]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[dash,youtube] Download live from start to end (#888)
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 #!/usr/bin/env python3
2 # coding: utf-8
3
4 from __future__ import absolute_import, unicode_literals
5
6 import collections
7 import contextlib
8 import datetime
9 import errno
10 import fileinput
11 import functools
12 import io
13 import itertools
14 import json
15 import locale
16 import operator
17 import os
18 import platform
19 import re
20 import shutil
21 import subprocess
22 import sys
23 import tempfile
24 import time
25 import tokenize
26 import traceback
27 import random
28 import unicodedata
29
30 from enum import Enum
31 from string import ascii_letters
32
33 from .compat import (
34 compat_basestring,
35 compat_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 DownloadCancelled,
59 DownloadError,
60 encode_compat_str,
61 encodeFilename,
62 EntryNotInPlaylist,
63 error_to_compat_str,
64 ExistingVideoReached,
65 expand_path,
66 ExtractorError,
67 float_or_none,
68 format_bytes,
69 format_field,
70 formatSeconds,
71 GeoRestrictedError,
72 get_domain,
73 HEADRequest,
74 int_or_none,
75 iri_to_uri,
76 ISO3166Utils,
77 join_nonempty,
78 LazyList,
79 LINK_TEMPLATES,
80 locked_file,
81 make_dir,
82 make_HTTPS_handler,
83 MaxDownloadsReached,
84 network_exceptions,
85 number_of_digits,
86 orderedSet,
87 OUTTMPL_TYPES,
88 PagedList,
89 parse_filesize,
90 PerRequestProxyHandler,
91 platform_name,
92 Popen,
93 PostProcessingError,
94 preferredencoding,
95 prepend_extension,
96 ReExtractInfo,
97 register_socks_protocols,
98 RejectedVideoReached,
99 remove_terminal_sequences,
100 render_table,
101 replace_extension,
102 SameFileError,
103 sanitize_filename,
104 sanitize_path,
105 sanitize_url,
106 sanitized_Request,
107 std_headers,
108 STR_FORMAT_RE_TMPL,
109 STR_FORMAT_TYPES,
110 str_or_none,
111 strftime_or_none,
112 subtitles_filename,
113 supports_terminal_sequences,
114 timetuple_from_msec,
115 to_high_limit_path,
116 traverse_obj,
117 try_get,
118 UnavailableVideoError,
119 url_basename,
120 variadic,
121 version_tuple,
122 write_json_file,
123 write_string,
124 YoutubeDLCookieProcessor,
125 YoutubeDLHandler,
126 YoutubeDLRedirectHandler,
127 )
128 from .cache import Cache
129 from .minicurses import format_text
130 from .extractor import (
131 gen_extractor_classes,
132 get_info_extractor,
133 _LAZY_LOADER,
134 _PLUGIN_CLASSES as plugin_extractors
135 )
136 from .extractor.openload import PhantomJSwrapper
137 from .downloader import (
138 FFmpegFD,
139 get_suitable_downloader,
140 shorten_protocol_name
141 )
142 from .downloader.rtmp import rtmpdump_version
143 from .postprocessor import (
144 get_postprocessor,
145 EmbedThumbnailPP,
146 FFmpegFixupDuplicateMoovPP,
147 FFmpegFixupDurationPP,
148 FFmpegFixupM3u8PP,
149 FFmpegFixupM4aPP,
150 FFmpegFixupStretchedPP,
151 FFmpegFixupTimestampPP,
152 FFmpegMergerPP,
153 FFmpegPostProcessor,
154 MoveFilesAfterDownloadPP,
155 _PLUGIN_CLASSES as plugin_postprocessors
156 )
157 from .update import detect_variant
158 from .version import __version__, RELEASE_GIT_HEAD
159
160 if compat_os_name == 'nt':
161 import ctypes
162
163
164 class YoutubeDL(object):
165 """YoutubeDL class.
166
167 YoutubeDL objects are the ones responsible of downloading the
168 actual video file and writing it to disk if the user has requested
169 it, among some other tasks. In most cases there should be one per
170 program. As, given a video URL, the downloader doesn't know how to
171 extract all the needed information, task that InfoExtractors do, it
172 has to pass the URL to one of them.
173
174 For this, YoutubeDL objects have a method that allows
175 InfoExtractors to be registered in a given order. When it is passed
176 a URL, the YoutubeDL object handles it to the first InfoExtractor it
177 finds that reports being able to handle it. The InfoExtractor extracts
178 all the information about the video or videos the URL refers to, and
179 YoutubeDL process the extracted information, possibly using a File
180 Downloader to download the video.
181
182 YoutubeDL objects accept a lot of parameters. In order not to saturate
183 the object constructor with arguments, it receives a dictionary of
184 options instead. These options are available through the params
185 attribute for the InfoExtractors to use. The YoutubeDL also
186 registers itself as the downloader in charge for the InfoExtractors
187 that are added to it, so this is a "mutual registration".
188
189 Available options:
190
191 username: Username for authentication purposes.
192 password: Password for authentication purposes.
193 videopassword: Password for accessing a video.
194 ap_mso: Adobe Pass multiple-system operator identifier.
195 ap_username: Multiple-system operator account username.
196 ap_password: Multiple-system operator account password.
197 usenetrc: Use netrc for authentication instead.
198 verbose: Print additional info to stdout.
199 quiet: Do not print messages to stdout.
200 no_warnings: Do not print out anything for warnings.
201 forceprint: A list of templates to force print
202 forceurl: Force printing final URL. (Deprecated)
203 forcetitle: Force printing title. (Deprecated)
204 forceid: Force printing ID. (Deprecated)
205 forcethumbnail: Force printing thumbnail URL. (Deprecated)
206 forcedescription: Force printing description. (Deprecated)
207 forcefilename: Force printing final filename. (Deprecated)
208 forceduration: Force printing duration. (Deprecated)
209 forcejson: Force printing info_dict as JSON.
210 dump_single_json: Force printing the info_dict of the whole playlist
211 (or video) as a single JSON line.
212 force_write_download_archive: Force writing download archive regardless
213 of 'skip_download' or 'simulate'.
214 simulate: Do not download the video files. If unset (or None),
215 simulate only if listsubtitles, listformats or list_thumbnails is used
216 format: Video format code. see "FORMAT SELECTION" for more details.
217 You can also pass a function. The function takes 'ctx' as
218 argument and returns the formats to download.
219 See "build_format_selector" for an implementation
220 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
221 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
222 extracting metadata even if the video is not actually
223 available for download (experimental)
224 format_sort: A list of fields by which to sort the video formats.
225 See "Sorting Formats" for more details.
226 format_sort_force: Force the given format_sort. see "Sorting Formats"
227 for more details.
228 allow_multiple_video_streams: Allow multiple video streams to be merged
229 into a single file
230 allow_multiple_audio_streams: Allow multiple audio streams to be merged
231 into a single file
232 check_formats Whether to test if the formats are downloadable.
233 Can be True (check all), False (check none),
234 'selected' (check selected formats),
235 or None (check only if requested by extractor)
236 paths: Dictionary of output paths. The allowed keys are 'home'
237 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
238 outtmpl: Dictionary of templates for output names. Allowed keys
239 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
240 For compatibility with youtube-dl, a single string can also be used
241 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
242 restrictfilenames: Do not allow "&" and spaces in file names
243 trim_file_name: Limit length of filename (extension excluded)
244 windowsfilenames: Force the filenames to be windows compatible
245 ignoreerrors: Do not stop on download/postprocessing errors.
246 Can be 'only_download' to ignore only download errors.
247 Default is 'only_download' for CLI, but False for API
248 skip_playlist_after_errors: Number of allowed failures until the rest of
249 the playlist is skipped
250 force_generic_extractor: Force downloader to use the generic extractor
251 overwrites: Overwrite all video and metadata files if True,
252 overwrite only non-video files if None
253 and don't overwrite any file if False
254 For compatibility with youtube-dl,
255 "nooverwrites" may also be used instead
256 playliststart: Playlist item to start at.
257 playlistend: Playlist item to end at.
258 playlist_items: Specific indices of playlist to download.
259 playlistreverse: Download playlist items in reverse order.
260 playlistrandom: Download playlist items in random order.
261 matchtitle: Download only matching titles.
262 rejecttitle: Reject downloads for matching titles.
263 logger: Log messages to a logging.Logger instance.
264 logtostderr: Log messages to stderr instead of stdout.
265 consoletitle: Display progress in console window's titlebar.
266 writedescription: Write the video description to a .description file
267 writeinfojson: Write the video description to a .info.json file
268 clean_infojson: Remove private fields from the infojson
269 getcomments: Extract video comments. This will not be written to disk
270 unless writeinfojson is also given
271 writeannotations: Write the video annotations to a .annotations.xml file
272 writethumbnail: Write the thumbnail image to a file
273 allow_playlist_files: Whether to write playlists' description, infojson etc
274 also to disk when using the 'write*' options
275 write_all_thumbnails: Write all thumbnail formats to files
276 writelink: Write an internet shortcut file, depending on the
277 current platform (.url/.webloc/.desktop)
278 writeurllink: Write a Windows internet shortcut file (.url)
279 writewebloclink: Write a macOS internet shortcut file (.webloc)
280 writedesktoplink: Write a Linux internet shortcut file (.desktop)
281 writesubtitles: Write the video subtitles to a file
282 writeautomaticsub: Write the automatically generated subtitles to a file
283 allsubtitles: Deprecated - Use subtitleslangs = ['all']
284 Downloads all the subtitles of the video
285 (requires writesubtitles or writeautomaticsub)
286 listsubtitles: Lists all available subtitles for the video
287 subtitlesformat: The format code for subtitles
288 subtitleslangs: List of languages of the subtitles to download (can be regex).
289 The list may contain "all" to refer to all the available
290 subtitles. The language can be prefixed with a "-" to
291 exclude it from the requested languages. Eg: ['all', '-live_chat']
292 keepvideo: Keep the video file after post-processing
293 daterange: A DateRange object, download only if the upload_date is in the range.
294 skip_download: Skip the actual download of the video file
295 cachedir: Location of the cache files in the filesystem.
296 False to disable filesystem cache.
297 noplaylist: Download single video instead of a playlist if in doubt.
298 age_limit: An integer representing the user's age in years.
299 Unsuitable videos for the given age are skipped.
300 min_views: An integer representing the minimum view count the video
301 must have in order to not be skipped.
302 Videos without view count information are always
303 downloaded. None for no limit.
304 max_views: An integer representing the maximum view count.
305 Videos that are more popular than that are not
306 downloaded.
307 Videos without view count information are always
308 downloaded. None for no limit.
309 download_archive: File name of a file where all downloads are recorded.
310 Videos already present in the file are not downloaded
311 again.
312 break_on_existing: Stop the download process after attempting to download a
313 file that is in the archive.
314 break_on_reject: Stop the download process when encountering a video that
315 has been filtered out.
316 break_per_url: Whether break_on_reject and break_on_existing
317 should act on each input URL as opposed to for the entire queue
318 cookiefile: File name where cookies should be read from and dumped to
319 cookiesfrombrowser: A tuple containing the name of the browser and the profile
320 name/path from where cookies are loaded.
321 Eg: ('chrome', ) or ('vivaldi', 'default')
322 nocheckcertificate:Do not verify SSL certificates
323 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
324 At the moment, this is only supported by YouTube.
325 proxy: URL of the proxy server to use
326 geo_verification_proxy: URL of the proxy to use for IP address verification
327 on geo-restricted sites.
328 socket_timeout: Time to wait for unresponsive hosts, in seconds
329 bidi_workaround: Work around buggy terminals without bidirectional text
330 support, using fridibi
331 debug_printtraffic:Print out sent and received HTTP traffic
332 include_ads: Download ads as well (deprecated)
333 default_search: Prepend this string if an input url is not valid.
334 'auto' for elaborate guessing
335 encoding: Use this encoding instead of the system-specified.
336 extract_flat: Do not resolve URLs, return the immediate result.
337 Pass in 'in_playlist' to only show this behavior for
338 playlist items.
339 wait_for_video: If given, wait for scheduled streams to become available.
340 The value should be a tuple containing the range
341 (min_secs, max_secs) to wait between retries
342 postprocessors: A list of dictionaries, each with an entry
343 * key: The name of the postprocessor. See
344 yt_dlp/postprocessor/__init__.py for a list.
345 * when: When to run the postprocessor. Can be one of
346 pre_process|before_dl|post_process|after_move.
347 Assumed to be 'post_process' if not given
348 post_hooks: Deprecated - Register a custom postprocessor instead
349 A list of functions that get called as the final step
350 for each video file, after all postprocessors have been
351 called. The filename will be passed as the only argument.
352 progress_hooks: A list of functions that get called on download
353 progress, with a dictionary with the entries
354 * status: One of "downloading", "error", or "finished".
355 Check this first and ignore unknown values.
356 * info_dict: The extracted info_dict
357
358 If status is one of "downloading", or "finished", the
359 following properties may also be present:
360 * filename: The final filename (always present)
361 * tmpfilename: The filename we're currently writing to
362 * downloaded_bytes: Bytes on disk
363 * total_bytes: Size of the whole file, None if unknown
364 * total_bytes_estimate: Guess of the eventual file size,
365 None if unavailable.
366 * elapsed: The number of seconds since download started.
