
506 'timestamp', 'release_timestamp',
507 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
508 'average_rating', 'comment_count', 'age_limit',
509 'start_time', 'end_time',
510 'chapter_number', 'season_number', 'episode_number',
511 'track_number', 'disc_number', 'release_year',
512 ))
513
514 _format_fields = {
515 # NB: Keep in sync with the docstring of extractor/common.py
516 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
517 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
518 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
519 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
520 'preference', 'language', 'language_preference', 'quality', 'source_preference',
521 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
522 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
523 }
524 _format_selection_exts = {
525 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
526 'video': {'mp4', 'flv', 'webm', '3gp'},
527 'storyboards': {'mhtml'},
528 }
529
530 def __init__(self, params=None, auto_init=True):
531 """Create a FileDownloader object with the given options.
532 @param auto_init Whether to load the default extractors and print header (if verbose).
533 Set to 'no_verbose_header' to not print the header
534 """
535 if params is None:
536 params = {}
537 self.params = params
538 self._ies = {}
539 self._ies_instances = {}
540 self._pps = {k: [] for k in POSTPROCESS_WHEN}
541 self._printed_messages = set()
542 self._first_webpage_request = True
543 self._post_hooks = []
544 self._progress_hooks = []
545 self._postprocessor_hooks = []
546 self._download_retcode = 0
547 self._num_downloads = 0
548 self._num_videos = 0
549 self._playlist_level = 0
550 self._playlist_urls = set()
551 self.cache = Cache(self)
552
553 windows_enable_vt_mode()
554 self._out_files = {
555 'error': sys.stderr,
556 'print': sys.stderr if self.params.get('logtostderr') else sys.stdout,
557 'console': None if compat_os_name == 'nt' else next(
558 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
559 }
560 self._out_files['screen'] = sys.stderr if self.params.get('quiet') else self._out_files['print']
561 self._allow_colors = {
562 type_: not self.params.get('no_color') and supports_terminal_sequences(self._out_files[type_])
563 for type_ in ('screen', 'error')
564 }
565
566 if sys.version_info < (3, 6):
567 self.report_warning(
568 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
569
570 if self.params.get('allow_unplayable_formats'):
571 self.report_warning(
572 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
573 'This is a developer option intended for debugging. \n'
574 ' If you experience any issues while using this option, '
575 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
576
577 def check_deprecated(param, option, suggestion):
578 if self.params.get(param) is not None:
579 self.report_warning('%s is deprecated. Use %s instead' % (option, suggestion))
580 return True
581 return False
582
583 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
584 if self.params.get('geo_verification_proxy') is None:
585 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
586
587 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
588 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
589 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
590
591 for msg in self.params.get('_warnings', []):
592 self.report_warning(msg)
593 for msg in self.params.get('_deprecation_warnings', []):
594 self.deprecation_warning(msg)
595
596 if 'list-formats' in self.params.get('compat_opts', []):
597 self.params['listformats_table'] = False
598
599 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
600 # nooverwrites was unnecessarily changed to overwrites
601 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
602 # This ensures compatibility with both keys
603 self.params['overwrites'] = not self.params['nooverwrites']
604 elif self.params.get('overwrites') is None:
605 self.params.pop('overwrites', None)
606 else:
607 self.params['nooverwrites'] = not self.params['overwrites']
608
609 self.params.setdefault('forceprint', {})
610 self.params.setdefault('print_to_file', {})
611
612 # Compatibility with older syntax
613 if not isinstance(params['forceprint'], dict):
614 self.params['forceprint'] = {'video': params['forceprint']}
615
616 if self.params.get('bidi_workaround', False):
617 try:
618 import pty
619 master, slave = pty.openpty()
620 width = compat_get_terminal_size().columns
621 if width is None:
622 width_args = []
623 else:
624 width_args = ['-w', str(width)]
625 sp_kwargs = dict(
626 stdin=subprocess.PIPE,
627 stdout=slave,
628 stderr=self._out_files['error'])
629 try:
630 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
631 except OSError:
632 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
633 self._output_channel = os.fdopen(master, 'rb')
634 except OSError as ose:
635 if ose.errno == errno.ENOENT:
636 self.report_warning(
637 'Could not find fribidi executable, ignoring --bidi-workaround. '
638 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
639 else:
640 raise
641
642 if auto_init:
643 if auto_init != 'no_verbose_header':
644 self.print_debug_header()
645 self.add_default_info_extractors()
646
647 if (sys.platform != 'win32'
648 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
649 and not self.params.get('restrictfilenames', False)):
650 # Unicode filesystem API will throw errors (#1474, #13027)
651 self.report_warning(
652 'Assuming --restrict-filenames since file system encoding '
653 'cannot encode all characters. '
654 'Set the LC_ALL environment variable to fix this.')
655 self.params['restrictfilenames'] = True
656
657 self.outtmpl_dict = self.parse_outtmpl()
658
659 # Creating format selector here allows us to catch syntax errors before the extraction
660 self.format_selector = (
661 self.params.get('format') if self.params.get('format') in (None, '-')
662 else self.params['format'] if callable(self.params['format'])
663 else self.build_format_selector(self.params['format']))
664
665 # Set http_headers defaults according to std_headers
666 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
667
668 hooks = {
669 'post_hooks': self.add_post_hook,
670 'progress_hooks': self.add_progress_hook,
671 'postprocessor_hooks': self.add_postprocessor_hook,
672 }
673 for opt, fn in hooks.items():
674 for ph in self.params.get(opt, []):
675 fn(ph)
676
677 for pp_def_raw in self.params.get('postprocessors', []):
678 pp_def = dict(pp_def_raw)
679 when = pp_def.pop('when', 'post_process')
680 self.add_post_processor(
681 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
682 when=when)
683
684 self._setup_opener()
685 register_socks_protocols()
686
687 def preload_download_archive(fn):
688 """Preload the archive, if any is specified"""
689 if fn is None:
690 return False
691 self.write_debug(f'Loading archive file {fn!r}')
692 try:
693 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
694 for line in archive_file:
695 self.archive.add(line.strip())
696 except IOError as ioe:
697 if ioe.errno != errno.ENOENT:
698 raise
699 return False
700 return True
701
702 self.archive = set()
703 preload_download_archive(self.params.get('download_archive'))
704
705 def warn_if_short_id(self, argv):
706 # short YouTube ID starting with dash?
707 idxs = [
708 i for i, a in enumerate(argv)
709 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
710 if idxs:
711 correct_argv = (
712 ['yt-dlp']
713 + [a for i, a in enumerate(argv) if i not in idxs]
714 + ['--'] + [argv[i] for i in idxs]
715 )
716 self.report_warning(
717 'Long argument string detected. '
718 'Use -- to separate parameters and URLs, like this:\n%s' %
719 args_to_str(correct_argv))
720
721 def add_info_extractor(self, ie):
722 """Add an InfoExtractor object to the end of the list."""
723 ie_key = ie.ie_key()
724 self._ies[ie_key] = ie
725 if not isinstance(ie, type):
726 self._ies_instances[ie_key] = ie
727 ie.set_downloader(self)
728
729 def _get_info_extractor_class(self, ie_key):
730 ie = self._ies.get(ie_key)
731 if ie is None:
732 ie = get_info_extractor(ie_key)
733 self.add_info_extractor(ie)
734 return ie
735
736 def get_info_extractor(self, ie_key):
737 """
738 Get an instance of an IE with name ie_key, it will try to get one from
739 the _ies list, if there's no instance it will create a new one and add
740 it to the extractor list.
741 """
742 ie = self._ies_instances.get(ie_key)
743 if ie is None:
744 ie = get_info_extractor(ie_key)()
745 self.add_info_extractor(ie)
746 return ie
747
748 def add_default_info_extractors(self):
749 """
750 Add the InfoExtractors returned by gen_extractors to the end of the list
751 """
752 for ie in gen_extractor_classes():
753 self.add_info_extractor(ie)
754
755 def add_post_processor(self, pp, when='post_process'):
756 """Add a PostProcessor object to the end of the chain."""
757 self._pps[when].append(pp)
758 pp.set_downloader(self)
759
760 def add_post_hook(self, ph):
761 """Add the post hook"""
762 self._post_hooks.append(ph)
763
764 def add_progress_hook(self, ph):
765 """Add the download progress hook"""
766 self._progress_hooks.append(ph)
767
768 def add_postprocessor_hook(self, ph):
769 """Add the postprocessing progress hook"""
770 self._postprocessor_hooks.append(ph)
771 for pps in self._pps.values():
772 for pp in pps:
773 pp.add_progress_hook(ph)
774
775 def _bidi_workaround(self, message):
776 if not hasattr(self, '_output_channel'):
777 return message
778
779 assert hasattr(self, '_output_process')
780 assert isinstance(message, compat_str)
781 line_count = message.count('\n') + 1
782 self._output_process.stdin.write((message + '\n').encode('utf-8'))
783 self._output_process.stdin.flush()
784 res = ''.join(self._output_channel.readline().decode('utf-8')
785 for _ in range(line_count))
786 return res[:-len('\n')]
787
788 def _write_string(self, message, out=None, only_once=False):
789 if only_once:
790 if message in self._printed_messages:
791 return
792 self._printed_messages.add(message)
793 write_string(message, out=out, encoding=self.params.get('encoding'))
794
795 def to_stdout(self, message, skip_eol=False, quiet=None):
796 """Print message to stdout"""
797 if quiet is not None:
798 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
799 self._write_string(
800 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
801 self._out_files['print'])
802
803 def to_screen(self, message, skip_eol=False, quiet=None):
804 """Print message to screen if not in quiet mode"""
805 if self.params.get('logger'):
806 self.params['logger'].debug(message)
807 return
808 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
809 return
810 self._write_string(
811 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
812 self._out_files['screen'])
813
814 def to_stderr(self, message, only_once=False):
815 """Print message to stderr"""
816 assert isinstance(message, compat_str)
817 if self.params.get('logger'):
818 self.params['logger'].error(message)
819 else:
820 self._write_string('%s\n' % self._bidi_workaround(message), self._out_files['error'], only_once=only_once)
821
822 def _send_console_code(self, code):
823 if compat_os_name == 'nt' or not self._out_files['console']:
824 return
825 self._write_string(code, self._out_files['console'])
826
827 def to_console_title(self, message):
828 if not self.params.get('consoletitle', False):
829 return
830 message = remove_terminal_sequences(message)
831 if compat_os_name == 'nt':
832 if ctypes.windll.kernel32.GetConsoleWindow():
833 # c_wchar_p() might not be necessary if `message` is
834 # already of type unicode()
835 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
836 else:
837 self._send_console_code(f'\033]0;{message}\007')
838
839 def save_console_title(self):
840 if not self.params.get('consoletitle') or self.params.get('simulate'):
841 return
842 self._send_console_code('\033[22;0t') # Save the title on stack
843
844 def restore_console_title(self):
845 if not self.params.get('consoletitle') or self.params.get('simulate'):
846 return
847 self._send_console_code('\033[23;0t') # Restore the title from stack
848
849 def __enter__(self):
850 self.save_console_title()
851 return self
852
853 def __exit__(self, *args):
854 self.restore_console_title()
855
856 if self.params.get('cookiefile') is not None:
857 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
858
859 def trouble(self, message=None, tb=None, is_error=True):
860 """Determine action to take when a download problem appears.
861
862 Depending on if the downloader has been configured to ignore
863 download errors or not, this method may throw an exception or
864 not when errors are found, after printing the message.
865
866 @param tb If given, is additional traceback information
867 @param is_error Whether to raise error according to ignorerrors
868 """
869 if message is not None:
870 self.to_stderr(message)
871 if self.params.get('verbose'):
872 if tb is None:
873 if sys.exc_info()[0]: # if .trouble has been called from an except block
874 tb = ''
875 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
876 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
877 tb += encode_compat_str(traceback.format_exc())
878 else:
879 tb_data = traceback.format_list(traceback.extract_stack())
880 tb = ''.join(tb_data)
881 if tb:
882 self.to_stderr(tb)
883 if not is_error:
884 return
885 if not self.params.get('ignoreerrors'):
886 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
887 exc_info = sys.exc_info()[1].exc_info
888 else:
889 exc_info = sys.exc_info()
890 raise DownloadError(message, exc_info)
891 self._download_retcode = 1
892
893 class Styles(Enum):
894 HEADERS = 'yellow'
895 EMPHASIS = 'light blue'
896 ID = 'green'
897 DELIM = 'blue'
898 ERROR = 'red'
899 WARNING = 'yellow'
900 SUPPRESS = 'light black'
901
902 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
903 if test_encoding:
904 original_text = text
905 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
906 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
907 text = text.encode(encoding, 'ignore').decode(encoding)
908 if fallback is not None and text != original_text:
909 text = fallback
910 if isinstance(f, self.Styles):
911 f = f.value
912 return format_text(text, f) if allow_colors else text if fallback is None else fallback
913
914 def _format_screen(self, *args, **kwargs):
915 return self._format_text(
916 self._out_files['screen'], self._allow_colors['screen'], *args, **kwargs)
917
918 def _format_err(self, *args, **kwargs):
919 return self._format_text(
920 self._out_files['error'], self._allow_colors['error'], *args, **kwargs)
921
922 def report_warning(self, message, only_once=False):
923 '''
924 Print the message to stderr, it will be prefixed with 'WARNING:'
925 If stderr is a tty file the 'WARNING:' will be colored
926 '''
927 if self.params.get('logger') is not None:
928 self.params['logger'].warning(message)
929 else:
930 if self.params.get('no_warnings'):
931 return
932 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
933
934 def deprecation_warning(self, message):
935 if self.params.get('logger') is not None:
936 self.params['logger'].warning(f'DeprecationWarning: {message}')
937 else:
938 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
939
940 def report_error(self, message, *args, **kwargs):
941 '''
942 Do the same as trouble, but prefixes the message with 'ERROR:', colored
943 in red if stderr is a tty file.
944 '''
945 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
946
947 def write_debug(self, message, only_once=False):
948 '''Log debug message or Print message to stderr'''
949 if not self.params.get('verbose', False):
950 return
951 message = '[debug] %s' % message
952 if self.params.get('logger'):
953 self.params['logger'].debug(message)
954 else:
955 self.to_stderr(message, only_once)
956
957 def report_file_already_downloaded(self, file_name):
958 """Report file has already been fully downloaded."""
