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