367 * eta: The estimated time in seconds, None if unknown
368 * speed: The download speed in bytes/second, None if
369 unknown
370 * fragment_index: The counter of the currently
371 downloaded video fragment.
372 * fragment_count: The number of fragments (= individual
373 files that will be merged)
374
375 Progress hooks are guaranteed to be called at least once
376 (with status "finished") if the download is successful.
377 postprocessor_hooks: A list of functions that get called on postprocessing
378 progress, with a dictionary with the entries
379 * status: One of "started", "processing", or "finished".
380 Check this first and ignore unknown values.
381 * postprocessor: Name of the postprocessor
382 * info_dict: The extracted info_dict
383
384 Progress hooks are guaranteed to be called at least twice
385 (with status "started" and "finished") if the processing is successful.
386 merge_output_format: Extension to use when merging formats.
387 final_ext: Expected final extension; used to detect when the file was
388 already downloaded and converted
389 fixup: Automatically correct known faults of the file.
390 One of:
391 - "never": do nothing
392 - "warn": only emit a warning
393 - "detect_or_warn": check whether we can do anything
394 about it, warn otherwise (default)
395 source_address: Client-side IP address to bind to.
396 call_home: Boolean, true iff we are allowed to contact the
397 yt-dlp servers for debugging. (BROKEN)
398 sleep_interval_requests: Number of seconds to sleep between requests
399 during extraction
400 sleep_interval: Number of seconds to sleep before each download when
401 used alone or a lower bound of a range for randomized
402 sleep before each download (minimum possible number
403 of seconds to sleep) when used along with
404 max_sleep_interval.
405 max_sleep_interval:Upper bound of a range for randomized sleep before each
406 download (maximum possible number of seconds to sleep).
407 Must only be used along with sleep_interval.
408 Actual sleep time will be a random float from range
409 [sleep_interval; max_sleep_interval].
410 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
411 listformats: Print an overview of available video formats and exit.
412 list_thumbnails: Print a table of all thumbnails and exit.
413 match_filter: A function that gets called with the info_dict of
414 every video.
415 If it returns a message, the video is ignored.
416 If it returns None, the video is downloaded.
417 match_filter_func in utils.py is one example for this.
418 no_color: Do not emit color codes in output.
419 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
420 HTTP header
421 geo_bypass_country:
422 Two-letter ISO 3166-2 country code that will be used for
423 explicit geographic restriction bypassing via faking
424 X-Forwarded-For HTTP header
425 geo_bypass_ip_block:
426 IP range in CIDR notation that will be used similarly to
427 geo_bypass_country
428
429 The following options determine which downloader is picked:
430 external_downloader: A dictionary of protocol keys and the executable of the
431 external downloader to use for it. The allowed protocols
432 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
433 Set the value to 'native' to use the native downloader
434 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
435 or {'m3u8': 'ffmpeg'} instead.
436 Use the native HLS downloader instead of ffmpeg/avconv
437 if True, otherwise use ffmpeg/avconv if False, otherwise
438 use downloader suggested by extractor if None.
439 compat_opts: Compatibility options. See "Differences in default behavior".
440 The following options do not work when used through the API:
441 filename, abort-on-error, multistreams, no-live-chat, format-sort
442 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
443 Refer __init__.py for their implementation
444 progress_template: Dictionary of templates for progress outputs.
445 Allowed keys are 'download', 'postprocess',
446 'download-title' (console title) and 'postprocess-title'.
447 The template is mapped on a dictionary with keys 'progress' and 'info'
448
449 The following parameters are not used by YoutubeDL itself, they are used by
450 the downloader (see yt_dlp/downloader/common.py):
451 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
452 max_filesize, test, noresizebuffer, retries, fragment_retries, continuedl,
453 noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
454 external_downloader_args, concurrent_fragment_downloads.
455
456 The following options are used by the post processors:
457 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
458 otherwise prefer ffmpeg. (avconv support is deprecated)
459 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
460 to the binary or its containing directory.
461 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
462 and a list of additional command-line arguments for the
463 postprocessor/executable. The dict can also have "PP+EXE" keys
464 which are used when the given exe is used by the given PP.
465 Use 'default' as the name for arguments to passed to all PP
466 For compatibility with youtube-dl, a single list of args
467 can also be used
468
469 The following options are used by the extractors:
470 extractor_retries: Number of times to retry for known errors
471 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
472 hls_split_discontinuity: Split HLS playlists to different formats at
473 discontinuities such as ad breaks (default: False)
474 extractor_args: A dictionary of arguments to be passed to the extractors.
475 See "EXTRACTOR ARGUMENTS" for details.
476 Eg: {'youtube': {'skip': ['dash', 'hls']}}
477 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
478 If True (default), DASH manifests and related
479 data will be downloaded and processed by extractor.
480 You can reduce network I/O by disabling it if you don't
481 care about DASH. (only for youtube)
482 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
483 If True (default), HLS manifests and related
484 data will be downloaded and processed by extractor.
485 You can reduce network I/O by disabling it if you don't
486 care about HLS. (only for youtube)
487 """
488
489 _NUMERIC_FIELDS = set((
490 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
491 'timestamp', 'release_timestamp',
492 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
493 'average_rating', 'comment_count', 'age_limit',
494 'start_time', 'end_time',
495 'chapter_number', 'season_number', 'episode_number',
496 'track_number', 'disc_number', 'release_year',
497 ))
498
499 _format_selection_exts = {
500 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
501 'video': {'mp4', 'flv', 'webm', '3gp'},
502 'storyboards': {'mhtml'},
503 }
504
505 params = None
506 _ies = {}
507 _pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
508 _printed_messages = set()
509 _first_webpage_request = True
510 _download_retcode = None
511 _num_downloads = None
512 _playlist_level = 0
513 _playlist_urls = set()
514 _screen_file = None
515
516 def __init__(self, params=None, auto_init=True):
517 """Create a FileDownloader object with the given options.
518 @param auto_init Whether to load the default extractors and print header (if verbose).
519 Set to 'no_verbose_header' to not print the header
520 """
521 if params is None:
522 params = {}
523 self._ies = {}
524 self._ies_instances = {}
525 self._pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
526 self._printed_messages = set()
527 self._first_webpage_request = True
528 self._post_hooks = []
529 self._progress_hooks = []
530 self._postprocessor_hooks = []
531 self._download_retcode = 0
532 self._num_downloads = 0
533 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
534 self._err_file = sys.stderr
535 self.params = params
536 self.cache = Cache(self)
537
538 windows_enable_vt_mode()
539 self._allow_colors = {
540 'screen': not self.params.get('no_color') and supports_terminal_sequences(self._screen_file),
541 'err': not self.params.get('no_color') and supports_terminal_sequences(self._err_file),
542 }
543
544 if sys.version_info < (3, 6):
545 self.report_warning(
546 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
547
548 if self.params.get('allow_unplayable_formats'):
549 self.report_warning(
550 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
551 'This is a developer option intended for debugging. \n'
552 ' If you experience any issues while using this option, '
553 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
554
555 def check_deprecated(param, option, suggestion):
556 if self.params.get(param) is not None:
557 self.report_warning('%s is deprecated. Use %s instead' % (option, suggestion))
558 return True
559 return False
560
561 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
562 if self.params.get('geo_verification_proxy') is None:
563 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
564
565 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
566 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
567 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
568
569 for msg in self.params.get('_warnings', []):
570 self.report_warning(msg)
571 for msg in self.params.get('_deprecation_warnings', []):
572 self.deprecation_warning(msg)
573
574 if 'list-formats' in self.params.get('compat_opts', []):
575 self.params['listformats_table'] = False
576
577 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
578 # nooverwrites was unnecessarily changed to overwrites
579 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
580 # This ensures compatibility with both keys
581 self.params['overwrites'] = not self.params['nooverwrites']
582 elif self.params.get('overwrites') is None:
583 self.params.pop('overwrites', None)
584 else:
585 self.params['nooverwrites'] = not self.params['overwrites']
586
587 if params.get('bidi_workaround', False):
588 try:
589 import pty
590 master, slave = pty.openpty()
591 width = compat_get_terminal_size().columns
592 if width is None:
593 width_args = []
594 else:
595 width_args = ['-w', str(width)]
596 sp_kwargs = dict(
597 stdin=subprocess.PIPE,
598 stdout=slave,
599 stderr=self._err_file)
600 try:
601 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
602 except OSError:
603 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
604 self._output_channel = os.fdopen(master, 'rb')
605 except OSError as ose:
606 if ose.errno == errno.ENOENT:
607 self.report_warning(
608 'Could not find fribidi executable, ignoring --bidi-workaround. '
609 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
610 else:
611 raise
612
613 if (sys.platform != 'win32'
614 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
615 and not params.get('restrictfilenames', False)):
616 # Unicode filesystem API will throw errors (#1474, #13027)
617 self.report_warning(
618 'Assuming --restrict-filenames since file system encoding '
619 'cannot encode all characters. '
620 'Set the LC_ALL environment variable to fix this.')
621 self.params['restrictfilenames'] = True
622
623 self.outtmpl_dict = self.parse_outtmpl()
624
625 # Creating format selector here allows us to catch syntax errors before the extraction
626 self.format_selector = (
627 None if self.params.get('format') is None
628 else self.params['format'] if callable(self.params['format'])
629 else self.build_format_selector(self.params['format']))
630
631 self._setup_opener()
632
633 if auto_init:
634 if auto_init != 'no_verbose_header':
635 self.print_debug_header()
636 self.add_default_info_extractors()
637
638 hooks = {
639 'post_hooks': self.add_post_hook,
640 'progress_hooks': self.add_progress_hook,
641 'postprocessor_hooks': self.add_postprocessor_hook,
642 }
643 for opt, fn in hooks.items():
644 for ph in self.params.get(opt, []):
645 fn(ph)
646
647 for pp_def_raw in self.params.get('postprocessors', []):
648 pp_def = dict(pp_def_raw)
649 when = pp_def.pop('when', 'post_process')
650 self.add_post_processor(
651 get_postprocessor(pp_def.pop('key'))(self, **compat_kwargs(pp_def)),
652 when=when)
653
654 register_socks_protocols()
655
656 def preload_download_archive(fn):
657 """Preload the archive, if any is specified"""
658 if fn is None:
659 return False
660 self.write_debug(f'Loading archive file {fn!r}')
661 try:
662 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
663 for line in archive_file:
664 self.archive.add(line.strip())
665 except IOError as ioe:
666 if ioe.errno != errno.ENOENT:
667 raise
668 return False
669 return True
670
671 self.archive = set()
672 preload_download_archive(self.params.get('download_archive'))
673
674 def warn_if_short_id(self, argv):
675 # short YouTube ID starting with dash?
676 idxs = [
677 i for i, a in enumerate(argv)
678 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
679 if idxs:
680 correct_argv = (
681 ['yt-dlp']
682 + [a for i, a in enumerate(argv) if i not in idxs]
683 + ['--'] + [argv[i] for i in idxs]
684 )
685 self.report_warning(
686 'Long argument string detected. '
687 'Use -- to separate parameters and URLs, like this:\n%s' %
688 args_to_str(correct_argv))
689
690 def add_info_extractor(self, ie):
691 """Add an InfoExtractor object to the end of the list."""
692 ie_key = ie.ie_key()
693 self._ies[ie_key] = ie
694 if not isinstance(ie, type):
695 self._ies_instances[ie_key] = ie
696 ie.set_downloader(self)
697
698 def _get_info_extractor_class(self, ie_key):
699 ie = self._ies.get(ie_key)
700 if ie is None:
701 ie = get_info_extractor(ie_key)
702 self.add_info_extractor(ie)
703 return ie
704
705 def get_info_extractor(self, ie_key):
706 """
707 Get an instance of an IE with name ie_key, it will try to get one from
708 the _ies list, if there's no instance it will create a new one and add
709 it to the extractor list.
710 """
711 ie = self._ies_instances.get(ie_key)
712 if ie is None:
713 ie = get_info_extractor(ie_key)()
714 self.add_info_extractor(ie)
715 return ie
716
717 def add_default_info_extractors(self):
718 """
719 Add the InfoExtractors returned by gen_extractors to the end of the list
720 """
721 for ie in gen_extractor_classes():
722 self.add_info_extractor(ie)
723
724 def add_post_processor(self, pp, when='post_process'):
725 """Add a PostProcessor object to the end of the chain."""