959 try:
960 self.to_screen('[download] %s has already been downloaded' % file_name)
961 except UnicodeEncodeError:
962 self.to_screen('[download] The file has already been downloaded')
963
964 def report_file_delete(self, file_name):
965 """Report that existing file will be deleted."""
966 try:
967 self.to_screen('Deleting existing file %s' % file_name)
968 except UnicodeEncodeError:
969 self.to_screen('Deleting existing file')
970
971 def raise_no_formats(self, info, forced=False, *, msg=None):
972 has_drm = info.get('__has_drm')
973 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
974 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
975 if forced or not ignored:
976 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
977 expected=has_drm or ignored or expected)
978 else:
979 self.report_warning(msg)
980
981 def parse_outtmpl(self):
982 outtmpl_dict = self.params.get('outtmpl', {})
983 if not isinstance(outtmpl_dict, dict):
984 outtmpl_dict = {'default': outtmpl_dict}
985 # Remove spaces in the default template
986 if self.params.get('restrictfilenames'):
987 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
988 else:
989 sanitize = lambda x: x
990 outtmpl_dict.update({
991 k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items()
992 if outtmpl_dict.get(k) is None})
993 for key, val in outtmpl_dict.items():
994 if isinstance(val, bytes):
995 self.report_warning(
996 'Parameter outtmpl is bytes, but should be a unicode string. '
997 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
998 return outtmpl_dict
999
1000 def get_output_path(self, dir_type='', filename=None):
1001 paths = self.params.get('paths', {})
1002 assert isinstance(paths, dict)
1003 path = os.path.join(
1004 expand_path(paths.get('home', '').strip()),
1005 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1006 filename or '')
1007 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1008
1009 @staticmethod
1010 def _outtmpl_expandpath(outtmpl):
1011 # expand_path translates '%%' into '%' and '$$' into '$'
1012 # correspondingly that is not what we want since we need to keep
1013 # '%%' intact for template dict substitution step. Working around
1014 # with boundary-alike separator hack.
1015 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1016 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
1017
1018 # outtmpl should be expand_path'ed before template dict substitution
1019 # because meta fields may contain env variables we don't want to
1020 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1021 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1022 return expand_path(outtmpl).replace(sep, '')
1023
1024 @staticmethod
1025 def escape_outtmpl(outtmpl):
1026 ''' Escape any remaining strings like %s, %abc% etc. '''
1027 return re.sub(
1028 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1029 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1030 outtmpl)
1031
1032 @classmethod
1033 def validate_outtmpl(cls, outtmpl):
1034 ''' @return None or Exception object '''
1035 outtmpl = re.sub(
1036 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1037 lambda mobj: f'{mobj.group(0)[:-1]}s',
1038 cls._outtmpl_expandpath(outtmpl))
1039 try:
1040 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1041 return None
1042 except ValueError as err:
1043 return err
1044
1045 @staticmethod
1046 def _copy_infodict(info_dict):
1047 info_dict = dict(info_dict)
1048 info_dict.pop('__postprocessors', None)
1049 return info_dict
1050
1051 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1052 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1053 @param sanitize Whether to sanitize the output as a filename.
1054 For backward compatibility, a function can also be passed
1055 """
1056
1057 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1058
1059 info_dict = self._copy_infodict(info_dict)
1060 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1061 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1062 if info_dict.get('duration', None) is not None
1063 else None)
1064 info_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
1065 info_dict['video_autonumber'] = self._num_videos
1066 if info_dict.get('resolution') is None:
1067 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1068
1069 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1070 # of %(field)s to %(field)0Nd for backward compatibility
1071 field_size_compat_map = {
1072 'playlist_index': number_of_digits(info_dict.get('_last_playlist_index') or 0),
1073 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1074 'autonumber': self.params.get('autonumber_size') or 5,
1075 }
1076
1077 TMPL_DICT = {}
1078 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1079 MATH_FUNCTIONS = {
1080 '+': float.__add__,
1081 '-': float.__sub__,
1082 }
1083 # Field is of the form key1.key2...
1084 # where keys (except first) can be string, int or slice
1085 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1086 MATH_FIELD_RE = r'''(?:{field}|{num})'''.format(field=FIELD_RE, num=r'-?\d+(?:.\d+)?')
1087 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1088 INTERNAL_FORMAT_RE = re.compile(r'''(?x)
1089 (?P<negate>-)?
1090 (?P<fields>{field})
1091 (?P<maths>(?:{math_op}{math_field})*)
1092 (?:>(?P<strf_format>.+?))?
1093 (?P<remaining>
1094 (?P<alternate>(?<!\\),[^|&)]+)?
1095 (?:&(?P<replacement>.*?))?
1096 (?:\|(?P<default>.*?))?
1097 )$'''.format(field=FIELD_RE, math_op=MATH_OPERATORS_RE, math_field=MATH_FIELD_RE))
1098
1099 def _traverse_infodict(k):
1100 k = k.split('.')
1101 if k[0] == '':
1102 k.pop(0)
1103 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1104
1105 def get_value(mdict):
1106 # Object traversal
1107 value = _traverse_infodict(mdict['fields'])
1108 # Negative
1109 if mdict['negate']:
1110 value = float_or_none(value)
1111 if value is not None:
1112 value *= -1
1113 # Do maths
1114 offset_key = mdict['maths']
1115 if offset_key:
1116 value = float_or_none(value)
1117 operator = None
1118 while offset_key:
1119 item = re.match(
1120 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1121 offset_key).group(0)
1122 offset_key = offset_key[len(item):]
1123 if operator is None:
1124 operator = MATH_FUNCTIONS[item]
1125 continue
1126 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1127 offset = float_or_none(item)
1128 if offset is None:
1129 offset = float_or_none(_traverse_infodict(item))
1130 try:
1131 value = operator(value, multiplier * offset)
1132 except (TypeError, ZeroDivisionError):
1133 return None
1134 operator = None
1135 # Datetime formatting
1136 if mdict['strf_format']:
1137 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1138
1139 return value
1140
1141 na = self.params.get('outtmpl_na_placeholder', 'NA')
1142
1143 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1144 return sanitize_filename(str(value), restricted=restricted, is_id=(
1145 bool(re.search(r'(^|[_.])id(\.|$)', key))
1146 if 'filename-sanitization' in self.params.get('compat_opts', [])
1147 else NO_DEFAULT))
1148
1149 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1150 sanitize = bool(sanitize)
1151
1152 def _dumpjson_default(obj):
1153 if isinstance(obj, (set, LazyList)):
1154 return list(obj)
1155 return repr(obj)
1156
1157 def create_key(outer_mobj):
1158 if not outer_mobj.group('has_key'):
1159 return outer_mobj.group(0)
1160 key = outer_mobj.group('key')
1161 mobj = re.match(INTERNAL_FORMAT_RE, key)
1162 initial_field = mobj.group('fields') if mobj else ''
1163 value, replacement, default = None, None, na
1164 while mobj:
1165 mobj = mobj.groupdict()
1166 default = mobj['default'] if mobj['default'] is not None else default
1167 value = get_value(mobj)
1168 replacement = mobj['replacement']
1169 if value is None and mobj['alternate']:
1170 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1171 else:
1172 break
1173
1174 fmt = outer_mobj.group('format')
1175 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1176 fmt = '0{:d}d'.format(field_size_compat_map[key])
1177
1178 value = default if value is None else value if replacement is None else replacement
1179
1180 flags = outer_mobj.group('conversion') or ''
1181 str_fmt = f'{fmt[:-1]}s'
1182 if fmt[-1] == 'l': # list
1183 delim = '\n' if '#' in flags else ', '
1184 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1185 elif fmt[-1] == 'j': # json
1186 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1187 elif fmt[-1] == 'q': # quoted
1188 value = map(str, variadic(value) if '#' in flags else [value])
1189 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1190 elif fmt[-1] == 'B': # bytes
1191 value = f'%{str_fmt}'.encode('utf-8') % str(value).encode('utf-8')
1192 value, fmt = value.decode('utf-8', 'ignore'), 's'
1193 elif fmt[-1] == 'U': # unicode normalized
1194 value, fmt = unicodedata.normalize(
1195 # "+" = compatibility equivalence, "#" = NFD
1196 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1197 value), str_fmt
1198 elif fmt[-1] == 'D': # decimal suffix
1199 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1200 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1201 factor=1024 if '#' in flags else 1000)
1202 elif fmt[-1] == 'S': # filename sanitization
1203 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1204 elif fmt[-1] == 'c':
1205 if value:
1206 value = str(value)[0]
1207 else:
1208 fmt = str_fmt
1209 elif fmt[-1] not in 'rs': # numeric
1210 value = float_or_none(value)
1211 if value is None:
1212 value, fmt = default, 's'
1213
1214 if sanitize:
1215 if fmt[-1] == 'r':
1216 # If value is an object, sanitize might convert it to a string
1217 # So we convert it to repr first
1218 value, fmt = repr(value), str_fmt
1219 if fmt[-1] in 'csr':
1220 value = sanitizer(initial_field, value)
1221
1222 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1223 TMPL_DICT[key] = value
1224 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1225
1226 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1227
1228 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1229 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1230 return self.escape_outtmpl(outtmpl) % info_dict
1231
1232 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1233 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1234 if outtmpl is None:
1235 outtmpl = self.outtmpl_dict.get(tmpl_type or 'default', self.outtmpl_dict['default'])
1236 try:
1237 outtmpl = self._outtmpl_expandpath(outtmpl)
1238 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1239 if not filename:
1240 return None
1241
1242 if tmpl_type in ('', 'temp'):
1243 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1244 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1245 filename = replace_extension(filename, ext, final_ext)
1246 elif tmpl_type:
1247 force_ext = OUTTMPL_TYPES[tmpl_type]
1248 if force_ext:
1249 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1250
1251 # https://github.com/blackjack4494/youtube-dlc/issues/85
1252 trim_file_name = self.params.get('trim_file_name', False)
1253 if trim_file_name:
1254 no_ext, *ext = filename.rsplit('.', 2)
1255 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1256
1257 return filename
1258 except ValueError as err:
1259 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1260 return None
1261
1262 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1263 """Generate the output filename"""
1264 if outtmpl:
1265 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1266 dir_type = None
1267 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1268 if not filename and dir_type not in ('', 'temp'):
1269 return ''
1270
1271 if warn:
1272 if not self.params.get('paths'):
1273 pass
1274 elif filename == '-':
1275 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1276 elif os.path.isabs(filename):
1277 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1278 if filename == '-' or not filename:
1279 return filename
1280
1281 return self.get_output_path(dir_type, filename)
1282
1283 def _match_entry(self, info_dict, incomplete=False, silent=False):
1284 """ Returns None if the file should be downloaded """
1285
1286 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1287
1288 def check_filter():
1289 if 'title' in info_dict:
1290 # This can happen when we're just evaluating the playlist
1291 title = info_dict['title']
1292 matchtitle = self.params.get('matchtitle', False)
1293 if matchtitle:
1294 if not re.search(matchtitle, title, re.IGNORECASE):
1295 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1296 rejecttitle = self.params.get('rejecttitle', False)
1297 if rejecttitle:
1298 if re.search(rejecttitle, title, re.IGNORECASE):
1299 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1300 date = info_dict.get('upload_date')
1301 if date is not None:
1302 dateRange = self.params.get('daterange', DateRange())
1303 if date not in dateRange:
1304 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
1305 view_count = info_dict.get('view_count')
1306 if view_count is not None:
1307 min_views = self.params.get('min_views')
1308 if min_views is not None and view_count < min_views:
1309 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1310 max_views = self.params.get('max_views')
1311 if max_views is not None and view_count > max_views:
1312 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1313 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1314 return 'Skipping "%s" because it is age restricted' % video_title
1315
1316 match_filter = self.params.get('match_filter')
1317 if match_filter is not None:
1318 try:
1319 ret = match_filter(info_dict, incomplete=incomplete)
1320 except TypeError:
1321 # For backward compatibility
1322 ret = None if incomplete else match_filter(info_dict)
1323 if ret is not None:
1324 return ret
1325 return None
1326
1327 if self.in_download_archive(info_dict):
1328 reason = '%s has already been recorded in the archive' % video_title
1329 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1330 else:
1331 reason = check_filter()
1332 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1333 if reason is not None:
1334 if not silent:
1335 self.to_screen('[download] ' + reason)
1336 if self.params.get(break_opt, False):
1337 raise break_err()
1338 return reason
1339
1340 @staticmethod
1341 def add_extra_info(info_dict, extra_info):
1342 '''Set the keys from extra_info in info dict if they are missing'''
1343 for key, value in extra_info.items():
1344 info_dict.setdefault(key, value)
1345
1346 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1347 process=True, force_generic_extractor=False):
1348 """
1349 Return a list with a dictionary for each video extracted.
1350
1351 Arguments:
1352 url -- URL to extract
1353
1354 Keyword arguments:
1355 download -- whether to download videos during extraction
1356 ie_key -- extractor key hint
1357 extra_info -- dictionary containing the extra values to add to each result
1358 process -- whether to resolve all unresolved references (URLs, playlist items),
1359 must be True for download to work.
1360 force_generic_extractor -- force using the generic extractor
1361 """
1362
1363 if extra_info is None:
1364 extra_info = {}
1365
1366 if not ie_key and force_generic_extractor:
1367 ie_key = 'Generic'
1368
1369 if ie_key:
1370 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1371 else:
1372 ies = self._ies
1373
1374 for ie_key, ie in ies.items():
1375 if not ie.suitable(url):
1376 continue
1377
1378 if not ie.working():
1379 self.report_warning('The program functionality for this site has been marked as broken, '
1380 'and will probably not work.')
