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