726 self._pps[when].append(pp)
727 pp.set_downloader(self)
728
729 def add_post_hook(self, ph):
730 """Add the post hook"""
731 self._post_hooks.append(ph)
732
733 def add_progress_hook(self, ph):
734 """Add the download progress hook"""
735 self._progress_hooks.append(ph)
736
737 def add_postprocessor_hook(self, ph):
738 """Add the postprocessing progress hook"""
739 self._postprocessor_hooks.append(ph)
740 for pps in self._pps.values():
741 for pp in pps:
742 pp.add_progress_hook(ph)
743
744 def _bidi_workaround(self, message):
745 if not hasattr(self, '_output_channel'):
746 return message
747
748 assert hasattr(self, '_output_process')
749 assert isinstance(message, compat_str)
750 line_count = message.count('\n') + 1
751 self._output_process.stdin.write((message + '\n').encode('utf-8'))
752 self._output_process.stdin.flush()
753 res = ''.join(self._output_channel.readline().decode('utf-8')
754 for _ in range(line_count))
755 return res[:-len('\n')]
756
757 def _write_string(self, message, out=None, only_once=False):
758 if only_once:
759 if message in self._printed_messages:
760 return
761 self._printed_messages.add(message)
762 write_string(message, out=out, encoding=self.params.get('encoding'))
763
764 def to_stdout(self, message, skip_eol=False, quiet=False):
765 """Print message to stdout"""
766 if self.params.get('logger'):
767 self.params['logger'].debug(message)
768 elif not quiet or self.params.get('verbose'):
769 self._write_string(
770 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
771 self._err_file if quiet else self._screen_file)
772
773 def to_stderr(self, message, only_once=False):
774 """Print message to stderr"""
775 assert isinstance(message, compat_str)
776 if self.params.get('logger'):
777 self.params['logger'].error(message)
778 else:
779 self._write_string('%s\n' % self._bidi_workaround(message), self._err_file, only_once=only_once)
780
781 def to_console_title(self, message):
782 if not self.params.get('consoletitle', False):
783 return
784 message = remove_terminal_sequences(message)
785 if compat_os_name == 'nt':
786 if ctypes.windll.kernel32.GetConsoleWindow():
787 # c_wchar_p() might not be necessary if `message` is
788 # already of type unicode()
789 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
790 elif 'TERM' in os.environ:
791 self._write_string('\033]0;%s\007' % message, self._screen_file)
792
793 def save_console_title(self):
794 if not self.params.get('consoletitle', False):
795 return
796 if self.params.get('simulate'):
797 return
798 if compat_os_name != 'nt' and 'TERM' in os.environ:
799 # Save the title on stack
800 self._write_string('\033[22;0t', self._screen_file)
801
802 def restore_console_title(self):
803 if not self.params.get('consoletitle', False):
804 return
805 if self.params.get('simulate'):
806 return
807 if compat_os_name != 'nt' and 'TERM' in os.environ:
808 # Restore the title from stack
809 self._write_string('\033[23;0t', self._screen_file)
810
811 def __enter__(self):
812 self.save_console_title()
813 return self
814
815 def __exit__(self, *args):
816 self.restore_console_title()
817
818 if self.params.get('cookiefile') is not None:
819 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
820
821 def trouble(self, message=None, tb=None):
822 """Determine action to take when a download problem appears.
823
824 Depending on if the downloader has been configured to ignore
825 download errors or not, this method may throw an exception or
826 not when errors are found, after printing the message.
827
828 tb, if given, is additional traceback information.
829 """
830 if message is not None:
831 self.to_stderr(message)
832 if self.params.get('verbose'):
833 if tb is None:
834 if sys.exc_info()[0]: # if .trouble has been called from an except block
835 tb = ''
836 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
837 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
838 tb += encode_compat_str(traceback.format_exc())
839 else:
840 tb_data = traceback.format_list(traceback.extract_stack())
841 tb = ''.join(tb_data)
842 if tb:
843 self.to_stderr(tb)
844 if not self.params.get('ignoreerrors'):
845 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
846 exc_info = sys.exc_info()[1].exc_info
847 else:
848 exc_info = sys.exc_info()
849 raise DownloadError(message, exc_info)
850 self._download_retcode = 1
851
852 def to_screen(self, message, skip_eol=False):
853 """Print message to stdout if not in quiet mode"""
854 self.to_stdout(
855 message, skip_eol, quiet=self.params.get('quiet', False))
856
857 class Styles(Enum):
858 HEADERS = 'yellow'
859 EMPHASIS = 'light blue'
860 ID = 'green'
861 DELIM = 'blue'
862 ERROR = 'red'
863 WARNING = 'yellow'
864 SUPPRESS = 'light black'
865
866 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
867 if test_encoding:
868 original_text = text
869 encoding = self.params.get('encoding') or getattr(handle, 'encoding', 'ascii')
870 text = text.encode(encoding, 'ignore').decode(encoding)
871 if fallback is not None and text != original_text:
872 text = fallback
873 if isinstance(f, self.Styles):
874 f = f.value
875 return format_text(text, f) if allow_colors else text if fallback is None else fallback
876
877 def _format_screen(self, *args, **kwargs):
878 return self._format_text(
879 self._screen_file, self._allow_colors['screen'], *args, **kwargs)
880
881 def _format_err(self, *args, **kwargs):
882 return self._format_text(
883 self._err_file, self._allow_colors['err'], *args, **kwargs)
884
885 def report_warning(self, message, only_once=False):
886 '''
887 Print the message to stderr, it will be prefixed with 'WARNING:'
888 If stderr is a tty file the 'WARNING:' will be colored
889 '''
890 if self.params.get('logger') is not None:
891 self.params['logger'].warning(message)
892 else:
893 if self.params.get('no_warnings'):
894 return
895 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
896
897 def deprecation_warning(self, message):
898 if self.params.get('logger') is not None:
899 self.params['logger'].warning('DeprecationWarning: {message}')
900 else:
901 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
902
903 def report_error(self, message, tb=None):
904 '''
905 Do the same as trouble, but prefixes the message with 'ERROR:', colored
906 in red if stderr is a tty file.
907 '''
908 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', tb)
909
910 def write_debug(self, message, only_once=False):
911 '''Log debug message or Print message to stderr'''
912 if not self.params.get('verbose', False):
913 return
914 message = '[debug] %s' % message
915 if self.params.get('logger'):
916 self.params['logger'].debug(message)
917 else:
918 self.to_stderr(message, only_once)
919
920 def report_file_already_downloaded(self, file_name):
921 """Report file has already been fully downloaded."""
922 try:
923 self.to_screen('[download] %s has already been downloaded' % file_name)
924 except UnicodeEncodeError:
925 self.to_screen('[download] The file has already been downloaded')
926
927 def report_file_delete(self, file_name):
928 """Report that existing file will be deleted."""
929 try:
930 self.to_screen('Deleting existing file %s' % file_name)
931 except UnicodeEncodeError:
932 self.to_screen('Deleting existing file')
933
934 def raise_no_formats(self, info, forced=False):
935 has_drm = info.get('__has_drm')
936 msg = 'This video is DRM protected' if has_drm else 'No video formats found!'
937 expected = self.params.get('ignore_no_formats_error')
938 if forced or not expected:
939 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
940 expected=has_drm or expected)
941 else:
942 self.report_warning(msg)
943
944 def parse_outtmpl(self):
945 outtmpl_dict = self.params.get('outtmpl', {})
946 if not isinstance(outtmpl_dict, dict):
947 outtmpl_dict = {'default': outtmpl_dict}
948 # Remove spaces in the default template
949 if self.params.get('restrictfilenames'):
950 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
951 else:
952 sanitize = lambda x: x
953 outtmpl_dict.update({
954 k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items()
955 if outtmpl_dict.get(k) is None})
956 for key, val in outtmpl_dict.items():
957 if isinstance(val, bytes):
958 self.report_warning(
959 'Parameter outtmpl is bytes, but should be a unicode string. '
960 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
961 return outtmpl_dict
962
963 def get_output_path(self, dir_type='', filename=None):
964 paths = self.params.get('paths', {})
965 assert isinstance(paths, dict)
966 path = os.path.join(
967 expand_path(paths.get('home', '').strip()),
968 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
969 filename or '')
970
971 # Temporary fix for #4787
972 # 'Treat' all problem characters by passing filename through preferredencoding
973 # to workaround encoding issues with subprocess on python2 @ Windows
974 if sys.version_info < (3, 0) and sys.platform == 'win32':
975 path = encodeFilename(path, True).decode(preferredencoding())
976 return sanitize_path(path, force=self.params.get('windowsfilenames'))
977
978 @staticmethod
979 def _outtmpl_expandpath(outtmpl):
980 # expand_path translates '%%' into '%' and '$$' into '$'
981 # correspondingly that is not what we want since we need to keep
982 # '%%' intact for template dict substitution step. Working around
983 # with boundary-alike separator hack.
984 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
985 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
986
987 # outtmpl should be expand_path'ed before template dict substitution
988 # because meta fields may contain env variables we don't want to
989 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
990 # title "Hello $PATH", we don't want `$PATH` to be expanded.
991 return expand_path(outtmpl).replace(sep, '')
992
993 @staticmethod
994 def escape_outtmpl(outtmpl):
995 ''' Escape any remaining strings like %s, %abc% etc. '''
996 return re.sub(
997 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
998 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
999 outtmpl)
1000
1001 @classmethod
1002 def validate_outtmpl(cls, outtmpl):
1003 ''' @return None or Exception object '''
1004 outtmpl = re.sub(
1005 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBU]'),
1006 lambda mobj: f'{mobj.group(0)[:-1]}s',
1007 cls._outtmpl_expandpath(outtmpl))
1008 try:
1009 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1010 return None
1011 except ValueError as err:
1012 return err
1013
1014 @staticmethod
1015 def _copy_infodict(info_dict):
1016 info_dict = dict(info_dict)
1017 for key in ('__original_infodict', '__postprocessors'):
1018 info_dict.pop(key, None)
1019 return info_dict
1020
1021 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=None):
1022 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict """
1023 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1024
1025 info_dict = self._copy_infodict(info_dict)
1026 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1027 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1028 if info_dict.get('duration', None) is not None
1029 else None)
1030 info_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
1031 if info_dict.get('resolution') is None:
1032 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1033
1034 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1035 # of %(field)s to %(field)0Nd for backward compatibility
1036 field_size_compat_map = {
1037 'playlist_index': number_of_digits(info_dict.get('_last_playlist_index') or 0),
1038 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1039 'autonumber': self.params.get('autonumber_size') or 5,
1040 }
1041
1042 TMPL_DICT = {}
1043 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBU]'))
1044 MATH_FUNCTIONS = {
1045 '+': float.__add__,
1046 '-': float.__sub__,
1047 }
1048 # Field is of the form key1.key2...
1049 # where keys (except first) can be string, int or slice
1050 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1051 MATH_FIELD_RE = r'''{field}|{num}'''.format(field=FIELD_RE, num=r'-?\d+(?:.\d+)?')
1052 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1053 INTERNAL_FORMAT_RE = re.compile(r'''(?x)
1054 (?P<negate>-)?
1055 (?P<fields>{field})
1056 (?P<maths>(?:{math_op}{math_field})*)
1057 (?:>(?P<strf_format>.+?))?
1058 (?P<alternate>(?<!\\),[^|&)]+)?
1059 (?:&(?P<replacement>.*?))?
1060 (?:\|(?P<default>.*?))?
1061 $'''.format(field=FIELD_RE, math_op=MATH_OPERATORS_RE, math_field=MATH_FIELD_RE))
1062
1063 def _traverse_infodict(k):
1064 k = k.split('.')