1381
1382 temp_id = ie.get_temp_id(url)
1383 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1384 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1385 if self.params.get('break_on_existing', False):
1386 raise ExistingVideoReached()
1387 break
1388 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1389 else:
1390 self.report_error('no suitable InfoExtractor for URL %s' % url)
1391
1392 def __handle_extraction_exceptions(func):
1393 @functools.wraps(func)
1394 def wrapper(self, *args, **kwargs):
1395 while True:
1396 try:
1397 return func(self, *args, **kwargs)
1398 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1399 raise
1400 except ReExtractInfo as e:
1401 if e.expected:
1402 self.to_screen(f'{e}; Re-extracting data')
1403 else:
1404 self.to_stderr('\r')
1405 self.report_warning(f'{e}; Re-extracting data')
1406 continue
1407 except GeoRestrictedError as e:
1408 msg = e.msg
1409 if e.countries:
1410 msg += '\nThis video is available in %s.' % ', '.join(
1411 map(ISO3166Utils.short2full, e.countries))
1412 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1413 self.report_error(msg)
1414 except ExtractorError as e: # An error we somewhat expected
1415 self.report_error(str(e), e.format_traceback())
1416 except Exception as e:
1417 if self.params.get('ignoreerrors'):
1418 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1419 else:
1420 raise
1421 break
1422 return wrapper
1423
1424 def _wait_for_video(self, ie_result):
1425 if (not self.params.get('wait_for_video')
1426 or ie_result.get('_type', 'video') != 'video'
1427 or ie_result.get('formats') or ie_result.get('url')):
1428 return
1429
1430 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1431 last_msg = ''
1432
1433 def progress(msg):
1434 nonlocal last_msg
1435 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1436 last_msg = msg
1437
1438 min_wait, max_wait = self.params.get('wait_for_video')
1439 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1440 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1441 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1442 self.report_warning('Release time of video is not known')
1443 elif (diff or 0) <= 0:
1444 self.report_warning('Video should already be available according to extracted info')
1445 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1446 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1447
1448 wait_till = time.time() + diff
1449 try:
1450 while True:
1451 diff = wait_till - time.time()
1452 if diff <= 0:
1453 progress('')
1454 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1455 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1456 time.sleep(1)
1457 except KeyboardInterrupt:
1458 progress('')
1459 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1460 except BaseException as e:
1461 if not isinstance(e, ReExtractInfo):
1462 self.to_screen('')
1463 raise
1464
1465 @__handle_extraction_exceptions
1466 def __extract_info(self, url, ie, download, extra_info, process):
1467 ie_result = ie.extract(url)
1468 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1469 return
1470 if isinstance(ie_result, list):
1471 # Backwards compatibility: old IE result format
1472 ie_result = {
1473 '_type': 'compat_list',
1474 'entries': ie_result,
1475 }
1476 if extra_info.get('original_url'):
1477 ie_result.setdefault('original_url', extra_info['original_url'])
1478 self.add_default_extra_info(ie_result, ie, url)
1479 if process:
1480 self._wait_for_video(ie_result)
1481 return self.process_ie_result(ie_result, download, extra_info)
1482 else:
1483 return ie_result
1484
1485 def add_default_extra_info(self, ie_result, ie, url):
1486 if url is not None:
1487 self.add_extra_info(ie_result, {
1488 'webpage_url': url,
1489 'original_url': url,
1490 })
1491 webpage_url = ie_result.get('webpage_url')
1492 if webpage_url:
1493 self.add_extra_info(ie_result, {
1494 'webpage_url_basename': url_basename(webpage_url),
1495 'webpage_url_domain': get_domain(webpage_url),
1496 })
1497 if ie is not None:
1498 self.add_extra_info(ie_result, {
1499 'extractor': ie.IE_NAME,
1500 'extractor_key': ie.ie_key(),
1501 })
1502
1503 def process_ie_result(self, ie_result, download=True, extra_info=None):
1504 """
1505 Take the result of the ie(may be modified) and resolve all unresolved
1506 references (URLs, playlist items).
1507
1508 It will also download the videos if 'download'.
1509 Returns the resolved ie_result.
1510 """
1511 if extra_info is None:
1512 extra_info = {}
1513 result_type = ie_result.get('_type', 'video')
1514
1515 if result_type in ('url', 'url_transparent'):
1516 ie_result['url'] = sanitize_url(ie_result['url'])
1517 if ie_result.get('original_url'):
1518 extra_info.setdefault('original_url', ie_result['original_url'])
1519
1520 extract_flat = self.params.get('extract_flat', False)
1521 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1522 or extract_flat is True):
1523 info_copy = ie_result.copy()
1524 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1525 if ie and not ie_result.get('id'):
1526 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1527 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1528 self.add_extra_info(info_copy, extra_info)
1529 info_copy, _ = self.pre_process(info_copy)
1530 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1531 if self.params.get('force_write_download_archive', False):
1532 self.record_download_archive(info_copy)
1533 return ie_result
1534
1535 if result_type == 'video':
1536 self.add_extra_info(ie_result, extra_info)
1537 ie_result = self.process_video_result(ie_result, download=download)
1538 additional_urls = (ie_result or {}).get('additional_urls')
1539 if additional_urls:
1540 # TODO: Improve MetadataParserPP to allow setting a list
1541 if isinstance(additional_urls, compat_str):
1542 additional_urls = [additional_urls]
1543 self.to_screen(
1544 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1545 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1546 ie_result['additional_entries'] = [
1547 self.extract_info(
1548 url, download, extra_info=extra_info,
1549 force_generic_extractor=self.params.get('force_generic_extractor'))
1550 for url in additional_urls
1551 ]
1552 return ie_result
1553 elif result_type == 'url':
1554 # We have to add extra_info to the results because it may be
1555 # contained in a playlist
1556 return self.extract_info(
1557 ie_result['url'], download,
1558 ie_key=ie_result.get('ie_key'),
1559 extra_info=extra_info)
1560 elif result_type == 'url_transparent':
1561 # Use the information from the embedding page
1562 info = self.extract_info(
1563 ie_result['url'], ie_key=ie_result.get('ie_key'),
1564 extra_info=extra_info, download=False, process=False)
1565
1566 # extract_info may return None when ignoreerrors is enabled and
1567 # extraction failed with an error, don't crash and return early
1568 # in this case
1569 if not info:
1570 return info
1571
1572 new_result = info.copy()
1573 new_result.update(filter_dict(ie_result, lambda k, v: (
1574 v is not None and k not in {'_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'})))
1575
1576 # Extracted info may not be a video result (i.e.
1577 # info.get('_type', 'video') != video) but rather an url or
1578 # url_transparent. In such cases outer metadata (from ie_result)
1579 # should be propagated to inner one (info). For this to happen
1580 # _type of info should be overridden with url_transparent. This
1581 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1582 if new_result.get('_type') == 'url':
1583 new_result['_type'] = 'url_transparent'
1584
1585 return self.process_ie_result(
1586 new_result, download=download, extra_info=extra_info)
1587 elif result_type in ('playlist', 'multi_video'):
1588 # Protect from infinite recursion due to recursively nested playlists
1589 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1590 webpage_url = ie_result['webpage_url']
1591 if webpage_url in self._playlist_urls:
1592 self.to_screen(
1593 '[download] Skipping already downloaded playlist: %s'
1594 % ie_result.get('title') or ie_result.get('id'))
1595 return
1596
1597 self._playlist_level += 1
1598 self._playlist_urls.add(webpage_url)
1599 self._fill_common_fields(ie_result, False)
1600 self._sanitize_thumbnails(ie_result)
1601 try:
1602 return self.__process_playlist(ie_result, download)
1603 finally:
1604 self._playlist_level -= 1
1605 if not self._playlist_level:
1606 self._playlist_urls.clear()
1607 elif result_type == 'compat_list':
1608 self.report_warning(
1609 'Extractor %s returned a compat_list result. '
1610 'It needs to be updated.' % ie_result.get('extractor'))
1611
1612 def _fixup(r):
1613 self.add_extra_info(r, {
1614 'extractor': ie_result['extractor'],
1615 'webpage_url': ie_result['webpage_url'],
1616 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1617 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1618 'extractor_key': ie_result['extractor_key'],
1619 })
1620 return r
1621 ie_result['entries'] = [
1622 self.process_ie_result(_fixup(r), download, extra_info)
1623 for r in ie_result['entries']
1624 ]
1625 return ie_result
1626 else:
1627 raise Exception('Invalid result type: %s' % result_type)
1628
1629 def _ensure_dir_exists(self, path):
1630 return make_dir(path, self.report_error)
1631
1632 @staticmethod
1633 def _playlist_infodict(ie_result, **kwargs):
1634 return {
1635 **ie_result,
1636 'playlist': ie_result.get('title') or ie_result.get('id'),
1637 'playlist_id': ie_result.get('id'),
1638 'playlist_title': ie_result.get('title'),
1639 'playlist_uploader': ie_result.get('uploader'),
1640 'playlist_uploader_id': ie_result.get('uploader_id'),
1641 'playlist_index': 0,
1642 **kwargs,
1643 }
1644
1645 def __process_playlist(self, ie_result, download):
1646 # We process each entry in the playlist
1647 playlist = ie_result.get('title') or ie_result.get('id')
1648 self.to_screen('[download] Downloading playlist: %s' % playlist)
1649
1650 if 'entries' not in ie_result:
1651 raise EntryNotInPlaylist('There are no entries')
1652
1653 MissingEntry = object()
1654 incomplete_entries = bool(ie_result.get('requested_entries'))
1655 if incomplete_entries:
1656 def fill_missing_entries(entries, indices):
1657 ret = [MissingEntry] * max(indices)
1658 for i, entry in zip(indices, entries):
1659 ret[i - 1] = entry
1660 return ret
1661 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1662
1663 playlist_results = []
1664
1665 playliststart = self.params.get('playliststart', 1)
1666 playlistend = self.params.get('playlistend')
1667 # For backwards compatibility, interpret -1 as whole list
1668 if playlistend == -1:
1669 playlistend = None
1670
1671 playlistitems_str = self.params.get('playlist_items')
1672 playlistitems = None
1673 if playlistitems_str is not None:
1674 def iter_playlistitems(format):
1675 for string_segment in format.split(','):
1676 if '-' in string_segment:
1677 start, end = string_segment.split('-')
1678 for item in range(int(start), int(end) + 1):
1679 yield int(item)
1680 else:
1681 yield int(string_segment)
1682 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1683
1684 ie_entries = ie_result['entries']
1685 if isinstance(ie_entries, list):
1686 playlist_count = len(ie_entries)
1687 msg = f'Collected {playlist_count} videos; downloading %d of them'
1688 ie_result['playlist_count'] = ie_result.get('playlist_count') or playlist_count
1689
1690 def get_entry(i):
1691 return ie_entries[i - 1]
1692 else:
1693 msg = 'Downloading %d videos'
1694 if not isinstance(ie_entries, (PagedList, LazyList)):
1695 ie_entries = LazyList(ie_entries)
1696 elif isinstance(ie_entries, InAdvancePagedList):
1697 if ie_entries._pagesize == 1:
1698 playlist_count = ie_entries._pagecount
1699
1700 def get_entry(i):
1701 return YoutubeDL.__handle_extraction_exceptions(
1702 lambda self, i: ie_entries[i - 1]
1703 )(self, i)
1704
1705 entries, broken = [], False
1706 items = playlistitems if playlistitems is not None else itertools.count(playliststart)
1707 for i in items:
1708 if i == 0:
1709 continue
1710 if playlistitems is None and playlistend is not None and playlistend < i:
1711 break
1712 entry = None
1713 try:
1714 entry = get_entry(i)
1715 if entry is MissingEntry:
1716 raise EntryNotInPlaylist()
1717 except (IndexError, EntryNotInPlaylist):
1718 if incomplete_entries:
1719 raise EntryNotInPlaylist(f'Entry {i} cannot be found')
1720 elif not playlistitems:
1721 break
1722 entries.append(entry)
1723 try:
1724 if entry is not None:
1725 self._match_entry(entry, incomplete=True, silent=True)
1726 except (ExistingVideoReached, RejectedVideoReached):
1727 broken = True
1728 break
1729 ie_result['entries'] = entries
1730
1731 # Save playlist_index before re-ordering
1732 entries = [
1733 ((playlistitems[i - 1] if playlistitems else i + playliststart - 1), entry)
1734 for i, entry in enumerate(entries, 1)
1735 if entry is not None]
1736 n_entries = len(entries)
1737
1738 if not (ie_result.get('playlist_count') or broken or playlistitems or playlistend):
1739 ie_result['playlist_count'] = n_entries
1740
1741 if not playlistitems and (playliststart != 1 or playlistend):
1742 playlistitems = list(range(playliststart, playliststart + n_entries))
1743 ie_result['requested_entries'] = playlistitems
1744
1745 _infojson_written = False
1746 write_playlist_files = self.params.get('allow_playlist_files', True)
1747 if write_playlist_files and self.params.get('list_thumbnails'):
1748 self.list_thumbnails(ie_result)
1749 if write_playlist_files and not self.params.get('simulate'):
1750 ie_copy = self._playlist_infodict(ie_result, n_entries=n_entries)
1751 _infojson_written = self._write_info_json(
1752 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1753 if _infojson_written is None:
1754 return
1755 if self._write_description('playlist', ie_result,
1756 self.prepare_filename(ie_copy, 'pl_description')) is None:
1757 return
1758 # TODO: This should be passed to ThumbnailsConvertor if necessary
1759 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1760
1761 if self.params.get('playlistreverse', False):
1762 entries = entries[::-1]
1763 if self.params.get('playlistrandom', False):
1764 random.shuffle(entries)
1765
1766 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1767
1768 self.to_screen('[%s] playlist %s: %s' % (ie_result['extractor'], playlist, msg % n_entries))
1769 failures = 0
1770 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1771 for i, entry_tuple in enumerate(entries, 1):
1772 playlist_index, entry = entry_tuple
1773 if 'playlist-index' in self.params.get('compat_opts', []):
1774 playlist_index = playlistitems[i - 1] if playlistitems else i + playliststart - 1
1775 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1776 # This __x_forwarded_for_ip thing is a bit ugly but requires
1777 # minimal changes
1778 if x_forwarded_for:
1779 entry['__x_forwarded_for_ip'] = x_forwarded_for
1780 extra = {
1781 'n_entries': n_entries,
1782 '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1783 'playlist_count': ie_result.get('playlist_count'),
1784 'playlist_index': playlist_index,
1785 'playlist_autonumber': i,
1786 'playlist': playlist,
1787 'playlist_id': ie_result.get('id'),
1788 'playlist_title': ie_result.get('title'),
1789 'playlist_uploader': ie_result.get('uploader'),
1790 'playlist_uploader_id': ie_result.get('uploader_id'),
1791 'extractor': ie_result['extractor'],
1792 'webpage_url': ie_result['webpage_url'],
1793 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1794 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1795 'extractor_key': ie_result['extractor_key'],
1796 }
1797
1798 if self._match_entry(entry, incomplete=True) is not None:
1799 continue
1800
1801 entry_result = self.__process_iterable_entry(entry, download, extra)
1802 if not entry_result:
1803 failures += 1
1804 if failures >= max_failures:
1805 self.report_error(
1806 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1807 break
1808 playlist_results.append(entry_result)
1809 ie_result['entries'] = playlist_results
1810
1811 # Write the updated info to json
1812 if _infojson_written is True and self._write_info_json(
1813 'updated playlist', ie_result,
1814 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1815 return
1816
1817 ie_result = self.run_all_pps('playlist', ie_result)
1818 self.to_screen(f'[download] Finished downloading playlist: {playlist}')
1819 return ie_result
1820
1821 @__handle_extraction_exceptions
1822 def __process_iterable_entry(self, entry, download, extra_info):
1823 return self.process_ie_result(
1824 entry, download=download, extra_info=extra_info)
1825
1826 def _build_format_filter(self, filter_spec):
1827 " Returns a function to filter the formats according to the filter_spec "
1828
1829 OPERATORS = {
1830 '<': operator.lt,
1831 '<=': operator.le,
1832 '>': operator.gt,
1833 '>=': operator.ge,
1834 '=': operator.eq,
1835 '!=': operator.ne,
1836 }
1837 operator_rex = re.compile(r'''(?x)\s*
1838 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1839 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1840 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1841 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1842 m = operator_rex.fullmatch(filter_spec)
1843 if m:
1844 try:
1845 comparison_value = int(m.group('value'))
1846 except ValueError:
1847 comparison_value = parse_filesize(m.group('value'))
1848 if comparison_value is None:
1849 comparison_value = parse_filesize(m.group('value') + 'B')
1850 if comparison_value is None:
1851 raise ValueError(
1852 'Invalid value %r in format specification %r' % (
1853 m.group('value'), filter_spec))
1854 op = OPERATORS[m.group('op')]
1855
1856 if not m:
1857 STR_OPERATORS = {
1858 '=': operator.eq,
1859 '^=': lambda attr, value: attr.startswith(value),
1860 '$=': lambda attr, value: attr.endswith(value),
1861 '*=': lambda attr, value: value in attr,
1862 '~=': lambda attr, value: value.search(attr) is not None
1863 }
1864 str_operator_rex = re.compile(r'''(?x)\s*
1865 (?P<key>[a-zA-Z0-9._-]+)\s*
1866 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1867 (?P<quote>["'])?