1065 if k[0] == '':
1066 k.pop(0)
1067 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1068
1069 def get_value(mdict):
1070 # Object traversal
1071 value = _traverse_infodict(mdict['fields'])
1072 # Negative
1073 if mdict['negate']:
1074 value = float_or_none(value)
1075 if value is not None:
1076 value *= -1
1077 # Do maths
1078 offset_key = mdict['maths']
1079 if offset_key:
1080 value = float_or_none(value)
1081 operator = None
1082 while offset_key:
1083 item = re.match(
1084 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1085 offset_key).group(0)
1086 offset_key = offset_key[len(item):]
1087 if operator is None:
1088 operator = MATH_FUNCTIONS[item]
1089 continue
1090 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1091 offset = float_or_none(item)
1092 if offset is None:
1093 offset = float_or_none(_traverse_infodict(item))
1094 try:
1095 value = operator(value, multiplier * offset)
1096 except (TypeError, ZeroDivisionError):
1097 return None
1098 operator = None
1099 # Datetime formatting
1100 if mdict['strf_format']:
1101 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1102
1103 return value
1104
1105 na = self.params.get('outtmpl_na_placeholder', 'NA')
1106
1107 def _dumpjson_default(obj):
1108 if isinstance(obj, (set, LazyList)):
1109 return list(obj)
1110 return repr(obj)
1111
1112 def create_key(outer_mobj):
1113 if not outer_mobj.group('has_key'):
1114 return outer_mobj.group(0)
1115 key = outer_mobj.group('key')
1116 mobj = re.match(INTERNAL_FORMAT_RE, key)
1117 initial_field = mobj.group('fields').split('.')[-1] if mobj else ''
1118 value, replacement, default = None, None, na
1119 while mobj:
1120 mobj = mobj.groupdict()
1121 default = mobj['default'] if mobj['default'] is not None else default
1122 value = get_value(mobj)
1123 replacement = mobj['replacement']
1124 if value is None and mobj['alternate']:
1125 mobj = re.match(INTERNAL_FORMAT_RE, mobj['alternate'][1:])
1126 else:
1127 break
1128
1129 fmt = outer_mobj.group('format')
1130 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1131 fmt = '0{:d}d'.format(field_size_compat_map[key])
1132
1133 value = default if value is None else value if replacement is None else replacement
1134
1135 flags = outer_mobj.group('conversion') or ''
1136 str_fmt = f'{fmt[:-1]}s'
1137 if fmt[-1] == 'l': # list
1138 delim = '\n' if '#' in flags else ', '
1139 value, fmt = delim.join(variadic(value)), str_fmt
1140 elif fmt[-1] == 'j': # json
1141 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1142 elif fmt[-1] == 'q': # quoted
1143 value = map(str, variadic(value) if '#' in flags else [value])
1144 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1145 elif fmt[-1] == 'B': # bytes
1146 value = f'%{str_fmt}'.encode('utf-8') % str(value).encode('utf-8')
1147 value, fmt = value.decode('utf-8', 'ignore'), 's'
1148 elif fmt[-1] == 'U': # unicode normalized
1149 value, fmt = unicodedata.normalize(
1150 # "+" = compatibility equivalence, "#" = NFD
1151 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1152 value), str_fmt
1153 elif fmt[-1] == 'c':
1154 if value:
1155 value = str(value)[0]
1156 else:
1157 fmt = str_fmt
1158 elif fmt[-1] not in 'rs': # numeric
1159 value = float_or_none(value)
1160 if value is None:
1161 value, fmt = default, 's'
1162
1163 if sanitize:
1164 if fmt[-1] == 'r':
1165 # If value is an object, sanitize might convert it to a string
1166 # So we convert it to repr first
1167 value, fmt = repr(value), str_fmt
1168 if fmt[-1] in 'csr':
1169 value = sanitize(initial_field, value)
1170
1171 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1172 TMPL_DICT[key] = value
1173 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1174
1175 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1176
1177 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1178 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1179 return self.escape_outtmpl(outtmpl) % info_dict
1180
1181 def _prepare_filename(self, info_dict, tmpl_type='default'):
1182 try:
1183 sanitize = lambda k, v: sanitize_filename(
1184 compat_str(v),
1185 restricted=self.params.get('restrictfilenames'),
1186 is_id=(k == 'id' or k.endswith('_id')))
1187 outtmpl = self._outtmpl_expandpath(self.outtmpl_dict.get(tmpl_type, self.outtmpl_dict['default']))
1188 filename = self.evaluate_outtmpl(outtmpl, info_dict, sanitize)
1189
1190 force_ext = OUTTMPL_TYPES.get(tmpl_type)
1191 if filename and force_ext is not None:
1192 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1193
1194 # https://github.com/blackjack4494/youtube-dlc/issues/85
1195 trim_file_name = self.params.get('trim_file_name', False)
1196 if trim_file_name:
1197 no_ext, *ext = filename.rsplit('.', 2)
1198 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1199
1200 return filename
1201 except ValueError as err:
1202 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1203 return None
1204
1205 def prepare_filename(self, info_dict, dir_type='', warn=False):
1206 """Generate the output filename."""
1207
1208 filename = self._prepare_filename(info_dict, dir_type or 'default')
1209 if not filename and dir_type not in ('', 'temp'):
1210 return ''
1211
1212 if warn:
1213 if not self.params.get('paths'):
1214 pass
1215 elif filename == '-':
1216 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1217 elif os.path.isabs(filename):
1218 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1219 if filename == '-' or not filename:
1220 return filename
1221
1222 return self.get_output_path(dir_type, filename)
1223
1224 def _match_entry(self, info_dict, incomplete=False, silent=False):
1225 """ Returns None if the file should be downloaded """
1226
1227 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1228
1229 def check_filter():
1230 if 'title' in info_dict:
1231 # This can happen when we're just evaluating the playlist
1232 title = info_dict['title']
1233 matchtitle = self.params.get('matchtitle', False)
1234 if matchtitle:
1235 if not re.search(matchtitle, title, re.IGNORECASE):
1236 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1237 rejecttitle = self.params.get('rejecttitle', False)
1238 if rejecttitle:
1239 if re.search(rejecttitle, title, re.IGNORECASE):
1240 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1241 date = info_dict.get('upload_date')
1242 if date is not None:
1243 dateRange = self.params.get('daterange', DateRange())
1244 if date not in dateRange:
1245 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
1246 view_count = info_dict.get('view_count')
1247 if view_count is not None:
1248 min_views = self.params.get('min_views')
1249 if min_views is not None and view_count < min_views:
1250 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1251 max_views = self.params.get('max_views')
1252 if max_views is not None and view_count > max_views:
1253 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1254 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1255 return 'Skipping "%s" because it is age restricted' % video_title
1256
1257 match_filter = self.params.get('match_filter')
1258 if match_filter is not None:
1259 try:
1260 ret = match_filter(info_dict, incomplete=incomplete)
1261 except TypeError:
1262 # For backward compatibility
1263 ret = None if incomplete else match_filter(info_dict)
1264 if ret is not None:
1265 return ret
1266 return None
1267
1268 if self.in_download_archive(info_dict):
1269 reason = '%s has already been recorded in the archive' % video_title
1270 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1271 else:
1272 reason = check_filter()
1273 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1274 if reason is not None:
1275 if not silent:
1276 self.to_screen('[download] ' + reason)
1277 if self.params.get(break_opt, False):
1278 raise break_err()
1279 return reason
1280
1281 @staticmethod
1282 def add_extra_info(info_dict, extra_info):
1283 '''Set the keys from extra_info in info dict if they are missing'''
1284 for key, value in extra_info.items():
1285 info_dict.setdefault(key, value)
1286
1287 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1288 process=True, force_generic_extractor=False):
1289 """
1290 Return a list with a dictionary for each video extracted.
1291
1292 Arguments:
1293 url -- URL to extract
1294
1295 Keyword arguments:
1296 download -- whether to download videos during extraction
1297 ie_key -- extractor key hint
1298 extra_info -- dictionary containing the extra values to add to each result
1299 process -- whether to resolve all unresolved references (URLs, playlist items),
1300 must be True for download to work.
1301 force_generic_extractor -- force using the generic extractor
1302 """
1303
1304 if extra_info is None:
1305 extra_info = {}
1306
1307 if not ie_key and force_generic_extractor:
1308 ie_key = 'Generic'
1309
1310 if ie_key:
1311 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1312 else:
1313 ies = self._ies
1314
1315 for ie_key, ie in ies.items():
1316 if not ie.suitable(url):
1317 continue
1318
1319 if not ie.working():
1320 self.report_warning('The program functionality for this site has been marked as broken, '
1321 'and will probably not work.')
1322
1323 temp_id = ie.get_temp_id(url)
1324 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1325 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1326 if self.params.get('break_on_existing', False):
1327 raise ExistingVideoReached()
1328 break
1329 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1330 else:
1331 self.report_error('no suitable InfoExtractor for URL %s' % url)
1332
1333 def __handle_extraction_exceptions(func):
1334 @functools.wraps(func)
1335 def wrapper(self, *args, **kwargs):
1336 try:
1337 return func(self, *args, **kwargs)
1338 except GeoRestrictedError as e:
1339 msg = e.msg
1340 if e.countries:
1341 msg += '\nThis video is available in %s.' % ', '.join(
1342 map(ISO3166Utils.short2full, e.countries))
1343 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1344 self.report_error(msg)
1345 except ExtractorError as e: # An error we somewhat expected
1346 self.report_error(compat_str(e), e.format_traceback())
1347 except ReExtractInfo as e:
1348 if e.expected:
1349 self.to_screen(f'{e}; Re-extracting data')
1350 else:
1351 self.to_stderr('\r')
1352 self.report_warning(f'{e}; Re-extracting data')
1353 return wrapper(self, *args, **kwargs)
1354 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1355 raise
1356 except Exception as e:
1357 if self.params.get('ignoreerrors'):
1358 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
1359 else:
1360 raise
1361 return wrapper
1362
1363 def _wait_for_video(self, ie_result):
1364 if (not self.params.get('wait_for_video')
1365 or ie_result.get('_type', 'video') != 'video'
1366 or ie_result.get('formats') or ie_result.get('url')):
1367 return
1368
1369 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1370 last_msg = ''
1371
1372 def progress(msg):
1373 nonlocal last_msg
1374 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1375 last_msg = msg
1376
1377 min_wait, max_wait = self.params.get('wait_for_video')
1378 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1379 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1380 diff = random.randrange(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait)
1381 self.report_warning('Release time of video is not known')
1382 elif (diff or 0) <= 0:
1383 self.report_warning('Video should already be available according to extracted info')
1384 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1385 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1386
1387 wait_till = time.time() + diff
1388 try:
1389 while True:
1390 diff = wait_till - time.time()
1391 if diff <= 0:
1392 progress('')
1393 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1394 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1395 time.sleep(1)
1396 except KeyboardInterrupt:
1397 progress('')
1398 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1399 except BaseException as e:
1400 if not isinstance(e, ReExtractInfo):
1401 self.to_screen('')
1402 raise
1403
1404 @__handle_extraction_exceptions
1405 def __extract_info(self, url, ie, download, extra_info, process):
1406 ie_result = ie.extract(url)
1407 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1408 return
1409 if isinstance(ie_result, list):
1410 # Backwards compatibility: old IE result format
1411 ie_result = {
1412 '_type': 'compat_list',
1413 'entries': ie_result,
1414 }
1415 if extra_info.get('original_url'):
1416 ie_result.setdefault('original_url', extra_info['original_url'])
1417 self.add_default_extra_info(ie_result, ie, url)
1418 if process:
1419 self._wait_for_video(ie_result)
1420 return self.process_ie_result(ie_result, download, extra_info)
1421 else:
1422 return ie_result
1423
1424 def add_default_extra_info(self, ie_result, ie, url):
1425 if url is not None:
1426 self.add_extra_info(ie_result, {
1427 'webpage_url': url,
1428 'original_url': url,
1429 'webpage_url_basename': url_basename(url),
1430 'webpage_url_domain': get_domain(url),
1431 })
1432 if ie is not None:
1433 self.add_extra_info(ie_result, {
1434 'extractor': ie.IE_NAME,
1435 'extractor_key': ie.ie_key(),
1436 })
1437
1438 def process_ie_result(self, ie_result, download=True, extra_info=None):
1439 """
1440 Take the result of the ie(may be modified) and resolve all unresolved
1441 references (URLs, playlist items).
1442
1443 It will also download the videos if 'download'.
1444 Returns the resolved ie_result.
1445 """
1446 if extra_info is None:
1447 extra_info = {}
1448 result_type = ie_result.get('_type', 'video')
1449
1450 if result_type in ('url', 'url_transparent'):
1451 ie_result['url'] = sanitize_url(ie_result['url'])
1452 if ie_result.get('original_url'):
1453 extra_info.setdefault('original_url', ie_result['original_url'])
1454
1455 extract_flat = self.params.get('extract_flat', False)
1456 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1457 or extract_flat is True):
1458 info_copy = ie_result.copy()
1459 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1460 if ie and not ie_result.get('id'):
1461 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1462 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1463 self.add_extra_info(info_copy, extra_info)
1464 info_copy, _ = self.pre_process(info_copy)
1465 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1466 if self.params.get('force_write_download_archive', False):
1467 self.record_download_archive(info_copy)
1468 return ie_result
1469
1470 if result_type == 'video':
1471 self.add_extra_info(ie_result, extra_info)
1472 ie_result = self.process_video_result(ie_result, download=download)
1473 additional_urls = (ie_result or {}).get('additional_urls')
1474 if additional_urls:
1475 # TODO: Improve MetadataParserPP to allow setting a list
1476 if isinstance(additional_urls, compat_str):
1477 additional_urls = [additional_urls]
1478 self.to_screen(
1479 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1480 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1481 ie_result['additional_entries'] = [
1482 self.extract_info(
1483 url, download, extra_info,
1484 force_generic_extractor=self.params.get('force_generic_extractor'))
1485 for url in additional_urls
1486 ]
1487 return ie_result
1488 elif result_type == 'url':
1489 # We have to add extra_info to the results because it may be
1490 # contained in a playlist
1491 return self.extract_info(
1492 ie_result['url'], download,
1493 ie_key=ie_result.get('ie_key'),
1494 extra_info=extra_info)
1495 elif result_type == 'url_transparent':
1496 # Use the information from the embedding page
1497 info = self.extract_info(
1498 ie_result['url'], ie_key=ie_result.get('ie_key'),
1499 extra_info=extra_info, download=False, process=False)
1500
1501 # extract_info may return None when ignoreerrors is enabled and
1502 # extraction failed with an error, don't crash and return early
1503 # in this case
1504 if not info:
1505 return info
1506
1507 force_properties = dict(
1508 (k, v) for k, v in ie_result.items() if v is not None)
1509 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1510 if f in force_properties:
1511 del force_properties[f]
1512 new_result = info.copy()
1513 new_result.update(force_properties)
1514
1515 # Extracted info may not be a video result (i.e.
1516 # info.get('_type', 'video') != video) but rather an url or
1517 # url_transparent. In such cases outer metadata (from ie_result)
1518 # should be propagated to inner one (info). For this to happen
1519 # _type of info should be overridden with url_transparent. This
1520 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1521 if new_result.get('_type') == 'url':
1522 new_result['_type'] = 'url_transparent'
1523
1524 return self.process_ie_result(
1525 new_result, download=download, extra_info=extra_info)
1526 elif result_type in ('playlist', 'multi_video'):
1527 # Protect from infinite recursion due to recursively nested playlists
1528 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1529 webpage_url = ie_result['webpage_url']
1530 if webpage_url in self._playlist_urls:
1531 self.to_screen(
1532 '[download] Skipping already downloaded playlist: %s'
1533 % ie_result.get('title') or ie_result.get('id'))
1534 return
1535
1536 self._playlist_level += 1
1537 self._playlist_urls.add(webpage_url)
1538 self._sanitize_thumbnails(ie_result)
1539 try:
1540 return self.__process_playlist(ie_result, download)
1541 finally:
1542 self._playlist_level -= 1
1543 if not self._playlist_level:
1544 self._playlist_urls.clear()
1545 elif result_type == 'compat_list':
1546 self.report_warning(
1547 'Extractor %s returned a compat_list result. '
1548 'It needs to be updated.' % ie_result.get('extractor'))
1549
1550 def _fixup(r):
1551 self.add_extra_info(r, {
1552 'extractor': ie_result['extractor'],
1553 'webpage_url': ie_result['webpage_url'],
1554 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1555 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1556 'extractor_key': ie_result['extractor_key'],
1557 })
1558 return r
1559 ie_result['entries'] = [
1560 self.process_ie_result(_fixup(r), download, extra_info)
1561 for r in ie_result['entries']
1562 ]
1563 return ie_result
1564 else:
1565 raise Exception('Invalid result type: %s' % result_type)
1566
1567 def _ensure_dir_exists(self, path):
1568 return make_dir(path, self.report_error)
1569
1570 def __process_playlist(self, ie_result, download):
1571 # We process each entry in the playlist
1572 playlist = ie_result.get('title') or ie_result.get('id')
1573 self.to_screen('[download] Downloading playlist: %s' % playlist)
1574
1575 if 'entries' not in ie_result:
1576 raise EntryNotInPlaylist('There are no entries')
1577
1578 MissingEntry = object()
1579 incomplete_entries = bool(ie_result.get('requested_entries'))
1580 if incomplete_entries:
1581 def fill_missing_entries(entries, indices):
1582 ret = [MissingEntry] * max(indices)
1583 for i, entry in zip(indices, entries):
1584 ret[i - 1] = entry
1585 return ret
1586 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1587
1588 playlist_results = []
1589
1590 playliststart = self.params.get('playliststart', 1)
1591 playlistend = self.params.get('playlistend')
1592 # For backwards compatibility, interpret -1 as whole list
1593 if playlistend == -1:
1594 playlistend = None
1595
1596 playlistitems_str = self.params.get('playlist_items')
1597 playlistitems = None
1598 if playlistitems_str is not None:
1599 def iter_playlistitems(format):
1600 for string_segment in format.split(','):
1601 if '-' in string_segment:
1602 start, end = string_segment.split('-')
1603 for item in range(int(start), int(end) + 1):
1604 yield int(item)
1605 else:
1606 yield int(string_segment)
1607 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1608
1609 ie_entries = ie_result['entries']
1610 msg = (
1611 'Downloading %d videos' if not isinstance(ie_entries, list)
1612 else 'Collected %d videos; downloading %%d of them' % len(ie_entries))
1613
1614 if isinstance(ie_entries, list):
1615 def get_entry(i):
1616 return ie_entries[i - 1]
1617 else:
1618 if not isinstance(ie_entries, (PagedList, LazyList)):
1619 ie_entries = LazyList(ie_entries)
1620
1621 def get_entry(i):
1622 return YoutubeDL.__handle_extraction_exceptions(
1623 lambda self, i: ie_entries[i - 1]
1624 )(self, i)
1625
1626 entries = []
1627 items = playlistitems if playlistitems is not None else itertools.count(playliststart)
1628 for i in items:
1629 if i == 0:
1630 continue
1631 if playlistitems is None and playlistend is not None and playlistend < i:
1632 break
1633 entry = None
1634 try:
1635 entry = get_entry(i)
1636 if entry is MissingEntry:
1637 raise EntryNotInPlaylist()
1638 except (IndexError, EntryNotInPlaylist):
1639 if incomplete_entries:
1640 raise EntryNotInPlaylist(f'Entry {i} cannot be found')
1641 elif not playlistitems:
1642 break
1643 entries.append(entry)
1644 try:
1645 if entry is not None:
1646 self._match_entry(entry, incomplete=True, silent=True)
1647 except (ExistingVideoReached, RejectedVideoReached):
1648 break
1649 ie_result['entries'] = entries
1650
1651 # Save playlist_index before re-ordering
1652 entries = [
1653 ((playlistitems[i - 1] if playlistitems else i + playliststart - 1), entry)
1654 for i, entry in enumerate(entries, 1)
1655 if entry is not None]
1656 n_entries = len(entries)
1657
1658 if not playlistitems and (playliststart != 1 or playlistend):
1659 playlistitems = list(range(playliststart, playliststart + n_entries))
1660 ie_result['requested_entries'] = playlistitems
1661
1662 _infojson_written = False
1663 if not self.params.get('simulate') and self.params.get('allow_playlist_files', True):
1664 ie_copy = {
1665 'playlist': playlist,
1666 'playlist_id': ie_result.get('id'),
1667 'playlist_title': ie_result.get('title'),
1668 'playlist_uploader': ie_result.get('uploader'),
1669 'playlist_uploader_id': ie_result.get('uploader_id'),
1670 'playlist_index': 0,
1671 'n_entries': n_entries,
1672 }
1673 ie_copy.update(dict(ie_result))
1674
1675 _infojson_written = self._write_info_json(
1676 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1677 if _infojson_written is None:
1678 return
1679 if self._write_description('playlist', ie_result,
1680 self.prepare_filename(ie_copy, 'pl_description')) is None:
1681 return
1682 # TODO: This should be passed to ThumbnailsConvertor if necessary
1683 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1684
1685 if self.params.get('playlistreverse', False):
1686 entries = entries[::-1]
1687 if self.params.get('playlistrandom', False):
1688 random.shuffle(entries)
1689
1690 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1691
1692 self.to_screen('[%s] playlist %s: %s' % (ie_result['extractor'], playlist, msg % n_entries))
1693 failures = 0
1694 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1695 for i, entry_tuple in enumerate(entries, 1):
1696 playlist_index, entry = entry_tuple
1697 if 'playlist-index' in self.params.get('compat_opts', []):
1698 playlist_index = playlistitems[i - 1] if playlistitems else i + playliststart - 1
1699 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1700 # This __x_forwarded_for_ip thing is a bit ugly but requires
1701 # minimal changes
1702 if x_forwarded_for:
1703 entry['__x_forwarded_for_ip'] = x_forwarded_for
1704 extra = {
1705 'n_entries': n_entries,
1706 '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1707 'playlist_index': playlist_index,
1708 'playlist_autonumber': i,
1709 'playlist': playlist,
1710 'playlist_id': ie_result.get('id'),
1711 'playlist_title': ie_result.get('title'),
1712 'playlist_uploader': ie_result.get('uploader'),
1713 'playlist_uploader_id': ie_result.get('uploader_id'),
1714 'extractor': ie_result['extractor'],
1715 'webpage_url': ie_result['webpage_url'],
1716 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1717 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1718 'extractor_key': ie_result['extractor_key'],
1719 }
1720
1721 if self._match_entry(entry, incomplete=True) is not None:
1722 continue
1723
1724 entry_result = self.__process_iterable_entry(entry, download, extra)
1725 if not entry_result:
1726 failures += 1
1727 if failures >= max_failures:
1728 self.report_error(
1729 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1730 break
1731 playlist_results.append(entry_result)
1732 ie_result['entries'] = playlist_results
1733
1734 # Write the updated info to json
1735 if _infojson_written and self._write_info_json(
1736 'updated playlist', ie_result,
1737 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1738 return
1739 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1740 return ie_result
1741
1742 @__handle_extraction_exceptions
1743 def __process_iterable_entry(self, entry, download, extra_info):
1744 return self.process_ie_result(
1745 entry, download=download, extra_info=extra_info)
1746
1747 def _build_format_filter(self, filter_spec):
1748 " Returns a function to filter the formats according to the filter_spec "
1749
1750 OPERATORS = {
1751 '<': operator.lt,
1752 '<=': operator.le,
1753 '>': operator.gt,
1754 '>=': operator.ge,
1755 '=': operator.eq,
1756 '!=': operator.ne,
1757 }
1758 operator_rex = re.compile(r'''(?x)\s*
1759 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1760 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1761 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1762 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1763 m = operator_rex.fullmatch(filter_spec)
1764 if m:
1765 try:
1766 comparison_value = int(m.group('value'))
1767 except ValueError:
1768 comparison_value = parse_filesize(m.group('value'))
1769 if comparison_value is None:
1770 comparison_value = parse_filesize(m.group('value') + 'B')
1771 if comparison_value is None:
1772 raise ValueError(
1773 'Invalid value %r in format specification %r' % (
1774 m.group('value'), filter_spec))
1775 op = OPERATORS[m.group('op')]
1776
1777 if not m:
1778 STR_OPERATORS = {
1779 '=': operator.eq,
1780 '^=': lambda attr, value: attr.startswith(value),
1781 '$=': lambda attr, value: attr.endswith(value),
1782 '*=': lambda attr, value: value in attr,
1783 }
1784 str_operator_rex = re.compile(r'''(?x)\s*
1785 (?P<key>[a-zA-Z0-9._-]+)\s*
1786 (?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1787 (?P<value>[a-zA-Z0-9._-]+)\s*
1788 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1789 m = str_operator_rex.fullmatch(filter_spec)
1790 if m:
1791 comparison_value = m.group('value')
1792 str_op = STR_OPERATORS[m.group('op')]
1793 if m.group('negation'):
1794 op = lambda attr, value: not str_op(attr, value)
1795 else:
1796 op = str_op
1797
1798 if not m:
1799 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1800
1801 def _filter(f):
1802 actual_value = f.get(m.group('key'))
1803 if actual_value is None:
1804 return m.group('none_inclusive')
1805 return op(actual_value, comparison_value)
1806 return _filter
1807
1808 def _check_formats(self, formats):
1809 for f in formats:
1810 self.to_screen('[info] Testing format %s' % f['format_id'])
1811 path = self.get_output_path('temp')
1812 if not self._ensure_dir_exists(f'{path}/'):
1813 continue
1814 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1815 temp_file.close()
1816 try:
1817 success, _ = self.dl(temp_file.name, f, test=True)
1818 except (DownloadError, IOError, OSError, ValueError) + network_exceptions:
1819 success = False
1820 finally:
1821 if os.path.exists(temp_file.name):
1822 try:
1823 os.remove(temp_file.name)
1824 except OSError:
1825 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1826 if success:
1827 yield f
1828 else:
1829 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1830
1831 def _default_format_spec(self, info_dict, download=True):
1832
1833 def can_merge():
1834 merger = FFmpegMergerPP(self)
1835 return merger.available and merger.can_merge()
1836
1837 prefer_best = (
1838 not self.params.get('simulate')
1839 and download
1840 and (
1841 not can_merge()
1842 or info_dict.get('is_live', False)
1843 or self.outtmpl_dict['default'] == '-'))
1844 compat = (
1845 prefer_best
1846 or self.params.get('allow_multiple_audio_streams', False)
1847 or 'format-spec' in self.params.get('compat_opts', []))
1848
1849 return (
1850 'best/bestvideo+bestaudio' if prefer_best
1851 else 'bestvideo*+bestaudio/best' if not compat
1852 else 'bestvideo+bestaudio/best')
1853
1854 def build_format_selector(self, format_spec):
1855 def syntax_error(note, start):
1856 message = (
1857 'Invalid format specification: '
1858 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1859 return SyntaxError(message)
1860
1861 PICKFIRST = 'PICKFIRST'
1862 MERGE = 'MERGE'
1863 SINGLE = 'SINGLE'
1864 GROUP = 'GROUP'
1865 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1866
1867 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1868 'video': self.params.get('allow_multiple_video_streams', False)}
1869
1870 check_formats = self.params.get('check_formats') == 'selected'
1871
1872 def _parse_filter(tokens):
1873 filter_parts = []
1874 for type, string, start, _, _ in tokens:
1875 if type == tokenize.OP and string == ']':
1876 return ''.join(filter_parts)
1877 else:
1878 filter_parts.append(string)
1879
1880 def _remove_unused_ops(tokens):
1881 # Remove operators that we don't use and join them with the surrounding strings
1882 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1883 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1884 last_string, last_start, last_end, last_line = None, None, None, None
1885 for type, string, start, end, line in tokens:
1886 if type == tokenize.OP and string == '[':
1887 if last_string:
1888 yield tokenize.NAME, last_string, last_start, last_end, last_line
1889 last_string = None
1890 yield type, string, start, end, line
1891 # everything inside brackets will be handled by _parse_filter
1892 for type, string, start, end, line in tokens:
1893 yield type, string, start, end, line
1894 if type == tokenize.OP and string == ']':
1895 break
1896 elif type == tokenize.OP and string in ALLOWED_OPS:
1897 if last_string:
1898 yield tokenize.NAME, last_string, last_start, last_end, last_line
1899 last_string = None
1900 yield type, string, start, end, line
1901 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1902 if not last_string:
1903 last_string = string
1904 last_start = start
1905 last_end = end
1906 else:
1907 last_string += string
1908 if last_string:
1909 yield tokenize.NAME, last_string, last_start, last_end, last_line
1910
1911 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1912 selectors = []
1913 current_selector = None
1914 for type, string, start, _, _ in tokens:
1915 # ENCODING is only defined in python 3.x
1916 if type == getattr(tokenize, 'ENCODING', None):
1917 continue
1918 elif type in [tokenize.NAME, tokenize.NUMBER]:
1919 current_selector = FormatSelector(SINGLE, string, [])
1920 elif type == tokenize.OP:
1921 if string == ')':
1922 if not inside_group:
1923 # ')' will be handled by the parentheses group
1924 tokens.restore_last_token()
1925 break
1926 elif inside_merge and string in ['/', ',']:
1927 tokens.restore_last_token()
1928 break
1929 elif inside_choice and string == ',':
1930 tokens.restore_last_token()
1931 break
1932 elif string == ',':
1933 if not current_selector:
1934 raise syntax_error('"," must follow a format selector', start)
1935 selectors.append(current_selector)
1936 current_selector = None
1937 elif string == '/':
1938 if not current_selector:
1939 raise syntax_error('"/" must follow a format selector', start)
1940 first_choice = current_selector
1941 second_choice = _parse_format_selection(tokens, inside_choice=True)
1942 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1943 elif string == '[':
1944 if not current_selector:
1945 current_selector = FormatSelector(SINGLE, 'best', [])
1946 format_filter = _parse_filter(tokens)
1947 current_selector.filters.append(format_filter)
1948 elif string == '(':
1949 if current_selector:
1950 raise syntax_error('Unexpected "("', start)
1951 group = _parse_format_selection(tokens, inside_group=True)
1952 current_selector = FormatSelector(GROUP, group, [])
1953 elif string == '+':
1954 if not current_selector:
1955 raise syntax_error('Unexpected "+"', start)
1956 selector_1 = current_selector
1957 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1958 if not selector_2:
1959 raise syntax_error('Expected a selector', start)
1960 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1961 else:
1962 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1963 elif type == tokenize.ENDMARKER:
1964 break
1965 if current_selector:
1966 selectors.append(current_selector)
1967 return selectors
1968
1969 def _merge(formats_pair):
1970 format_1, format_2 = formats_pair
1971
1972 formats_info = []
1973 formats_info.extend(format_1.get('requested_formats', (format_1,)))
1974 formats_info.extend(format_2.get('requested_formats', (format_2,)))
1975
1976 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
1977 get_no_more = {'video': False, 'audio': False}
1978 for (i, fmt_info) in enumerate(formats_info):
1979 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
1980 formats_info.pop(i)
1981 continue
1982 for aud_vid in ['audio', 'video']:
1983 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
1984 if get_no_more[aud_vid]:
1985 formats_info.pop(i)
1986 break
1987 get_no_more[aud_vid] = True
1988
1989 if len(formats_info) == 1:
1990 return formats_info[0]
1991
1992 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
1993 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
1994
1995 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
1996 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
1997
1998 output_ext = self.params.get('merge_output_format')
1999 if not output_ext:
2000 if the_only_video:
2001 output_ext = the_only_video['ext']
2002 elif the_only_audio and not video_fmts:
2003 output_ext = the_only_audio['ext']
2004 else:
2005 output_ext = 'mkv'
2006
2007 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2008
2009 new_dict = {
2010 'requested_formats': formats_info,
2011 'format': '+'.join(filtered('format')),
2012 'format_id': '+'.join(filtered('format_id')),
2013 'ext': output_ext,
2014 'protocol': '+'.join(map(determine_protocol, formats_info)),
2015 'language': '+'.join(orderedSet(filtered('language'))) or None,
2016 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2017 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2018 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2019 }
2020
2021 if the_only_video:
2022 new_dict.update({
2023 'width': the_only_video.get('width'),
2024 'height': the_only_video.get('height'),
2025 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2026 'fps': the_only_video.get('fps'),
2027 'dynamic_range': the_only_video.get('dynamic_range'),
2028 'vcodec': the_only_video.get('vcodec'),
2029 'vbr': the_only_video.get('vbr'),
2030 'stretched_ratio': the_only_video.get('stretched_ratio'),
2031 })
2032
2033 if the_only_audio:
2034 new_dict.update({
2035 'acodec': the_only_audio.get('acodec'),
2036 'abr': the_only_audio.get('abr'),
2037 'asr': the_only_audio.get('asr'),
2038 })
2039
2040 return new_dict
2041
2042 def _check_formats(formats):
2043 if not check_formats:
2044 yield from formats
2045 return
2046 yield from self._check_formats(formats)
2047
2048 def _build_selector_function(selector):
2049 if isinstance(selector, list): # ,
2050 fs = [_build_selector_function(s) for s in selector]
2051
2052 def selector_function(ctx):
2053 for f in fs:
2054 yield from f(ctx)
2055 return selector_function
2056
2057 elif selector.type == GROUP: # ()
2058 selector_function = _build_selector_function(selector.selector)
2059
2060 elif selector.type == PICKFIRST: # /
2061 fs = [_build_selector_function(s) for s in selector.selector]
2062
2063 def selector_function(ctx):
2064 for f in fs:
2065 picked_formats = list(f(ctx))
2066 if picked_formats:
2067 return picked_formats
2068 return []
2069
2070 elif selector.type == MERGE: # +
2071 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2072
2073 def selector_function(ctx):
2074 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2075 yield _merge(pair)
2076
2077 elif selector.type == SINGLE: # atom
2078 format_spec = selector.selector or 'best'
2079
2080 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2081 if format_spec == 'all':
2082 def selector_function(ctx):
2083 yield from _check_formats(ctx['formats'][::-1])
2084 elif format_spec == 'mergeall':
2085 def selector_function(ctx):
2086 formats = list(_check_formats(ctx['formats']))
2087 if not formats:
2088 return
2089 merged_format = formats[-1]
2090 for f in formats[-2::-1]:
2091 merged_format = _merge((merged_format, f))
2092 yield merged_format
2093
2094 else:
2095 format_fallback, format_reverse, format_idx = False, True, 1
2096 mobj = re.match(
2097 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2098 format_spec)
2099 if mobj is not None:
2100 format_idx = int_or_none(mobj.group('n'), default=1)
2101 format_reverse = mobj.group('bw')[0] == 'b'
2102 format_type = (mobj.group('type') or [None])[0]
2103 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2104 format_modified = mobj.group('mod') is not None
2105
2106 format_fallback = not format_type and not format_modified # for b, w
2107 _filter_f = (
2108 (lambda f: f.get('%scodec' % format_type) != 'none')
2109 if format_type and format_modified # bv*, ba*, wv*, wa*
2110 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2111 if format_type # bv, ba, wv, wa
2112 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2113 if not format_modified # b, w
2114 else lambda f: True) # b*, w*
2115 filter_f = lambda f: _filter_f(f) and (
2116 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2117 else:
2118 if format_spec in self._format_selection_exts['audio']:
2119 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2120 elif format_spec in self._format_selection_exts['video']:
2121 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2122 elif format_spec in self._format_selection_exts['storyboards']:
2123 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2124 else:
2125 filter_f = lambda f: f.get('format_id') == format_spec # id
2126
2127 def selector_function(ctx):
2128 formats = list(ctx['formats'])
2129 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2130 if format_fallback and ctx['incomplete_formats'] and not matches:
2131 # for extractors with incomplete formats (audio only (soundcloud)
2132 # or video only (imgur)) best/worst will fallback to
2133 # best/worst {video,audio}-only format
2134 matches = formats
2135 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2136 try:
2137 yield matches[format_idx - 1]
2138 except IndexError:
2139 return
2140
2141 filters = [self._build_format_filter(f) for f in selector.filters]
2142
2143 def final_selector(ctx):
2144 ctx_copy = dict(ctx)
2145 for _filter in filters:
2146 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2147 return selector_function(ctx_copy)
2148 return final_selector
2149
2150 stream = io.BytesIO(format_spec.encode('utf-8'))
2151 try:
2152 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
2153 except tokenize.TokenError:
2154 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2155
2156 class TokenIterator(object):
2157 def __init__(self, tokens):
2158 self.tokens = tokens
2159 self.counter = 0
2160
2161 def __iter__(self):
2162 return self
2163
2164 def __next__(self):
2165 if self.counter >= len(self.tokens):
2166 raise StopIteration()
2167 value = self.tokens[self.counter]
2168 self.counter += 1
2169 return value
2170
2171 next = __next__
2172
2173 def restore_last_token(self):
2174 self.counter -= 1
2175
2176 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2177 return _build_selector_function(parsed_selector)
2178
2179 def _calc_headers(self, info_dict):
2180 res = std_headers.copy()
2181
2182 add_headers = info_dict.get('http_headers')
2183 if add_headers:
2184 res.update(add_headers)
2185
2186 cookies = self._calc_cookies(info_dict)
2187 if cookies:
2188 res['Cookie'] = cookies
2189
2190 if 'X-Forwarded-For' not in res:
2191 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2192 if x_forwarded_for_ip:
2193 res['X-Forwarded-For'] = x_forwarded_for_ip
2194
2195 return res
2196
2197 def _calc_cookies(self, info_dict):
2198 pr = sanitized_Request(info_dict['url'])
2199 self.cookiejar.add_cookie_header(pr)
2200 return pr.get_header('Cookie')
2201
2202 def _sort_thumbnails(self, thumbnails):
2203 thumbnails.sort(key=lambda t: (
2204 t.get('preference') if t.get('preference') is not None else -1,
2205 t.get('width') if t.get('width') is not None else -1,
2206 t.get('height') if t.get('height') is not None else -1,
2207 t.get('id') if t.get('id') is not None else '',
2208 t.get('url')))
2209
2210 def _sanitize_thumbnails(self, info_dict):
2211 thumbnails = info_dict.get('thumbnails')
2212 if thumbnails is None:
2213 thumbnail = info_dict.get('thumbnail')
2214 if thumbnail:
2215 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2216 if not thumbnails:
2217 return
2218
2219 def check_thumbnails(thumbnails):
2220 for t in thumbnails:
2221 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2222 try:
2223 self.urlopen(HEADRequest(t['url']))
2224 except network_exceptions as err:
2225 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2226 continue
2227 yield t
2228
2229 self._sort_thumbnails(thumbnails)
2230 for i, t in enumerate(thumbnails):
2231 if t.get('id') is None:
2232 t['id'] = '%d' % i
2233 if t.get('width') and t.get('height'):
2234 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2235 t['url'] = sanitize_url(t['url'])
2236
2237 if self.params.get('check_formats') is True:
2238 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2239 else:
2240 info_dict['thumbnails'] = thumbnails
2241
2242 def process_video_result(self, info_dict, download=True):
2243 assert info_dict.get('_type', 'video') == 'video'
2244
2245 if 'id' not in info_dict:
2246 raise ExtractorError('Missing "id" field in extractor result')
2247 if 'title' not in info_dict:
2248 raise ExtractorError('Missing "title" field in extractor result',
2249 video_id=info_dict['id'], ie=info_dict['extractor'])
2250
2251 def report_force_conversion(field, field_not, conversion):
2252 self.report_warning(
2253 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2254 % (field, field_not, conversion))
2255
2256 def sanitize_string_field(info, string_field):
2257 field = info.get(string_field)
2258 if field is None or isinstance(field, compat_str):
2259 return
2260 report_force_conversion(string_field, 'a string', 'string')
2261 info[string_field] = compat_str(field)
2262
2263 def sanitize_numeric_fields(info):
2264 for numeric_field in self._NUMERIC_FIELDS:
2265 field = info.get(numeric_field)
2266 if field is None or isinstance(field, compat_numeric_types):
2267 continue
2268 report_force_conversion(numeric_field, 'numeric', 'int')
2269 info[numeric_field] = int_or_none(field)
2270
2271 sanitize_string_field(info_dict, 'id')
2272 sanitize_numeric_fields(info_dict)
2273
2274 if 'playlist' not in info_dict:
2275 # It isn't part of a playlist
2276 info_dict['playlist'] = None
2277 info_dict['playlist_index'] = None
2278
2279 self._sanitize_thumbnails(info_dict)
2280
2281 thumbnail = info_dict.get('thumbnail')
2282 thumbnails = info_dict.get('thumbnails')
2283 if thumbnail:
2284 info_dict['thumbnail'] = sanitize_url(thumbnail)
2285 elif thumbnails:
2286 info_dict['thumbnail'] = thumbnails[-1]['url']
2287
2288 if info_dict.get('display_id') is None and 'id' in info_dict:
2289 info_dict['display_id'] = info_dict['id']
2290
2291 if info_dict.get('duration') is not None:
2292 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2293
2294 for ts_key, date_key in (
2295 ('timestamp', 'upload_date'),
2296 ('release_timestamp', 'release_date'),
2297 ):
2298 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2299 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2300 # see http://bugs.python.org/issue1646728)
2301 try:
2302 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2303 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2304 except (ValueError, OverflowError, OSError):
2305 pass
2306
2307 live_keys = ('is_live', 'was_live')
2308 live_status = info_dict.get('live_status')
2309 if live_status is None:
2310 for key in live_keys:
2311 if info_dict.get(key) is False:
2312 continue
2313 if info_dict.get(key):
2314 live_status = key
2315 break
2316 if all(info_dict.get(key) is False for key in live_keys):
2317 live_status = 'not_live'
2318 if live_status:
2319 info_dict['live_status'] = live_status
2320 for key in live_keys:
2321 if info_dict.get(key) is None:
2322 info_dict[key] = (live_status == key)
2323
2324 # Auto generate title fields corresponding to the *_number fields when missing
2325 # in order to always have clean titles. This is very common for TV series.