1868 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1869 (?(quote)(?P=quote))\s*
1870 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1871 m = str_operator_rex.fullmatch(filter_spec)
1872 if m:
1873 if m.group('op') == '~=':
1874 comparison_value = re.compile(m.group('value'))
1875 else:
1876 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1877 str_op = STR_OPERATORS[m.group('op')]
1878 if m.group('negation'):
1879 op = lambda attr, value: not str_op(attr, value)
1880 else:
1881 op = str_op
1882
1883 if not m:
1884 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1885
1886 def _filter(f):
1887 actual_value = f.get(m.group('key'))
1888 if actual_value is None:
1889 return m.group('none_inclusive')
1890 return op(actual_value, comparison_value)
1891 return _filter
1892
1893 def _check_formats(self, formats):
1894 for f in formats:
1895 self.to_screen('[info] Testing format %s' % f['format_id'])
1896 path = self.get_output_path('temp')
1897 if not self._ensure_dir_exists(f'{path}/'):
1898 continue
1899 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1900 temp_file.close()
1901 try:
1902 success, _ = self.dl(temp_file.name, f, test=True)
1903 except (DownloadError, IOError, OSError, ValueError) + network_exceptions:
1904 success = False
1905 finally:
1906 if os.path.exists(temp_file.name):
1907 try:
1908 os.remove(temp_file.name)
1909 except OSError:
1910 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1911 if success:
1912 yield f
1913 else:
1914 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1915
1916 def _default_format_spec(self, info_dict, download=True):
1917
1918 def can_merge():
1919 merger = FFmpegMergerPP(self)
1920 return merger.available and merger.can_merge()
1921
1922 prefer_best = (
1923 not self.params.get('simulate')
1924 and download
1925 and (
1926 not can_merge()
1927 or info_dict.get('is_live', False)
1928 or self.outtmpl_dict['default'] == '-'))
1929 compat = (
1930 prefer_best
1931 or self.params.get('allow_multiple_audio_streams', False)
1932 or 'format-spec' in self.params.get('compat_opts', []))
1933
1934 return (
1935 'best/bestvideo+bestaudio' if prefer_best
1936 else 'bestvideo*+bestaudio/best' if not compat
1937 else 'bestvideo+bestaudio/best')
1938
1939 def build_format_selector(self, format_spec):
1940 def syntax_error(note, start):
1941 message = (
1942 'Invalid format specification: '
1943 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1944 return SyntaxError(message)
1945
1946 PICKFIRST = 'PICKFIRST'
1947 MERGE = 'MERGE'
1948 SINGLE = 'SINGLE'
1949 GROUP = 'GROUP'
1950 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1951
1952 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1953 'video': self.params.get('allow_multiple_video_streams', False)}
1954
1955 check_formats = self.params.get('check_formats') == 'selected'
1956
1957 def _parse_filter(tokens):
1958 filter_parts = []
1959 for type, string, start, _, _ in tokens:
1960 if type == tokenize.OP and string == ']':
1961 return ''.join(filter_parts)
1962 else:
1963 filter_parts.append(string)
1964
1965 def _remove_unused_ops(tokens):
1966 # Remove operators that we don't use and join them with the surrounding strings
1967 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1968 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1969 last_string, last_start, last_end, last_line = None, None, None, None
1970 for type, string, start, end, line in tokens:
1971 if type == tokenize.OP and string == '[':
1972 if last_string:
1973 yield tokenize.NAME, last_string, last_start, last_end, last_line
1974 last_string = None
1975 yield type, string, start, end, line
1976 # everything inside brackets will be handled by _parse_filter
1977 for type, string, start, end, line in tokens:
1978 yield type, string, start, end, line
1979 if type == tokenize.OP and string == ']':
1980 break
1981 elif type == tokenize.OP and string in ALLOWED_OPS:
1982 if last_string:
1983 yield tokenize.NAME, last_string, last_start, last_end, last_line
1984 last_string = None
1985 yield type, string, start, end, line
1986 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1987 if not last_string:
1988 last_string = string
1989 last_start = start
1990 last_end = end
1991 else:
1992 last_string += string
1993 if last_string:
1994 yield tokenize.NAME, last_string, last_start, last_end, last_line
1995
1996 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1997 selectors = []
1998 current_selector = None
1999 for type, string, start, _, _ in tokens:
2000 # ENCODING is only defined in python 3.x
2001 if type == getattr(tokenize, 'ENCODING', None):
2002 continue
2003 elif type in [tokenize.NAME, tokenize.NUMBER]:
2004 current_selector = FormatSelector(SINGLE, string, [])
2005 elif type == tokenize.OP:
2006 if string == ')':
2007 if not inside_group:
2008 # ')' will be handled by the parentheses group
2009 tokens.restore_last_token()
2010 break
2011 elif inside_merge and string in ['/', ',']:
2012 tokens.restore_last_token()
2013 break
2014 elif inside_choice and string == ',':
2015 tokens.restore_last_token()
2016 break
2017 elif string == ',':
2018 if not current_selector:
2019 raise syntax_error('"," must follow a format selector', start)
2020 selectors.append(current_selector)
2021 current_selector = None
2022 elif string == '/':
2023 if not current_selector:
2024 raise syntax_error('"/" must follow a format selector', start)
2025 first_choice = current_selector
2026 second_choice = _parse_format_selection(tokens, inside_choice=True)
2027 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2028 elif string == '[':
2029 if not current_selector:
2030 current_selector = FormatSelector(SINGLE, 'best', [])
2031 format_filter = _parse_filter(tokens)
2032 current_selector.filters.append(format_filter)
2033 elif string == '(':
2034 if current_selector:
2035 raise syntax_error('Unexpected "("', start)
2036 group = _parse_format_selection(tokens, inside_group=True)
2037 current_selector = FormatSelector(GROUP, group, [])
2038 elif string == '+':
2039 if not current_selector:
2040 raise syntax_error('Unexpected "+"', start)
2041 selector_1 = current_selector
2042 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2043 if not selector_2:
2044 raise syntax_error('Expected a selector', start)
2045 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2046 else:
2047 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
2048 elif type == tokenize.ENDMARKER:
2049 break
2050 if current_selector:
2051 selectors.append(current_selector)
2052 return selectors
2053
2054 def _merge(formats_pair):
2055 format_1, format_2 = formats_pair
2056
2057 formats_info = []
2058 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2059 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2060
2061 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2062 get_no_more = {'video': False, 'audio': False}
2063 for (i, fmt_info) in enumerate(formats_info):
2064 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2065 formats_info.pop(i)
2066 continue
2067 for aud_vid in ['audio', 'video']:
2068 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2069 if get_no_more[aud_vid]:
2070 formats_info.pop(i)
2071 break
2072 get_no_more[aud_vid] = True
2073
2074 if len(formats_info) == 1:
2075 return formats_info[0]
2076
2077 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2078 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2079
2080 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2081 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2082
2083 output_ext = self.params.get('merge_output_format')
2084 if not output_ext:
2085 if the_only_video:
2086 output_ext = the_only_video['ext']
2087 elif the_only_audio and not video_fmts:
2088 output_ext = the_only_audio['ext']
2089 else:
2090 output_ext = 'mkv'
2091
2092 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2093
2094 new_dict = {
2095 'requested_formats': formats_info,
2096 'format': '+'.join(filtered('format')),
2097 'format_id': '+'.join(filtered('format_id')),
2098 'ext': output_ext,
2099 'protocol': '+'.join(map(determine_protocol, formats_info)),
2100 'language': '+'.join(orderedSet(filtered('language'))) or None,
2101 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2102 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2103 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2104 }
2105
2106 if the_only_video:
2107 new_dict.update({
2108 'width': the_only_video.get('width'),
2109 'height': the_only_video.get('height'),
2110 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2111 'fps': the_only_video.get('fps'),
2112 'dynamic_range': the_only_video.get('dynamic_range'),
2113 'vcodec': the_only_video.get('vcodec'),
2114 'vbr': the_only_video.get('vbr'),
2115 'stretched_ratio': the_only_video.get('stretched_ratio'),
2116 })
2117
2118 if the_only_audio:
2119 new_dict.update({
2120 'acodec': the_only_audio.get('acodec'),
2121 'abr': the_only_audio.get('abr'),
2122 'asr': the_only_audio.get('asr'),
2123 })
2124
2125 return new_dict
2126
2127 def _check_formats(formats):
2128 if not check_formats:
2129 yield from formats
2130 return
2131 yield from self._check_formats(formats)
2132
2133 def _build_selector_function(selector):
2134 if isinstance(selector, list): # ,
2135 fs = [_build_selector_function(s) for s in selector]
2136
2137 def selector_function(ctx):
2138 for f in fs:
2139 yield from f(ctx)
2140 return selector_function
2141
2142 elif selector.type == GROUP: # ()
2143 selector_function = _build_selector_function(selector.selector)
2144
2145 elif selector.type == PICKFIRST: # /
2146 fs = [_build_selector_function(s) for s in selector.selector]
2147
2148 def selector_function(ctx):
2149 for f in fs:
2150 picked_formats = list(f(ctx))
2151 if picked_formats:
2152 return picked_formats
2153 return []
2154
2155 elif selector.type == MERGE: # +
2156 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2157
2158 def selector_function(ctx):
2159 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2160 yield _merge(pair)
2161
2162 elif selector.type == SINGLE: # atom
2163 format_spec = selector.selector or 'best'
2164
2165 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2166 if format_spec == 'all':
2167 def selector_function(ctx):
2168 yield from _check_formats(ctx['formats'][::-1])
2169 elif format_spec == 'mergeall':
2170 def selector_function(ctx):
2171 formats = list(_check_formats(
2172 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2173 if not formats:
2174 return
2175 merged_format = formats[-1]
2176 for f in formats[-2::-1]:
2177 merged_format = _merge((merged_format, f))
2178 yield merged_format
2179
2180 else:
2181 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2182 mobj = re.match(
2183 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2184 format_spec)
2185 if mobj is not None:
2186 format_idx = int_or_none(mobj.group('n'), default=1)
2187 format_reverse = mobj.group('bw')[0] == 'b'
2188 format_type = (mobj.group('type') or [None])[0]
2189 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2190 format_modified = mobj.group('mod') is not None
2191
2192 format_fallback = not format_type and not format_modified # for b, w
2193 _filter_f = (
2194 (lambda f: f.get('%scodec' % format_type) != 'none')
2195 if format_type and format_modified # bv*, ba*, wv*, wa*
2196 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2197 if format_type # bv, ba, wv, wa
2198 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2199 if not format_modified # b, w
2200 else lambda f: True) # b*, w*
2201 filter_f = lambda f: _filter_f(f) and (
2202 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2203 else:
2204 if format_spec in self._format_selection_exts['audio']:
2205 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2206 elif format_spec in self._format_selection_exts['video']:
2207 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2208 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2209 elif format_spec in self._format_selection_exts['storyboards']:
2210 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2211 else:
2212 filter_f = lambda f: f.get('format_id') == format_spec # id
2213
2214 def selector_function(ctx):
2215 formats = list(ctx['formats'])
2216 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2217 if not matches:
2218 if format_fallback and ctx['incomplete_formats']:
2219 # for extractors with incomplete formats (audio only (soundcloud)
2220 # or video only (imgur)) best/worst will fallback to
2221 # best/worst {video,audio}-only format
2222 matches = formats
2223 elif seperate_fallback and not ctx['has_merged_format']:
2224 # for compatibility with youtube-dl when there is no pre-merged format
2225 matches = list(filter(seperate_fallback, formats))
2226 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2227 try:
2228 yield matches[format_idx - 1]
2229 except LazyList.IndexError:
2230 return
2231
2232 filters = [self._build_format_filter(f) for f in selector.filters]
2233
2234 def final_selector(ctx):
2235 ctx_copy = dict(ctx)
2236 for _filter in filters:
2237 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2238 return selector_function(ctx_copy)
2239 return final_selector
2240
2241 stream = io.BytesIO(format_spec.encode('utf-8'))
2242 try:
2243 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2244 except tokenize.TokenError:
2245 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2246
2247 class TokenIterator(object):
2248 def __init__(self, tokens):
2249 self.tokens = tokens
2250 self.counter = 0
2251
2252 def __iter__(self):
2253 return self
2254
2255 def __next__(self):
2256 if self.counter >= len(self.tokens):
2257 raise StopIteration()
2258 value = self.tokens[self.counter]
2259 self.counter += 1
2260 return value
2261
2262 next = __next__
2263
2264 def restore_last_token(self):
2265 self.counter -= 1
2266