2326 for field in ('chapter', 'season', 'episode'):
2327 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2328 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2329
2330 for cc_kind in ('subtitles', 'automatic_captions'):
2331 cc = info_dict.get(cc_kind)
2332 if cc:
2333 for _, subtitle in cc.items():
2334 for subtitle_format in subtitle:
2335 if subtitle_format.get('url'):
2336 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2337 if subtitle_format.get('ext') is None:
2338 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2339
2340 automatic_captions = info_dict.get('automatic_captions')
2341 subtitles = info_dict.get('subtitles')
2342
2343 info_dict['requested_subtitles'] = self.process_subtitles(
2344 info_dict['id'], subtitles, automatic_captions)
2345
2346 if info_dict.get('formats') is None:
2347 # There's only one format available
2348 formats = [info_dict]
2349 else:
2350 formats = info_dict['formats']
2351
2352 info_dict['__has_drm'] = any(f.get('has_drm') for f in formats)
2353 if not self.params.get('allow_unplayable_formats'):
2354 formats = [f for f in formats if not f.get('has_drm')]
2355
2356 if info_dict.get('is_live'):
2357 get_from_start = bool(self.params.get('live_from_start'))
2358 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2359
2360 if not formats:
2361 self.raise_no_formats(info_dict)
2362
2363 def is_wellformed(f):
2364 url = f.get('url')
2365 if not url:
2366 self.report_warning(
2367 '"url" field is missing or empty - skipping format, '
2368 'there is an error in extractor')
2369 return False
2370 if isinstance(url, bytes):
2371 sanitize_string_field(f, 'url')
2372 return True
2373
2374 # Filter out malformed formats for better extraction robustness
2375 formats = list(filter(is_wellformed, formats))
2376
2377 formats_dict = {}
2378
2379 # We check that all the formats have the format and format_id fields
2380 for i, format in enumerate(formats):
2381 sanitize_string_field(format, 'format_id')
2382 sanitize_numeric_fields(format)
2383 format['url'] = sanitize_url(format['url'])
2384 if not format.get('format_id'):
2385 format['format_id'] = compat_str(i)
2386 else:
2387 # Sanitize format_id from characters used in format selector expression
2388 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2389 format_id = format['format_id']
2390 if format_id not in formats_dict:
2391 formats_dict[format_id] = []
2392 formats_dict[format_id].append(format)
2393
2394 # Make sure all formats have unique format_id
2395 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2396 for format_id, ambiguous_formats in formats_dict.items():
2397 ambigious_id = len(ambiguous_formats) > 1
2398 for i, format in enumerate(ambiguous_formats):
2399 if ambigious_id:
2400 format['format_id'] = '%s-%d' % (format_id, i)
2401 if format.get('ext') is None:
2402 format['ext'] = determine_ext(format['url']).lower()
2403 # Ensure there is no conflict between id and ext in format selection
2404 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2405 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2406 format['format_id'] = 'f%s' % format['format_id']
2407
2408 for i, format in enumerate(formats):
2409 if format.get('format') is None:
2410 format['format'] = '{id} - {res}{note}'.format(
2411 id=format['format_id'],
2412 res=self.format_resolution(format),
2413 note=format_field(format, 'format_note', ' (%s)'),
2414 )
2415 if format.get('protocol') is None:
2416 format['protocol'] = determine_protocol(format)
2417 if format.get('resolution') is None:
2418 format['resolution'] = self.format_resolution(format, default=None)
2419 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2420 format['dynamic_range'] = 'SDR'
2421 if (info_dict.get('duration') and format.get('tbr')
2422 and not format.get('filesize') and not format.get('filesize_approx')):
2423 format['filesize_approx'] = info_dict['duration'] * format['tbr'] * (1024 / 8)
2424
2425 # Add HTTP headers, so that external programs can use them from the
2426 # json output
2427 full_format_info = info_dict.copy()
2428 full_format_info.update(format)
2429 format['http_headers'] = self._calc_headers(full_format_info)
2430 # Remove private housekeeping stuff
2431 if '__x_forwarded_for_ip' in info_dict:
2432 del info_dict['__x_forwarded_for_ip']
2433
2434 # TODO Central sorting goes here
2435
2436 if self.params.get('check_formats') is True:
2437 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2438
2439 if not formats or formats[0] is not info_dict:
2440 # only set the 'formats' fields if the original info_dict list them
2441 # otherwise we end up with a circular reference, the first (and unique)
2442 # element in the 'formats' field in info_dict is info_dict itself,
2443 # which can't be exported to json
2444 info_dict['formats'] = formats
2445
2446 info_dict, _ = self.pre_process(info_dict)
2447
2448 # The pre-processors may have modified the formats
2449 formats = info_dict.get('formats', [info_dict])
2450
2451 if self.params.get('list_thumbnails'):
2452 self.list_thumbnails(info_dict)
2453 if self.params.get('listformats'):
2454 if not info_dict.get('formats') and not info_dict.get('url'):
2455 self.to_screen('%s has no formats' % info_dict['id'])
2456 else:
2457 self.list_formats(info_dict)
2458 if self.params.get('listsubtitles'):
2459 if 'automatic_captions' in info_dict:
2460 self.list_subtitles(
2461 info_dict['id'], automatic_captions, 'automatic captions')
2462 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2463 list_only = self.params.get('simulate') is None and (
2464 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2465 if list_only:
2466 # Without this printing, -F --print-json will not work
2467 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2468 return
2469
2470 format_selector = self.format_selector
2471 if format_selector is None:
2472 req_format = self._default_format_spec(info_dict, download=download)
2473 self.write_debug('Default format spec: %s' % req_format)
2474 format_selector = self.build_format_selector(req_format)
2475
2476 # While in format selection we may need to have an access to the original
2477 # format set in order to calculate some metrics or do some processing.
2478 # For now we need to be able to guess whether original formats provided
2479 # by extractor are incomplete or not (i.e. whether extractor provides only
2480 # video-only or audio-only formats) for proper formats selection for
2481 # extractors with such incomplete formats (see
2482 # https://github.com/ytdl-org/youtube-dl/pull/5556).
2483 # Since formats may be filtered during format selection and may not match
2484 # the original formats the results may be incorrect. Thus original formats
2485 # or pre-calculated metrics should be passed to format selection routines
2486 # as well.
2487 # We will pass a context object containing all necessary additional data
2488 # instead of just formats.
2489 # This fixes incorrect format selection issue (see
2490 # https://github.com/ytdl-org/youtube-dl/issues/10083).
2491 incomplete_formats = (
2492 # All formats are video-only or
2493 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2494 # all formats are audio-only
2495 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
2496
2497 ctx = {
2498 'formats': formats,
2499 'incomplete_formats': incomplete_formats,
2500 }
2501
2502 formats_to_download = list(format_selector(ctx))
2503 if not formats_to_download:
2504 if not self.params.get('ignore_no_formats_error'):
2505 raise ExtractorError('Requested format is not available', expected=True,
2506 video_id=info_dict['id'], ie=info_dict['extractor'])
2507 else:
2508 self.report_warning('Requested format is not available')
2509 # Process what we can, even without any available formats.
2510 self.process_info(dict(info_dict))
2511 elif download:
2512 self.to_screen(
2513 '[info] %s: Downloading %d format(s): %s' % (
2514 info_dict['id'], len(formats_to_download),
2515 ", ".join([f['format_id'] for f in formats_to_download])))
2516 for fmt in formats_to_download:
2517 new_info = dict(info_dict)
2518 # Save a reference to the original info_dict so that it can be modified in process_info if needed
2519 new_info['__original_infodict'] = info_dict
2520 new_info.update(fmt)
2521 self.process_info(new_info)
2522 # We update the info dict with the selected best quality format (backwards compatibility)
2523 if formats_to_download:
2524 info_dict.update(formats_to_download[-1])
2525 return info_dict
2526
2527 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2528 """Select the requested subtitles and their format"""
2529 available_subs = {}
2530 if normal_subtitles and self.params.get('writesubtitles'):
2531 available_subs.update(normal_subtitles)
2532 if automatic_captions and self.params.get('writeautomaticsub'):
2533 for lang, cap_info in automatic_captions.items():
2534 if lang not in available_subs:
2535 available_subs[lang] = cap_info
2536
2537 if (not self.params.get('writesubtitles') and not
2538 self.params.get('writeautomaticsub') or not
2539 available_subs):
2540 return None
2541
2542 all_sub_langs = available_subs.keys()
2543 if self.params.get('allsubtitles', False):
2544 requested_langs = all_sub_langs
2545 elif self.params.get('subtitleslangs', False):
2546 # A list is used so that the order of languages will be the same as
2547 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2548 requested_langs = []
2549 for lang_re in self.params.get('subtitleslangs'):
2550 if lang_re == 'all':
2551 requested_langs.extend(all_sub_langs)
2552 continue
2553 discard = lang_re[0] == '-'
2554 if discard:
2555 lang_re = lang_re[1:]
2556 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2557 if discard:
2558 for lang in current_langs:
2559 while lang in requested_langs:
2560 requested_langs.remove(lang)
2561 else:
2562 requested_langs.extend(current_langs)
2563 requested_langs = orderedSet(requested_langs)
2564 elif 'en' in available_subs:
2565 requested_langs = ['en']
2566 else:
2567 requested_langs = [list(all_sub_langs)[0]]
2568 if requested_langs:
2569 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2570
2571 formats_query = self.params.get('subtitlesformat', 'best')
2572 formats_preference = formats_query.split('/') if formats_query else []
2573 subs = {}
2574 for lang in requested_langs:
2575 formats = available_subs.get(lang)
2576 if formats is None:
2577 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
2578 continue
2579 for ext in formats_preference:
2580 if ext == 'best':
2581 f = formats[-1]
2582 break
2583 matches = list(filter(lambda f: f['ext'] == ext, formats))
2584 if matches:
2585 f = matches[-1]
2586 break
2587 else:
2588 f = formats[-1]
2589 self.report_warning(
2590 'No subtitle format found matching "%s" for language %s, '
2591 'using %s' % (formats_query, lang, f['ext']))
2592 subs[lang] = f
2593 return subs
2594
2595 def __forced_printings(self, info_dict, filename, incomplete):
2596 def print_mandatory(field, actual_field=None):
2597 if actual_field is None:
2598 actual_field = field
2599 if (self.params.get('force%s' % field, False)
2600 and (not incomplete or info_dict.get(actual_field) is not None)):
2601 self.to_stdout(info_dict[actual_field])
2602
2603 def print_optional(field):
2604 if (self.params.get('force%s' % field, False)
2605 and info_dict.get(field) is not None):
2606 self.to_stdout(info_dict[field])
2607
2608 info_dict = info_dict.copy()
2609 if filename is not None:
2610 info_dict['filename'] = filename
2611 if info_dict.get('requested_formats') is not None:
2612 # For RTMP URLs, also include the playpath
2613 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2614 elif 'url' in info_dict:
2615 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2616
2617 if self.params.get('forceprint') or self.params.get('forcejson'):
2618 self.post_extract(info_dict)
2619 for tmpl in self.params.get('forceprint', []):
2620 mobj = re.match(r'\w+(=?)$', tmpl)
2621 if mobj and mobj.group(1):
2622 tmpl = f'{tmpl[:-1]} = %({tmpl[:-1]})s'
2623 elif mobj:
2624 tmpl = '%({})s'.format(tmpl)
2625 self.to_stdout(self.evaluate_outtmpl(tmpl, info_dict))
2626
2627 print_mandatory('title')
2628 print_mandatory('id')
2629 print_mandatory('url', 'urls')
2630 print_optional('thumbnail')
2631 print_optional('description')
2632 print_optional('filename')
2633 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2634 self.to_stdout(formatSeconds(info_dict['duration']))
2635 print_mandatory('format')
2636
2637 if self.params.get('forcejson'):
2638 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2639
2640 def dl(self, name, info, subtitle=False, test=False):
2641 if not info.get('url'):
2642 self.raise_no_formats(info, True)
2643
2644 if test:
2645 verbose = self.params.get('verbose')
2646 params = {
2647 'test': True,
2648 'quiet': self.params.get('quiet') or not verbose,
2649 'verbose': verbose,
2650 'noprogress': not verbose,
2651 'nopart': True,
2652 'skip_unavailable_fragments': False,
2653 'keep_fragments': False,
2654 'overwrites': True,
2655 '_no_ytdl_file': True,
2656 }
2657 else:
2658 params = self.params
2659 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2660 if not test:
2661 for ph in self._progress_hooks:
2662 fd.add_progress_hook(ph)
2663 urls = '", "'.join([f['url'] for f in info.get('requested_formats', [])] or [info['url']])
2664 self.write_debug('Invoking downloader on "%s"' % urls)
2665
2666 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2667 # But it may contain objects that are not deep-copyable
2668 new_info = self._copy_infodict(info)
2669 if new_info.get('http_headers') is None:
2670 new_info['http_headers'] = self._calc_headers(new_info)
2671 return fd.download(name, new_info, subtitle)
2672
2673 def process_info(self, info_dict):
2674 """Process a single resolved IE result."""