2267 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2268 return _build_selector_function(parsed_selector)
2269
2270 def _calc_headers(self, info_dict):
2271 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2272
2273 cookies = self._calc_cookies(info_dict)
2274 if cookies:
2275 res['Cookie'] = cookies
2276
2277 if 'X-Forwarded-For' not in res:
2278 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2279 if x_forwarded_for_ip:
2280 res['X-Forwarded-For'] = x_forwarded_for_ip
2281
2282 return res
2283
2284 def _calc_cookies(self, info_dict):
2285 pr = sanitized_Request(info_dict['url'])
2286 self.cookiejar.add_cookie_header(pr)
2287 return pr.get_header('Cookie')
2288
2289 def _sort_thumbnails(self, thumbnails):
2290 thumbnails.sort(key=lambda t: (
2291 t.get('preference') if t.get('preference') is not None else -1,
2292 t.get('width') if t.get('width') is not None else -1,
2293 t.get('height') if t.get('height') is not None else -1,
2294 t.get('id') if t.get('id') is not None else '',
2295 t.get('url')))
2296
2297 def _sanitize_thumbnails(self, info_dict):
2298 thumbnails = info_dict.get('thumbnails')
2299 if thumbnails is None:
2300 thumbnail = info_dict.get('thumbnail')
2301 if thumbnail:
2302 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2303 if not thumbnails:
2304 return
2305
2306 def check_thumbnails(thumbnails):
2307 for t in thumbnails:
2308 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2309 try:
2310 self.urlopen(HEADRequest(t['url']))
2311 except network_exceptions as err:
2312 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2313 continue
2314 yield t
2315
2316 self._sort_thumbnails(thumbnails)
2317 for i, t in enumerate(thumbnails):
2318 if t.get('id') is None:
2319 t['id'] = '%d' % i
2320 if t.get('width') and t.get('height'):
2321 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2322 t['url'] = sanitize_url(t['url'])
2323
2324 if self.params.get('check_formats') is True:
2325 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2326 else:
2327 info_dict['thumbnails'] = thumbnails
2328
2329 def _fill_common_fields(self, info_dict, is_video=True):
2330 # TODO: move sanitization here
2331 if is_video:
2332 # playlists are allowed to lack "title"
2333 info_dict['fulltitle'] = info_dict.get('title')
2334 if 'title' not in info_dict:
2335 raise ExtractorError('Missing "title" field in extractor result',
2336 video_id=info_dict['id'], ie=info_dict['extractor'])
2337 elif not info_dict.get('title'):
2338 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2339 info_dict['title'] = f'{info_dict["extractor"]} video #{info_dict["id"]}'
2340
2341 if info_dict.get('duration') is not None:
2342 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2343
2344 for ts_key, date_key in (
2345 ('timestamp', 'upload_date'),
2346 ('release_timestamp', 'release_date'),
2347 ('modified_timestamp', 'modified_date'),
2348 ):
2349 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2350 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2351 # see http://bugs.python.org/issue1646728)
2352 try:
2353 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2354 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2355 except (ValueError, OverflowError, OSError):
2356 pass
2357
2358 live_keys = ('is_live', 'was_live')
2359 live_status = info_dict.get('live_status')
2360 if live_status is None:
2361 for key in live_keys:
2362 if info_dict.get(key) is False:
2363 continue
2364 if info_dict.get(key):
2365 live_status = key
2366 break
2367 if all(info_dict.get(key) is False for key in live_keys):
2368 live_status = 'not_live'
2369 if live_status:
2370 info_dict['live_status'] = live_status
2371 for key in live_keys:
2372 if info_dict.get(key) is None:
2373 info_dict[key] = (live_status == key)
2374
2375 # Auto generate title fields corresponding to the *_number fields when missing
2376 # in order to always have clean titles. This is very common for TV series.
2377 for field in ('chapter', 'season', 'episode'):
2378 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2379 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2380
2381 def process_video_result(self, info_dict, download=True):
2382 assert info_dict.get('_type', 'video') == 'video'
2383 self._num_videos += 1
2384
2385 if 'id' not in info_dict:
2386 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2387 elif not info_dict.get('id'):
2388 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2389
2390 def report_force_conversion(field, field_not, conversion):
2391 self.report_warning(
2392 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2393 % (field, field_not, conversion))
2394
2395 def sanitize_string_field(info, string_field):
2396 field = info.get(string_field)
2397 if field is None or isinstance(field, compat_str):
2398 return
2399 report_force_conversion(string_field, 'a string', 'string')
2400 info[string_field] = compat_str(field)
2401
2402 def sanitize_numeric_fields(info):
2403 for numeric_field in self._NUMERIC_FIELDS:
2404 field = info.get(numeric_field)
2405 if field is None or isinstance(field, (int, float)):
2406 continue
2407 report_force_conversion(numeric_field, 'numeric', 'int')
2408 info[numeric_field] = int_or_none(field)
2409
2410 sanitize_string_field(info_dict, 'id')
2411 sanitize_numeric_fields(info_dict)
2412 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2413 self.report_warning('"duration" field is negative, there is an error in extractor')
2414
2415 if 'playlist' not in info_dict:
2416 # It isn't part of a playlist
2417 info_dict['playlist'] = None
2418 info_dict['playlist_index'] = None
2419
2420 self._sanitize_thumbnails(info_dict)
2421
2422 thumbnail = info_dict.get('thumbnail')
2423 thumbnails = info_dict.get('thumbnails')
2424 if thumbnail:
2425 info_dict['thumbnail'] = sanitize_url(thumbnail)
2426 elif thumbnails:
2427 info_dict['thumbnail'] = thumbnails[-1]['url']
2428
2429 if info_dict.get('display_id') is None and 'id' in info_dict:
2430 info_dict['display_id'] = info_dict['id']
2431
2432 self._fill_common_fields(info_dict)
2433
2434 for cc_kind in ('subtitles', 'automatic_captions'):
2435 cc = info_dict.get(cc_kind)
2436 if cc:
2437 for _, subtitle in cc.items():
2438 for subtitle_format in subtitle:
2439 if subtitle_format.get('url'):
2440 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2441 if subtitle_format.get('ext') is None:
2442 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2443
2444 automatic_captions = info_dict.get('automatic_captions')
2445 subtitles = info_dict.get('subtitles')
2446
2447 info_dict['requested_subtitles'] = self.process_subtitles(
2448 info_dict['id'], subtitles, automatic_captions)
2449
2450 if info_dict.get('formats') is None:
2451 # There's only one format available
2452 formats = [info_dict]
2453 else:
2454 formats = info_dict['formats']
2455
2456 info_dict['__has_drm'] = any(f.get('has_drm') for f in formats)
2457 if not self.params.get('allow_unplayable_formats'):
2458 formats = [f for f in formats if not f.get('has_drm')]
2459 if info_dict['__has_drm'] and all(
2460 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2461 self.report_warning(
2462 'This video is DRM protected and only images are available for download. '
2463 'Use --list-formats to see them')
2464
2465 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2466 if not get_from_start:
2467 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2468 if info_dict.get('is_live') and formats:
2469 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2470 if get_from_start and not formats:
2471 self.raise_no_formats(info_dict, msg=(
2472 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2473 'If you want to download from the current time, use --no-live-from-start'))
2474
2475 if not formats:
2476 self.raise_no_formats(info_dict)
2477
2478 def is_wellformed(f):
2479 url = f.get('url')
2480 if not url:
2481 self.report_warning(
2482 '"url" field is missing or empty - skipping format, '
2483 'there is an error in extractor')
2484 return False
2485 if isinstance(url, bytes):
2486 sanitize_string_field(f, 'url')
2487 return True
2488
2489 # Filter out malformed formats for better extraction robustness
2490 formats = list(filter(is_wellformed, formats))
2491
2492 formats_dict = {}
2493
2494 # We check that all the formats have the format and format_id fields
2495 for i, format in enumerate(formats):
2496 sanitize_string_field(format, 'format_id')
2497 sanitize_numeric_fields(format)
2498 format['url'] = sanitize_url(format['url'])
2499 if not format.get('format_id'):
2500 format['format_id'] = compat_str(i)
2501 else:
2502 # Sanitize format_id from characters used in format selector expression
2503 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2504 format_id = format['format_id']
2505 if format_id not in formats_dict:
2506 formats_dict[format_id] = []
2507 formats_dict[format_id].append(format)
2508
2509 # Make sure all formats have unique format_id
2510 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2511 for format_id, ambiguous_formats in formats_dict.items():
2512 ambigious_id = len(ambiguous_formats) > 1
2513 for i, format in enumerate(ambiguous_formats):
2514 if ambigious_id:
2515 format['format_id'] = '%s-%d' % (format_id, i)
2516 if format.get('ext') is None:
2517 format['ext'] = determine_ext(format['url']).lower()
2518 # Ensure there is no conflict between id and ext in format selection
2519 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2520 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2521 format['format_id'] = 'f%s' % format['format_id']
2522
2523 for i, format in enumerate(formats):
2524 if format.get('format') is None:
2525 format['format'] = '{id} - {res}{note}'.format(
2526 id=format['format_id'],
2527 res=self.format_resolution(format),
2528 note=format_field(format, 'format_note', ' (%s)'),
2529 )
2530 if format.get('protocol') is None:
2531 format['protocol'] = determine_protocol(format)
2532 if format.get('resolution') is None:
2533 format['resolution'] = self.format_resolution(format, default=None)
2534 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2535 format['dynamic_range'] = 'SDR'
2536 if (info_dict.get('duration') and format.get('tbr')
2537 and not format.get('filesize') and not format.get('filesize_approx')):
2538 format['filesize_approx'] = info_dict['duration'] * format['tbr'] * (1024 / 8)
2539
2540 # Add HTTP headers, so that external programs can use them from the
2541 # json output
2542 full_format_info = info_dict.copy()
2543 full_format_info.update(format)
2544 format['http_headers'] = self._calc_headers(full_format_info)
2545 # Remove private housekeeping stuff
2546 if '__x_forwarded_for_ip' in info_dict:
2547 del info_dict['__x_forwarded_for_ip']
2548
2549 if self.params.get('check_formats') is True:
2550 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2551
2552 if not formats or formats[0] is not info_dict:
2553 # only set the 'formats' fields if the original info_dict list them
2554 # otherwise we end up with a circular reference, the first (and unique)
2555 # element in the 'formats' field in info_dict is info_dict itself,
2556 # which can't be exported to json
2557 info_dict['formats'] = formats
2558
2559 info_dict, _ = self.pre_process(info_dict)
2560
2561 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2562 return info_dict
2563
2564 self.post_extract(info_dict)
2565 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2566
2567 # The pre-processors may have modified the formats
2568 formats = info_dict.get('formats', [info_dict])
2569
2570 list_only = self.params.get('simulate') is None and (
2571 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2572 interactive_format_selection = not list_only and self.format_selector == '-'
2573 if self.params.get('list_thumbnails'):
2574 self.list_thumbnails(info_dict)
2575 if self.params.get('listsubtitles'):
2576 if 'automatic_captions' in info_dict:
2577 self.list_subtitles(
2578 info_dict['id'], automatic_captions, 'automatic captions')
2579 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2580 if self.params.get('listformats') or interactive_format_selection:
2581 self.list_formats(info_dict)
2582 if list_only:
2583 # Without this printing, -F --print-json will not work
2584 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2585 return
2586
2587 format_selector = self.format_selector
2588 if format_selector is None:
2589 req_format = self._default_format_spec(info_dict, download=download)
2590 self.write_debug('Default format spec: %s' % req_format)
2591 format_selector = self.build_format_selector(req_format)
2592
2593 while True:
2594 if interactive_format_selection:
2595 req_format = input(
2596 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2597 try:
2598 format_selector = self.build_format_selector(req_format)
2599 except SyntaxError as err:
2600 self.report_error(err, tb=False, is_error=False)
2601 continue
2602
2603 formats_to_download = list(format_selector({
2604 'formats': formats,
2605 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2606 'incomplete_formats': (
2607 # All formats are video-only or
2608 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2609 # all formats are audio-only
2610 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2611 }))
2612 if interactive_format_selection and not formats_to_download:
2613 self.report_error('Requested format is not available', tb=False, is_error=False)
2614 continue
2615 break
2616
2617 if not formats_to_download:
2618 if not self.params.get('ignore_no_formats_error'):
2619 raise ExtractorError(
2620 'Requested format is not available. Use --list-formats for a list of available formats',
2621 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2622 self.report_warning('Requested format is not available')
2623 # Process what we can, even without any available formats.
2624 formats_to_download = [{}]
2625
2626 best_format = formats_to_download[-1]
2627 if download:
2628 if best_format:
2629 self.to_screen(
2630 f'[info] {info_dict["id"]}: Downloading {len(formats_to_download)} format(s): '
2631 + ', '.join([f['format_id'] for f in formats_to_download]))
2632 max_downloads_reached = False
2633 for i, fmt in enumerate(formats_to_download):
2634 formats_to_download[i] = new_info = self._copy_infodict(info_dict)
2635 new_info.update(fmt)
2636 try:
2637 self.process_info(new_info)
2638 except MaxDownloadsReached:
2639 max_downloads_reached = True
2640 # Remove copied info
2641 for key, val in tuple(new_info.items()):
2642 if info_dict.get(key) == val:
2643 new_info.pop(key)
2644 if max_downloads_reached:
2645 break
2646
2647 write_archive = set(f.get('__write_download_archive', False) for f in formats_to_download)
2648 assert write_archive.issubset({True, False, 'ignore'})
2649 if True in write_archive and False not in write_archive:
2650 self.record_download_archive(info_dict)
2651
2652 info_dict['requested_downloads'] = formats_to_download
2653 info_dict = self.run_all_pps('after_video', info_dict)
2654 if max_downloads_reached:
2655 raise MaxDownloadsReached()
2656
2657 # We update the info dict with the selected best quality format (backwards compatibility)
2658 info_dict.update(best_format)
2659 return info_dict
2660
2661 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2662 """Select the requested subtitles and their format"""
2663 available_subs, normal_sub_langs = {}, []
2664 if normal_subtitles and self.params.get('writesubtitles'):
2665 available_subs.update(normal_subtitles)
2666 normal_sub_langs = tuple(normal_subtitles.keys())
2667 if automatic_captions and self.params.get('writeautomaticsub'):
2668 for lang, cap_info in automatic_captions.items():
2669 if lang not in available_subs:
2670 available_subs[lang] = cap_info
2671
2672 if (not self.params.get('writesubtitles') and not
2673 self.params.get('writeautomaticsub') or not
2674 available_subs):
2675 return None
2676
2677 all_sub_langs = tuple(available_subs.keys())
2678 if self.params.get('allsubtitles', False):
2679 requested_langs = all_sub_langs
2680 elif self.params.get('subtitleslangs', False):
2681 # A list is used so that the order of languages will be the same as
2682 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2683 requested_langs = []
2684 for lang_re in self.params.get('subtitleslangs'):
2685 discard = lang_re[0] == '-'
2686 if discard:
2687 lang_re = lang_re[1:]
2688 if lang_re == 'all':
2689 if discard:
2690 requested_langs = []
2691 else:
2692 requested_langs.extend(all_sub_langs)
2693 continue
2694 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2695 if discard:
2696 for lang in current_langs:
2697 while lang in requested_langs:
2698 requested_langs.remove(lang)
2699 else:
2700 requested_langs.extend(current_langs)
2701 requested_langs = orderedSet(requested_langs)
2702 elif normal_sub_langs:
2703 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2704 else:
2705 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2706 if requested_langs:
2707 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2708
2709 formats_query = self.params.get('subtitlesformat', 'best')
2710 formats_preference = formats_query.split('/') if formats_query else []
2711 subs = {}
2712 for lang in requested_langs:
2713 formats = available_subs.get(lang)
2714 if formats is None:
2715 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
2716 continue
2717 for ext in formats_preference:
2718 if ext == 'best':
2719 f = formats[-1]
2720 break
2721 matches = list(filter(lambda f: f['ext'] == ext, formats))
2722 if matches:
2723 f = matches[-1]
2724 break
2725 else:
2726 f = formats[-1]
2727 self.report_warning(
2728 'No subtitle format found matching "%s" for language %s, '
2729 'using %s' % (formats_query, lang, f['ext']))
2730 subs[lang] = f
2731 return subs
2732
2733 def _forceprint(self, key, info_dict):
2734 if info_dict is None:
2735 return
2736 info_copy = info_dict.copy()
2737 info_copy['formats_table'] = self.render_formats_table(info_dict)
2738 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2739 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2740 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2741
2742 def format_tmpl(tmpl):
2743 mobj = re.match(r'\w+(=?)$', tmpl)
2744 if mobj and mobj.group(1):
2745 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2746 elif mobj:
2747 return f'%({tmpl})s'
2748 return tmpl
2749
2750 for tmpl in self.params['forceprint'].get(key, []):
2751 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2752
2753 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2754 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2755 tmpl = format_tmpl(tmpl)
2756 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2757 if self._ensure_dir_exists(filename):
2758 with io.open(filename, 'a', encoding='utf-8') as f:
2759 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2760
2761 def __forced_printings(self, info_dict, filename, incomplete):
2762 def print_mandatory(field, actual_field=None):
2763 if actual_field is None:
2764 actual_field = field
2765 if (self.params.get('force%s' % field, False)
2766 and (not incomplete or info_dict.get(actual_field) is not None)):
2767 self.to_stdout(info_dict[actual_field])
2768
2769 def print_optional(field):
2770 if (self.params.get('force%s' % field, False)
2771 and info_dict.get(field) is not None):
2772 self.to_stdout(info_dict[field])
2773
2774 info_dict = info_dict.copy()
2775 if filename is not None:
2776 info_dict['filename'] = filename
2777 if info_dict.get('requested_formats') is not None:
2778 # For RTMP URLs, also include the playpath
2779 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2780 elif info_dict.get('url'):
2781 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2782
2783 if (self.params.get('forcejson')
2784 or self.params['forceprint'].get('video')
2785 or self.params['print_to_file'].get('video')):
2786 self.post_extract(info_dict)
2787 self._forceprint('video', info_dict)
2788
2789 print_mandatory('title')
2790 print_mandatory('id')
2791 print_mandatory('url', 'urls')
2792 print_optional('thumbnail')
2793 print_optional('description')
2794 print_optional('filename')
2795 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2796 self.to_stdout(formatSeconds(info_dict['duration']))
2797 print_mandatory('format')
2798
2799 if self.params.get('forcejson'):
2800 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2801
2802 def dl(self, name, info, subtitle=False, test=False):
2803 if not info.get('url'):
2804 self.raise_no_formats(info, True)
2805
2806 if test:
2807 verbose = self.params.get('verbose')
2808 params = {
2809 'test': True,
2810 'quiet': self.params.get('quiet') or not verbose,
2811 'verbose': verbose,
2812 'noprogress': not verbose,
2813 'nopart': True,
2814 'skip_unavailable_fragments': False,
2815 'keep_fragments': False,
2816 'overwrites': True,
2817 '_no_ytdl_file': True,
2818 }
2819 else:
2820 params = self.params
2821 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2822 if not test:
2823 for ph in self._progress_hooks:
2824 fd.add_progress_hook(ph)
2825 urls = '", "'.join(
2826 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2827 for f in info.get('requested_formats', []) or [info])
2828 self.write_debug('Invoking downloader on "%s"' % urls)
2829
2830 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2831 # But it may contain objects that are not deep-copyable
2832 new_info = self._copy_infodict(info)
2833 if new_info.get('http_headers') is None:
2834 new_info['http_headers'] = self._calc_headers(new_info)
2835 return fd.download(name, new_info, subtitle)
2836
2837 def existing_file(self, filepaths, *, default_overwrite=True):
2838 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2839 if existing_files and not self.params.get('overwrites', default_overwrite):
2840 return existing_files[0]
2841
2842 for file in existing_files:
2843 self.report_file_delete(file)
2844 os.remove(file)
2845 return None
2846
2847 def process_info(self, info_dict):
2848 """Process a single resolved IE result. (Modifies it in-place)"""
2849
2850 assert info_dict.get('_type', 'video') == 'video'
2851 original_infodict = info_dict
2852
2853 if 'format' not in info_dict and 'ext' in info_dict:
2854 info_dict['format'] = info_dict['ext']
2855
2856 # This is mostly just for backward compatibility of process_info
2857 # As a side-effect, this allows for format-specific filters
2858 if self._match_entry(info_dict) is not None:
2859 info_dict['__write_download_archive'] = 'ignore'
2860 return
2861
2862 # Does nothing under normal operation - for backward compatibility of process_info
2863 self.post_extract(info_dict)
2864 self._num_downloads += 1
2865
2866 # info_dict['_filename'] needs to be set for backward compatibility
2867 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2868 temp_filename = self.prepare_filename(info_dict, 'temp')
2869 files_to_move = {}
2870
2871 # Forced printings
2872 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2873
2874 if self.params.get('simulate'):
2875 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2876 return
2877
2878 if full_filename is None:
2879 return
2880 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2881 return
2882 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2883 return
2884
2885 if self._write_description('video', info_dict,
2886 self.prepare_filename(info_dict, 'description')) is None:
2887 return
2888
2889 sub_files = self._write_subtitles(info_dict, temp_filename)
2890 if sub_files is None:
2891 return
2892 files_to_move.update(dict(sub_files))
2893
2894 thumb_files = self._write_thumbnails(
2895 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2896 if thumb_files is None:
2897 return
2898 files_to_move.update(dict(thumb_files))
2899
2900 infofn = self.prepare_filename(info_dict, 'infojson')
2901 _infojson_written = self._write_info_json('video', info_dict, infofn)
2902 if _infojson_written:
2903 info_dict['infojson_filename'] = infofn
2904 # For backward compatibility, even though it was a private field
2905 info_dict['__infojson_filename'] = infofn
2906 elif _infojson_written is None:
2907 return
2908
2909 # Note: Annotations are deprecated
2910 annofn = None
2911 if self.params.get('writeannotations', False):
2912 annofn = self.prepare_filename(info_dict, 'annotation')
2913 if annofn:
2914 if not self._ensure_dir_exists(encodeFilename(annofn)):
2915 return
2916 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2917 self.to_screen('[info] Video annotations are already present')
2918 elif not info_dict.get('annotations'):
2919 self.report_warning('There are no annotations to write.')
2920 else:
2921 try:
2922 self.to_screen('[info] Writing video annotations to: ' + annofn)
2923 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2924 annofile.write(info_dict['annotations'])
2925 except (KeyError, TypeError):
2926 self.report_warning('There are no annotations to write.')
2927 except (OSError, IOError):
2928 self.report_error('Cannot write annotations file: ' + annofn)
2929 return
2930
2931 # Write internet shortcut files
2932 def _write_link_file(link_type):
2933 url = try_get(info_dict['webpage_url'], iri_to_uri)
2934 if not url:
2935 self.report_warning(
2936 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2937 return True
2938 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2939 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2940 return False
2941 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2942 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2943 return True
2944 try:
2945 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2946 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2947 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2948 template_vars = {'url': url}
2949 if link_type == 'desktop':
2950 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2951 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2952 except (OSError, IOError):
2953 self.report_error(f'Cannot write internet shortcut {linkfn}')
2954 return False
2955 return True
2956
2957 write_links = {
2958 'url': self.params.get('writeurllink'),
2959 'webloc': self.params.get('writewebloclink'),
2960 'desktop': self.params.get('writedesktoplink'),
2961 }
2962 if self.params.get('writelink'):
2963 link_type = ('webloc' if sys.platform == 'darwin'
2964 else 'desktop' if sys.platform.startswith('linux')
2965 else 'url')
2966 write_links[link_type] = True
2967
2968 if any(should_write and not _write_link_file(link_type)
2969 for link_type, should_write in write_links.items()):
2970 return
2971
2972 def replace_info_dict(new_info):
2973 nonlocal info_dict
2974 if new_info == info_dict:
2975 return
2976 info_dict.clear()
2977 info_dict.update(new_info)
2978
2979 try:
2980 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2981 replace_info_dict(new_info)
2982 except PostProcessingError as err:
2983 self.report_error('Preprocessing: %s' % str(err))
2984 return
2985
2986 if self.params.get('skip_download'):
2987 info_dict['filepath'] = temp_filename
2988 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2989 info_dict['__files_to_move'] = files_to_move
2990 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2991 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2992 else:
2993 # Download
2994 info_dict.setdefault('__postprocessors', [])
2995 try:
2996
2997 def existing_video_file(*filepaths):
2998 ext = info_dict.get('ext')
2999 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3000 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3001 default_overwrite=False)
3002 if file:
3003 info_dict['ext'] = os.path.splitext(file)[1][1:]
3004 return file
3005
3006 success = True
3007 if info_dict.get('requested_formats') is not None:
3008
3009 def compatible_formats(formats):
3010 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3011 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3012 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3013 if len(video_formats) > 2 or len(audio_formats) > 2:
3014 return False
3015
3016 # Check extension
3017 exts = set(format.get('ext') for format in formats)
3018 COMPATIBLE_EXTS = (
3019 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
3020 set(('webm',)),
3021 )
3022 for ext_sets in COMPATIBLE_EXTS:
3023 if ext_sets.issuperset(exts):
3024 return True
3025 # TODO: Check acodec/vcodec
3026 return False
3027
3028 requested_formats = info_dict['requested_formats']
3029 old_ext = info_dict['ext']
3030 if self.params.get('merge_output_format') is None:
3031 if not compatible_formats(requested_formats):
3032 info_dict['ext'] = 'mkv'
3033 self.report_warning(
3034 'Requested formats are incompatible for merge and will be merged into mkv')
3035 if (info_dict['ext'] == 'webm'
3036 and info_dict.get('thumbnails')
3037 # check with type instead of pp_key, __name__, or isinstance
3038 # since we dont want any custom PPs to trigger this
3039 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])):
3040 info_dict['ext'] = 'mkv'
3041 self.report_warning(
3042 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3043 new_ext = info_dict['ext']
3044
3045 def correct_ext(filename, ext=new_ext):
3046 if filename == '-':
3047 return filename
3048 filename_real_ext = os.path.splitext(filename)[1][1:]
3049 filename_wo_ext = (
3050 os.path.splitext(filename)[0]
3051 if filename_real_ext in (old_ext, new_ext)
3052 else filename)
3053 return '%s.%s' % (filename_wo_ext, ext)
3054
3055 # Ensure filename always has a correct extension for successful merge
3056 full_filename = correct_ext(full_filename)
3057 temp_filename = correct_ext(temp_filename)
3058 dl_filename = existing_video_file(full_filename, temp_filename)
3059 info_dict['__real_download'] = False
3060
3061 downloaded = []
3062 merger = FFmpegMergerPP(self)
3063
3064 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3065 if dl_filename is not None:
3066 self.report_file_already_downloaded(dl_filename)
3067 elif fd:
3068 for f in requested_formats if fd != FFmpegFD else []:
3069 f['filepath'] = fname = prepend_extension(
3070 correct_ext(temp_filename, info_dict['ext']),
3071 'f%s' % f['format_id'], info_dict['ext'])
3072 downloaded.append(fname)
3073 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3074 success, real_download = self.dl(temp_filename, info_dict)
3075 info_dict['__real_download'] = real_download
3076 else:
3077 if self.params.get('allow_unplayable_formats'):
3078 self.report_warning(
3079 'You have requested merging of multiple formats '
3080 'while also allowing unplayable formats to be downloaded. '
3081 'The formats won\'t be merged to prevent data corruption.')
3082 elif not merger.available:
3083 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3084 if not self.params.get('ignoreerrors'):
3085 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3086 return
3087 self.report_warning(f'{msg}. The formats won\'t be merged')
3088
3089 if temp_filename == '-':
3090 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3091 else 'but the formats are incompatible for simultaneous download' if merger.available
3092 else 'but ffmpeg is not installed')
3093 self.report_warning(
3094 f'You have requested downloading multiple formats to stdout {reason}. '
3095 'The formats will be streamed one after the other')
3096 fname = temp_filename
3097 for f in requested_formats:
3098 new_info = dict(info_dict)
3099 del new_info['requested_formats']
3100 new_info.update(f)
3101 if temp_filename != '-':
3102 fname = prepend_extension(
3103 correct_ext(temp_filename, new_info['ext']),
3104 'f%s' % f['format_id'], new_info['ext'])
3105 if not self._ensure_dir_exists(fname):
3106 return
3107 f['filepath'] = fname
3108 downloaded.append(fname)
3109 partial_success, real_download = self.dl(fname, new_info)
3110 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3111 success = success and partial_success
3112
3113 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3114 info_dict['__postprocessors'].append(merger)
3115 info_dict['__files_to_merge'] = downloaded
3116 # Even if there were no downloads, it is being merged only now
3117 info_dict['__real_download'] = True
3118 else:
3119 for file in downloaded:
3120 files_to_move[file] = None
3121 else:
3122 # Just a single file
3123 dl_filename = existing_video_file(full_filename, temp_filename)
3124 if dl_filename is None or dl_filename == temp_filename:
3125 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3126 # So we should try to resume the download
3127 success, real_download = self.dl(temp_filename, info_dict)
3128 info_dict['__real_download'] = real_download
3129 else:
3130 self.report_file_already_downloaded(dl_filename)
3131
3132 dl_filename = dl_filename or temp_filename
3133 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3134
3135 except network_exceptions as err:
3136 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3137 return
3138 except (OSError, IOError) as err:
3139 raise UnavailableVideoError(err)
3140 except (ContentTooShortError, ) as err:
3141 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
3142 return
3143
3144 if success and full_filename != '-':
3145
3146 def fixup():
3147 do_fixup = True
3148 fixup_policy = self.params.get('fixup')
3149 vid = info_dict['id']
3150
3151 if fixup_policy in ('ignore', 'never'):
3152 return
3153 elif fixup_policy == 'warn':
3154 do_fixup = False
3155 elif fixup_policy != 'force':
3156 assert fixup_policy in ('detect_or_warn', None)
3157 if not info_dict.get('__real_download'):
3158 do_fixup = False
3159
3160 def ffmpeg_fixup(cndn, msg, cls):
3161 if not cndn:
3162 return
3163 if not do_fixup:
3164 self.report_warning(f'{vid}: {msg}')
3165 return
3166 pp = cls(self)
3167 if pp.available:
3168 info_dict['__postprocessors'].append(pp)
3169 else:
3170 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3171
3172 stretched_ratio = info_dict.get('stretched_ratio')
3173 ffmpeg_fixup(
3174 stretched_ratio not in (1, None),
3175 f'Non-uniform pixel ratio {stretched_ratio}',
3176 FFmpegFixupStretchedPP)
3177
3178 ffmpeg_fixup(
3179 (info_dict.get('requested_formats') is None
3180 and info_dict.get('container') == 'm4a_dash'
3181 and info_dict.get('ext') == 'm4a'),
3182 'writing DASH m4a. Only some players support this container',
3183 FFmpegFixupM4aPP)
3184
3185 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3186 downloader = downloader.__name__ if downloader else None
3187
3188 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3189 ffmpeg_fixup(downloader == 'HlsFD',
3190 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3191 FFmpegFixupM3u8PP)
3192 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3193 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3194
3195 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3196 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed duration detected', FFmpegFixupDurationPP)
3197
3198 fixup()
3199 try:
3200 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3201 except PostProcessingError as err:
3202 self.report_error('Postprocessing: %s' % str(err))
3203 return
3204 try:
3205 for ph in self._post_hooks:
3206 ph(info_dict['filepath'])
3207 except Exception as err:
3208 self.report_error('post hooks: %s' % str(err))
3209 return
3210 info_dict['__write_download_archive'] = True
3211
3212 if self.params.get('force_write_download_archive'):
3213 info_dict['__write_download_archive'] = True
3214
3215 # Make sure the info_dict was modified in-place
3216 assert info_dict is original_infodict
3217
3218 max_downloads = self.params.get('max_downloads')
3219 if max_downloads is not None and self._num_downloads >= int(max_downloads):
3220 raise MaxDownloadsReached()
3221
3222 def __download_wrapper(self, func):
3223 @functools.wraps(func)
3224 def wrapper(*args, **kwargs):
3225 try:
3226 res = func(*args, **kwargs)
3227 except UnavailableVideoError as e:
3228 self.report_error(e)
3229 except MaxDownloadsReached as e:
3230 self.to_screen(f'[info] {e}')
3231 raise
3232 except DownloadCancelled as e:
3233 self.to_screen(f'[info] {e}')
3234 if not self.params.get('break_per_url'):
3235 raise
3236 else:
3237 if self.params.get('dump_single_json', False):
3238 self.post_extract(res)
3239 self.to_stdout(json.dumps(self.sanitize_info(res)))
3240 return wrapper
3241
3242 def download(self, url_list):
3243 """Download a given list of URLs."""
3244 url_list = variadic(url_list) # Passing a single URL is a common mistake
3245 outtmpl = self.outtmpl_dict['default']
3246 if (len(url_list) > 1
3247 and outtmpl != '-'
3248 and '%' not in outtmpl
3249 and self.params.get('max_downloads') != 1):
3250 raise SameFileError(outtmpl)
3251
3252 for url in url_list:
3253 self.__download_wrapper(self.extract_info)(
3254 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3255
3256 return self._download_retcode
3257
3258 def download_with_info_file(self, info_filename):
3259 with contextlib.closing(fileinput.FileInput(
3260 [info_filename], mode='r',
3261 openhook=fileinput.hook_encoded('utf-8'))) as f:
3262 # FileInput doesn't have a read method, we can't call json.load
3263 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3264 try:
3265 self.__download_wrapper(self.process_ie_result)(info, download=True)
3266 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3267 if not isinstance(e, EntryNotInPlaylist):
3268 self.to_stderr('\r')
3269 webpage_url = info.get('webpage_url')
3270 if webpage_url is not None:
3271 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3272 return self.download([webpage_url])
3273 else:
3274 raise
3275 return self._download_retcode
3276
3277 @staticmethod
3278 def sanitize_info(info_dict, remove_private_keys=False):
3279 ''' Sanitize the infodict for converting to json '''
3280 if info_dict is None:
3281 return info_dict
3282 info_dict.setdefault('epoch', int(time.time()))
3283 info_dict.setdefault('_type', 'video')
3284
3285 if remove_private_keys:
3286 reject = lambda k, v: v is None or (k.startswith('_') and k != '_type') or k in {
3287 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3288 'entries', 'filepath', 'infojson_filename', 'original_url', 'playlist_autonumber',
3289 }
3290 else:
3291 reject = lambda k, v: False
3292
3293 def filter_fn(obj):
3294 if isinstance(obj, dict):
3295 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3296 elif isinstance(obj, (list, tuple, set, LazyList)):
3297 return list(map(filter_fn, obj))
3298 elif obj is None or isinstance(obj, (str, int, float, bool)):
3299 return obj
3300 else:
3301 return repr(obj)
3302
3303 return filter_fn(info_dict)
3304
3305 @staticmethod
3306 def filter_requested_info(info_dict, actually_filter=True):
3307 ''' Alias of sanitize_info for backward compatibility '''
3308 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3309
3310 @staticmethod
3311 def post_extract(info_dict):
3312 def actual_post_extract(info_dict):
3313 if info_dict.get('_type') in ('playlist', 'multi_video'):
3314 for video_dict in info_dict.get('entries', {}):
3315 actual_post_extract(video_dict or {})
3316 return
3317
3318 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3319 info_dict.update(post_extractor())
3320
3321 actual_post_extract(info_dict or {})
3322
3323 def run_pp(self, pp, infodict):
3324 files_to_delete = []
3325 if '__files_to_move' not in infodict:
3326 infodict['__files_to_move'] = {}
3327 try:
3328 files_to_delete, infodict = pp.run(infodict)
3329 except PostProcessingError as e:
3330 # Must be True and not 'only_download'
3331 if self.params.get('ignoreerrors') is True:
3332 self.report_error(e)
3333 return infodict
3334 raise
3335
3336 if not files_to_delete:
3337 return infodict
3338 if self.params.get('keepvideo', False):
3339 for f in files_to_delete:
3340 infodict['__files_to_move'].setdefault(f, '')
3341 else:
3342 for old_filename in set(files_to_delete):
3343 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
3344 try:
3345 os.remove(encodeFilename(old_filename))
3346 except (IOError, OSError):
3347 self.report_warning('Unable to remove downloaded original file')
3348 if old_filename in infodict['__files_to_move']:
3349 del infodict['__files_to_move'][old_filename]
3350 return infodict
3351
3352 def run_all_pps(self, key, info, *, additional_pps=None):
3353 self._forceprint(key, info)
3354 for pp in (additional_pps or []) + self._pps[key]:
3355 info = self.run_pp(pp, info)
3356 return info
3357
3358 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3359 info = dict(ie_info)
3360 info['__files_to_move'] = files_to_move or {}
3361 info = self.run_all_pps(key, info)
3362 return info, info.pop('__files_to_move', None)
3363
3364 def post_process(self, filename, info, files_to_move=None):
3365 """Run all the postprocessors on the given file."""
3366 info['filepath'] = filename
3367 info['__files_to_move'] = files_to_move or {}
3368 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3369 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3370 del info['__files_to_move']
3371 return self.run_all_pps('after_move', info)
3372
3373 def _make_archive_id(self, info_dict):
3374 video_id = info_dict.get('id')
3375 if not video_id:
3376 return
3377 # Future-proof against any change in case
3378 # and backwards compatibility with prior versions
3379 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3380 if extractor is None:
3381 url = str_or_none(info_dict.get('url'))
3382 if not url:
3383 return
3384 # Try to find matching extractor for the URL and take its ie_key
3385 for ie_key, ie in self._ies.items():
3386 if ie.suitable(url):
3387 extractor = ie_key
3388 break
3389 else:
3390 return
3391 return '%s %s' % (extractor.lower(), video_id)
3392
3393 def in_download_archive(self, info_dict):
3394 fn = self.params.get('download_archive')
3395 if fn is None:
3396 return False
3397
3398 vid_id = self._make_archive_id(info_dict)
3399 if not vid_id:
3400 return False # Incomplete video information
3401
3402 return vid_id in self.archive
3403
3404 def record_download_archive(self, info_dict):
3405 fn = self.params.get('download_archive')
3406 if fn is None:
3407 return
3408 vid_id = self._make_archive_id(info_dict)
3409 assert vid_id
3410 self.write_debug(f'Adding to archive: {vid_id}')
3411 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3412 archive_file.write(vid_id + '\n')
3413 self.archive.add(vid_id)
3414
3415 @staticmethod
3416 def format_resolution(format, default='unknown'):
3417 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3418 return 'audio only'
3419 if format.get('resolution') is not None:
3420 return format['resolution']
3421 if format.get('width') and format.get('height'):
3422 return '%dx%d' % (format['width'], format['height'])
3423 elif format.get('height'):
3424 return '%sp' % format['height']
3425 elif format.get('width'):
3426 return '%dx?' % format['width']
3427 return default
3428
3429 def _list_format_headers(self, *headers):
3430 if self.params.get('listformats_table', True) is not False:
3431 return [self._format_screen(header, self.Styles.HEADERS) for header in headers]
3432 return headers
3433
3434 def _format_note(self, fdict):
3435 res = ''
3436 if fdict.get('ext') in ['f4f', 'f4m']:
3437 res += '(unsupported)'
3438 if fdict.get('language'):
3439 if res:
3440 res += ' '
3441 res += '[%s]' % fdict['language']
3442 if fdict.get('format_note') is not None:
3443 if res:
3444 res += ' '
3445 res += fdict['format_note']
3446 if fdict.get('tbr') is not None:
3447 if res:
3448 res += ', '
3449 res += '%4dk' % fdict['tbr']
3450 if fdict.get('container') is not None:
3451 if res:
3452 res += ', '
3453 res += '%s container' % fdict['container']
3454 if (fdict.get('vcodec') is not None
3455 and fdict.get('vcodec') != 'none'):
3456 if res:
3457 res += ', '
3458 res += fdict['vcodec']
3459 if fdict.get('vbr') is not None:
3460 res += '@'
3461 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3462 res += 'video@'
3463 if fdict.get('vbr') is not None:
3464 res += '%4dk' % fdict['vbr']
3465 if fdict.get('fps') is not None:
3466 if res:
3467 res += ', '
3468 res += '%sfps' % fdict['fps']
3469 if fdict.get('acodec') is not None:
3470 if res:
3471 res += ', '
3472 if fdict['acodec'] == 'none':
3473 res += 'video only'
3474 else:
3475 res += '%-5s' % fdict['acodec']
3476 elif fdict.get('abr') is not None:
3477 if res:
3478 res += ', '
3479 res += 'audio'
3480 if fdict.get('abr') is not None:
3481 res += '@%3dk' % fdict['abr']
3482 if fdict.get('asr') is not None:
3483 res += ' (%5dHz)' % fdict['asr']
3484 if fdict.get('filesize') is not None:
3485 if res:
3486 res += ', '
3487 res += format_bytes(fdict['filesize'])
3488 elif fdict.get('filesize_approx') is not None:
3489 if res:
3490 res += ', '
3491 res += '~' + format_bytes(fdict['filesize_approx'])
3492 return res
3493
3494 def render_formats_table(self, info_dict):
3495 if not info_dict.get('formats') and not info_dict.get('url'):
3496 return None
3497
3498 formats = info_dict.get('formats', [info_dict])
3499 if not self.params.get('listformats_table', True) is not False:
3500 table = [
3501 [
3502 format_field(f, 'format_id'),
3503 format_field(f, 'ext'),
3504 self.format_resolution(f),
3505 self._format_note(f)
3506 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3507 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3508
3509 delim = self._format_screen('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3510 table = [
3511 [
3512 self._format_screen(format_field(f, 'format_id'), self.Styles.ID),
3513 format_field(f, 'ext'),
3514 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3515 format_field(f, 'fps', '\t%d'),
3516 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3517 delim,
3518 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3519 format_field(f, 'tbr', '\t%dk'),
3520 shorten_protocol_name(f.get('protocol', '')),
3521 delim,
3522 format_field(f, 'vcodec', default='unknown').replace(
3523 'none', 'images' if f.get('acodec') == 'none'
3524 else self._format_screen('audio only', self.Styles.SUPPRESS)),
3525 format_field(f, 'vbr', '\t%dk'),
3526 format_field(f, 'acodec', default='unknown').replace(
3527 'none', '' if f.get('vcodec') == 'none'
3528 else self._format_screen('video only', self.Styles.SUPPRESS)),
3529 format_field(f, 'abr', '\t%dk'),
3530 format_field(f, 'asr', '\t%dHz'),
3531 join_nonempty(
3532 self._format_screen('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3533 format_field(f, 'language', '[%s]'),
3534 join_nonempty(format_field(f, 'format_note'),
3535 format_field(f, 'container', ignore=(None, f.get('ext'))),
3536 delim=', '),
3537 delim=' '),
3538 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3539 header_line = self._list_format_headers(
3540 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3541 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3542
3543 return render_table(
3544 header_line, table, hide_empty=True,
3545 delim=self._format_screen('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3546
3547 def render_thumbnails_table(self, info_dict):
3548 thumbnails = list(info_dict.get('thumbnails') or [])
3549 if not thumbnails:
3550 return None
3551 return render_table(
3552 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3553 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3554
3555 def render_subtitles_table(self, video_id, subtitles):
3556 def _row(lang, formats):
3557 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3558 if len(set(names)) == 1:
3559 names = [] if names[0] == 'unknown' else names[:1]
3560 return [lang, ', '.join(names), ', '.join(exts)]
3561
3562 if not subtitles:
3563 return None
3564 return render_table(
3565 self._list_format_headers('Language', 'Name', 'Formats'),
3566 [_row(lang, formats) for lang, formats in subtitles.items()],
3567 hide_empty=True)
3568
3569 def __list_table(self, video_id, name, func, *args):
3570 table = func(*args)
3571 if not table:
3572 self.to_screen(f'{video_id} has no {name}')
3573 return
3574 self.to_screen(f'[info] Available {name} for {video_id}:')
3575 self.to_stdout(table)
3576
3577 def list_formats(self, info_dict):
3578 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3579
3580 def list_thumbnails(self, info_dict):
3581 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3582
3583 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3584 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3585
3586 def urlopen(self, req):
3587 """ Start an HTTP download """
3588 if isinstance(req, str):
3589 req = sanitized_Request(req)
3590 return self._opener.open(req, timeout=self._socket_timeout)
3591
3592 def print_debug_header(self):
3593 if not self.params.get('verbose'):
3594 return
3595
3596 def get_encoding(stream):
3597 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3598 if not supports_terminal_sequences(stream):
3599 from .compat import WINDOWS_VT_MODE
3600 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3601 return ret
3602
3603 encoding_str = 'Encodings: locale %s, fs %s, out %s, err %s, pref %s' % (
3604 locale.getpreferredencoding(),
3605 sys.getfilesystemencoding(),
3606 get_encoding(self._out_files['screen']), get_encoding(self._out_files['error']),
3607 self.get_encoding())
3608
3609 logger = self.params.get('logger')
3610 if logger:
3611 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3612 write_debug(encoding_str)
3613 else:
3614 write_string(f'[debug] {encoding_str}\n', encoding=None)
3615 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3616
3617 source = detect_variant()
3618 write_debug(join_nonempty(
3619 'yt-dlp version', __version__,
3620 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3621 '' if source == 'unknown' else f'({source})',
3622 delim=' '))
3623 if not _LAZY_LOADER:
3624 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3625 write_debug('Lazy loading extractors is forcibly disabled')
3626 else:
3627 write_debug('Lazy loading extractors is disabled')
3628 if plugin_extractors or plugin_postprocessors:
3629 write_debug('Plugins: %s' % [
3630 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3631 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3632 if self.params.get('compat_opts'):
3633 write_debug('Compatibility options: %s' % ', '.join(self.params.get('compat_opts')))
3634
3635 if source == 'source':
3636 try:
3637 sp = Popen(
3638 ['git', 'rev-parse', '--short', 'HEAD'],
3639 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3640 cwd=os.path.dirname(os.path.abspath(__file__)))
3641 out, err = sp.communicate_or_kill()
3642 out = out.decode().strip()
3643 if re.match('[0-9a-f]+', out):
3644 write_debug('Git HEAD: %s' % out)
3645 except Exception:
3646 try:
3647 sys.exc_clear()
3648 except Exception:
3649 pass
3650
3651 def python_implementation():
3652 impl_name = platform.python_implementation()
3653 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3654 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3655 return impl_name
3656
3657 write_debug('Python version %s (%s %s) - %s' % (
3658 platform.python_version(),
3659 python_implementation(),
3660 platform.architecture()[0],
3661 platform_name()))
3662
3663 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3664 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3665 if ffmpeg_features:
3666 exe_versions['ffmpeg'] += ' (%s)' % ','.join(ffmpeg_features)
3667
3668 exe_versions['rtmpdump'] = rtmpdump_version()
3669 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3670 exe_str = ', '.join(
3671 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3672 ) or 'none'
3673 write_debug('exe versions: %s' % exe_str)
3674
3675 from .downloader.websocket import has_websockets
3676 from .postprocessor.embedthumbnail import has_mutagen
3677 from .cookies import SQLITE_AVAILABLE, SECRETSTORAGE_AVAILABLE
3678
3679 lib_str = join_nonempty(
3680 compat_brotli and compat_brotli.__name__,
3681 has_certifi and 'certifi',
3682 compat_pycrypto_AES and compat_pycrypto_AES.__name__.split('.')[0],
3683 SECRETSTORAGE_AVAILABLE and 'secretstorage',
3684 has_mutagen and 'mutagen',
3685 SQLITE_AVAILABLE and 'sqlite',
3686 has_websockets and 'websockets',
3687 delim=', ') or 'none'
3688 write_debug('Optional libraries: %s' % lib_str)
3689
3690 self._setup_opener()
3691 proxy_map = {}
3692 for handler in self._opener.handlers:
3693 if hasattr(handler, 'proxies'):
3694 proxy_map.update(handler.proxies)
3695 write_debug(f'Proxy map: {proxy_map}')
3696
3697 # Not implemented
3698 if False and self.params.get('call_home'):
3699 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3700 write_debug('Public IP address: %s' % ipaddr)
3701 latest_version = self.urlopen(
3702 'https://yt-dl.org/latest/version').read().decode('utf-8')
3703 if version_tuple(latest_version) > version_tuple(__version__):
3704 self.report_warning(
3705 'You are using an outdated version (newest version: %s)! '
3706 'See https://yt-dl.org/update if you need help updating.' %
3707 latest_version)
3708
3709 def _setup_opener(self):
3710 if hasattr(self, '_opener'):
3711 return
3712 timeout_val = self.params.get('socket_timeout')
3713 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3714
3715 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3716 opts_cookiefile = self.params.get('cookiefile')
3717 opts_proxy = self.params.get('proxy')
3718
3719 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3720
3721 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3722 if opts_proxy is not None:
3723 if opts_proxy == '':
3724 proxies = {}
3725 else:
3726 proxies = {'http': opts_proxy, 'https': opts_proxy}
3727 else:
3728 proxies = compat_urllib_request.getproxies()
3729 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3730 if 'http' in proxies and 'https' not in proxies:
3731 proxies['https'] = proxies['http']
3732 proxy_handler = PerRequestProxyHandler(proxies)
3733
3734 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3735 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3736 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3737 redirect_handler = YoutubeDLRedirectHandler()
3738 data_handler = urllib.request.DataHandler()
3739
3740 # When passing our own FileHandler instance, build_opener won't add the
3741 # default FileHandler and allows us to disable the file protocol, which
3742 # can be used for malicious purposes (see
3743 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3744 file_handler = compat_urllib_request.FileHandler()
3745
3746 def file_open(*args, **kwargs):
3747 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3748 file_handler.file_open = file_open
3749
3750 opener = compat_urllib_request.build_opener(
3751 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3752
3753 # Delete the default user-agent header, which would otherwise apply in
3754 # cases where our custom HTTP handler doesn't come into play
3755 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3756 opener.addheaders = []
3757 self._opener = opener
3758
3759 def encode(self, s):
3760 if isinstance(s, bytes):
3761 return s # Already encoded
3762
3763 try:
3764 return s.encode(self.get_encoding())
3765 except UnicodeEncodeError as err:
3766 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3767 raise
3768
3769 def get_encoding(self):
3770 encoding = self.params.get('encoding')
3771 if encoding is None:
3772 encoding = preferredencoding()
3773 return encoding
3774
3775 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3776 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3777 if overwrite is None:
3778 overwrite = self.params.get('overwrites', True)
3779 if not self.params.get('writeinfojson'):
3780 return False
3781 elif not infofn:
3782 self.write_debug(f'Skipping writing {label} infojson')
3783 return False
3784 elif not self._ensure_dir_exists(infofn):
3785 return None
3786 elif not overwrite and os.path.exists(infofn):
3787 self.to_screen(f'[info] {label.title()} metadata is already present')
3788 return 'exists'
3789
3790 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3791 try:
3792 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3793 return True
3794 except (OSError, IOError):
3795 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3796 return None
3797
3798 def _write_description(self, label, ie_result, descfn):
3799 ''' Write description and returns True = written, False = skip, None = error '''
3800 if not self.params.get('writedescription'):
3801 return False
3802 elif not descfn:
3803 self.write_debug(f'Skipping writing {label} description')
3804 return False
3805 elif not self._ensure_dir_exists(descfn):
3806 return None
3807 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3808 self.to_screen(f'[info] {label.title()} description is already present')
3809 elif ie_result.get('description') is None:
3810 self.report_warning(f'There\'s no {label} description to write')
3811 return False
3812 else:
3813 try:
3814 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3815 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3816 descfile.write(ie_result['description'])
3817 except (OSError, IOError):
3818 self.report_error(f'Cannot write {label} description file {descfn}')
3819 return None
3820 return True
3821
3822 def _write_subtitles(self, info_dict, filename):
3823 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3824 ret = []
3825 subtitles = info_dict.get('requested_subtitles')
3826 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3827 # subtitles download errors are already managed as troubles in relevant IE
3828 # that way it will silently go on when used with unsupporting IE
3829 return ret
3830
3831 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3832 if not sub_filename_base:
3833 self.to_screen('[info] Skipping writing video subtitles')
3834 return ret
3835 for sub_lang, sub_info in subtitles.items():
3836 sub_format = sub_info['ext']
3837 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3838 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3839 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3840 if existing_sub:
3841 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3842 sub_info['filepath'] = existing_sub
3843 ret.append((existing_sub, sub_filename_final))
3844 continue
3845
3846 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3847 if sub_info.get('data') is not None:
3848 try:
3849 # Use newline='' to prevent conversion of newline characters
3850 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3851 with io.open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3852 subfile.write(sub_info['data'])
3853 sub_info['filepath'] = sub_filename
3854 ret.append((sub_filename, sub_filename_final))
3855 continue
3856 except (OSError, IOError):
3857 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3858 return None
3859
3860 try:
3861 sub_copy = sub_info.copy()
3862 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3863 self.dl(sub_filename, sub_copy, subtitle=True)
3864 sub_info['filepath'] = sub_filename
3865 ret.append((sub_filename, sub_filename_final))
3866 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3867 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3868 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3869 if not self.params.get('ignoreerrors'):
3870 self.report_error(msg)
3871 raise DownloadError(msg)
3872 self.report_warning(msg)
3873 return ret
3874
3875 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3876 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3877 write_all = self.params.get('write_all_thumbnails', False)
3878 thumbnails, ret = [], []
3879 if write_all or self.params.get('writethumbnail', False):
3880 thumbnails = info_dict.get('thumbnails') or []
3881 multiple = write_all and len(thumbnails) > 1
3882
3883 if thumb_filename_base is None:
3884 thumb_filename_base = filename
3885 if thumbnails and not thumb_filename_base:
3886 self.write_debug(f'Skipping writing {label} thumbnail')
3887 return ret
3888
3889 for idx, t in list(enumerate(thumbnails))[::-1]:
3890 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3891 thumb_display_id = f'{label} thumbnail {t["id"]}'
3892 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3893 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3894
3895 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3896 if existing_thumb:
3897 self.to_screen('[info] %s is already present' % (
3898 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3899 t['filepath'] = existing_thumb
3900 ret.append((existing_thumb, thumb_filename_final))
3901 else:
3902 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3903 try:
3904 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3905 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3906 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3907 shutil.copyfileobj(uf, thumbf)
3908 ret.append((thumb_filename, thumb_filename_final))
3909 t['filepath'] = thumb_filename
3910 except network_exceptions as err:
3911 thumbnails.pop(idx)
3912 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3913 if ret and not write_all:
3914 break
3915 return ret