2675
2676 assert info_dict.get('_type', 'video') == 'video'
2677
2678 max_downloads = self.params.get('max_downloads')
2679 if max_downloads is not None:
2680 if self._num_downloads >= int(max_downloads):
2681 raise MaxDownloadsReached()
2682
2683 if info_dict.get('is_live') and not self.params.get('live_from_start'):
2684 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2685
2686 # TODO: backward compatibility, to be removed
2687 info_dict['fulltitle'] = info_dict['title']
2688
2689 if 'format' not in info_dict and 'ext' in info_dict:
2690 info_dict['format'] = info_dict['ext']
2691
2692 if self._match_entry(info_dict) is not None:
2693 return
2694
2695 self.post_extract(info_dict)
2696 self._num_downloads += 1
2697
2698 # info_dict['_filename'] needs to be set for backward compatibility
2699 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2700 temp_filename = self.prepare_filename(info_dict, 'temp')
2701 files_to_move = {}
2702
2703 # Forced printings
2704 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2705
2706 if self.params.get('simulate'):
2707 if self.params.get('force_write_download_archive', False):
2708 self.record_download_archive(info_dict)
2709 # Do nothing else if in simulate mode
2710 return
2711
2712 if full_filename is None:
2713 return
2714 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2715 return
2716 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2717 return
2718
2719 if self._write_description('video', info_dict,
2720 self.prepare_filename(info_dict, 'description')) is None:
2721 return
2722
2723 sub_files = self._write_subtitles(info_dict, temp_filename)
2724 if sub_files is None:
2725 return
2726 files_to_move.update(dict(sub_files))
2727
2728 thumb_files = self._write_thumbnails(
2729 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2730 if thumb_files is None:
2731 return
2732 files_to_move.update(dict(thumb_files))
2733
2734 infofn = self.prepare_filename(info_dict, 'infojson')
2735 _infojson_written = self._write_info_json('video', info_dict, infofn)
2736 if _infojson_written:
2737 info_dict['infojson_filename'] = infofn
2738 # For backward compatibility, even though it was a private field
2739 info_dict['__infojson_filename'] = infofn
2740 elif _infojson_written is None:
2741 return
2742
2743 # Note: Annotations are deprecated
2744 annofn = None
2745 if self.params.get('writeannotations', False):
2746 annofn = self.prepare_filename(info_dict, 'annotation')
2747 if annofn:
2748 if not self._ensure_dir_exists(encodeFilename(annofn)):
2749 return
2750 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2751 self.to_screen('[info] Video annotations are already present')
2752 elif not info_dict.get('annotations'):
2753 self.report_warning('There are no annotations to write.')
2754 else:
2755 try:
2756 self.to_screen('[info] Writing video annotations to: ' + annofn)
2757 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2758 annofile.write(info_dict['annotations'])
2759 except (KeyError, TypeError):
2760 self.report_warning('There are no annotations to write.')
2761 except (OSError, IOError):
2762 self.report_error('Cannot write annotations file: ' + annofn)
2763 return
2764
2765 # Write internet shortcut files
2766 def _write_link_file(link_type):
2767 if 'webpage_url' not in info_dict:
2768 self.report_error('Cannot write internet shortcut file because the "webpage_url" field is missing in the media information')
2769 return False
2770 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2771 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2772 return False
2773 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2774 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2775 return True
2776 try:
2777 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2778 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2779 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2780 template_vars = {'url': iri_to_uri(info_dict['webpage_url'])}
2781 if link_type == 'desktop':
2782 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2783 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2784 except (OSError, IOError):
2785 self.report_error(f'Cannot write internet shortcut {linkfn}')
2786 return False
2787 return True
2788
2789 write_links = {
2790 'url': self.params.get('writeurllink'),
2791 'webloc': self.params.get('writewebloclink'),
2792 'desktop': self.params.get('writedesktoplink'),
2793 }
2794 if self.params.get('writelink'):
2795 link_type = ('webloc' if sys.platform == 'darwin'
2796 else 'desktop' if sys.platform.startswith('linux')
2797 else 'url')
2798 write_links[link_type] = True
2799
2800 if any(should_write and not _write_link_file(link_type)
2801 for link_type, should_write in write_links.items()):
2802 return
2803
2804 try:
2805 info_dict, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2806 except PostProcessingError as err:
2807 self.report_error('Preprocessing: %s' % str(err))
2808 return
2809
2810 must_record_download_archive = False
2811 if self.params.get('skip_download', False):
2812 info_dict['filepath'] = temp_filename
2813 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2814 info_dict['__files_to_move'] = files_to_move
2815 info_dict = self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict)
2816 else:
2817 # Download
2818 info_dict.setdefault('__postprocessors', [])
2819 try:
2820
2821 def existing_file(*filepaths):
2822 ext = info_dict.get('ext')
2823 final_ext = self.params.get('final_ext', ext)
2824 existing_files = []
2825 for file in orderedSet(filepaths):
2826 if final_ext != ext:
2827 converted = replace_extension(file, final_ext, ext)
2828 if os.path.exists(encodeFilename(converted)):
2829 existing_files.append(converted)
2830 if os.path.exists(encodeFilename(file)):
2831 existing_files.append(file)
2832
2833 if not existing_files or self.params.get('overwrites', False):
2834 for file in orderedSet(existing_files):
2835 self.report_file_delete(file)
2836 os.remove(encodeFilename(file))
2837 return None
2838
2839 info_dict['ext'] = os.path.splitext(existing_files[0])[1][1:]
2840 return existing_files[0]
2841
2842 success = True
2843 if info_dict.get('requested_formats') is not None:
2844
2845 def compatible_formats(formats):
2846 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2847 video_formats = [format for format in formats if format.get('vcodec') != 'none']
2848 audio_formats = [format for format in formats if format.get('acodec') != 'none']
2849 if len(video_formats) > 2 or len(audio_formats) > 2:
2850 return False
2851
2852 # Check extension
2853 exts = set(format.get('ext') for format in formats)
2854 COMPATIBLE_EXTS = (
2855 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
2856 set(('webm',)),
2857 )
2858 for ext_sets in COMPATIBLE_EXTS:
2859 if ext_sets.issuperset(exts):
2860 return True
2861 # TODO: Check acodec/vcodec
2862 return False
2863
2864 requested_formats = info_dict['requested_formats']
2865 old_ext = info_dict['ext']
2866 if self.params.get('merge_output_format') is None:
2867 if not compatible_formats(requested_formats):
2868 info_dict['ext'] = 'mkv'
2869 self.report_warning(
2870 'Requested formats are incompatible for merge and will be merged into mkv')
2871 if (info_dict['ext'] == 'webm'
2872 and info_dict.get('thumbnails')
2873 # check with type instead of pp_key, __name__, or isinstance
2874 # since we dont want any custom PPs to trigger this
2875 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])):
2876 info_dict['ext'] = 'mkv'
2877 self.report_warning(
2878 'webm doesn\'t support embedding a thumbnail, mkv will be used')
2879 new_ext = info_dict['ext']
2880
2881 def correct_ext(filename, ext=new_ext):
2882 if filename == '-':
2883 return filename
2884 filename_real_ext = os.path.splitext(filename)[1][1:]
2885 filename_wo_ext = (
2886 os.path.splitext(filename)[0]
2887 if filename_real_ext in (old_ext, new_ext)
2888 else filename)
2889 return '%s.%s' % (filename_wo_ext, ext)
2890
2891 # Ensure filename always has a correct extension for successful merge
2892 full_filename = correct_ext(full_filename)
2893 temp_filename = correct_ext(temp_filename)
2894 dl_filename = existing_file(full_filename, temp_filename)
2895 info_dict['__real_download'] = False
2896
2897 downloaded = []
2898 merger = FFmpegMergerPP(self)
2899
2900 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
2901 if dl_filename is not None:
2902 self.report_file_already_downloaded(dl_filename)
2903 elif fd:
2904 for f in requested_formats if fd != FFmpegFD else []:
2905 f['filepath'] = fname = prepend_extension(
2906 correct_ext(temp_filename, info_dict['ext']),
2907 'f%s' % f['format_id'], info_dict['ext'])
2908 downloaded.append(fname)
2909 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
2910 success, real_download = self.dl(temp_filename, info_dict)
2911 info_dict['__real_download'] = real_download
2912 else:
2913 if self.params.get('allow_unplayable_formats'):
2914 self.report_warning(
2915 'You have requested merging of multiple formats '
2916 'while also allowing unplayable formats to be downloaded. '
2917 'The formats won\'t be merged to prevent data corruption.')
2918 elif not merger.available:
2919 self.report_warning(
2920 'You have requested merging of multiple formats but ffmpeg is not installed. '
2921 'The formats won\'t be merged.')
2922
2923 if temp_filename == '-':
2924 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
2925 else 'but the formats are incompatible for simultaneous download' if merger.available
2926 else 'but ffmpeg is not installed')
2927 self.report_warning(
2928 f'You have requested downloading multiple formats to stdout {reason}. '
2929 'The formats will be streamed one after the other')
2930 fname = temp_filename
2931 for f in requested_formats:
2932 new_info = dict(info_dict)
2933 del new_info['requested_formats']
2934 new_info.update(f)
2935 if temp_filename != '-':
2936 fname = prepend_extension(
2937 correct_ext(temp_filename, new_info['ext']),
2938 'f%s' % f['format_id'], new_info['ext'])
2939 if not self._ensure_dir_exists(fname):
2940 return
2941 f['filepath'] = fname
2942 downloaded.append(fname)
2943 partial_success, real_download = self.dl(fname, new_info)
2944 info_dict['__real_download'] = info_dict['__real_download'] or real_download
2945 success = success and partial_success
2946
2947 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
2948 info_dict['__postprocessors'].append(merger)
2949 info_dict['__files_to_merge'] = downloaded
2950 # Even if there were no downloads, it is being merged only now
2951 info_dict['__real_download'] = True
2952 else:
2953 for file in downloaded:
2954 files_to_move[file] = None
2955 else:
2956 # Just a single file
2957 dl_filename = existing_file(full_filename, temp_filename)
2958 if dl_filename is None or dl_filename == temp_filename:
2959 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
2960 # So we should try to resume the download
2961 success, real_download = self.dl(temp_filename, info_dict)
2962 info_dict['__real_download'] = real_download
2963 else:
2964 self.report_file_already_downloaded(dl_filename)
2965
2966 dl_filename = dl_filename or temp_filename
2967 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2968
2969 except network_exceptions as err:
2970 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
2971 return
2972 except (OSError, IOError) as err:
2973 raise UnavailableVideoError(err)
2974 except (ContentTooShortError, ) as err:
2975 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
2976 return
2977
2978 if success and full_filename != '-':
2979
2980 def fixup():
2981 do_fixup = True
2982 fixup_policy = self.params.get('fixup')
2983 vid = info_dict['id']
2984
2985 if fixup_policy in ('ignore', 'never'):
2986 return
2987 elif fixup_policy == 'warn':
2988 do_fixup = False
2989 elif fixup_policy != 'force':
2990 assert fixup_policy in ('detect_or_warn', None)
2991 if not info_dict.get('__real_download'):
2992 do_fixup = False
2993
2994 def ffmpeg_fixup(cndn, msg, cls):
2995 if not cndn:
2996 return
2997 if not do_fixup:
2998 self.report_warning(f'{vid}: {msg}')
2999 return
3000 pp = cls(self)
3001 if pp.available:
3002 info_dict['__postprocessors'].append(pp)
3003 else:
3004 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3005
3006 stretched_ratio = info_dict.get('stretched_ratio')
3007 ffmpeg_fixup(
3008 stretched_ratio not in (1, None),
3009 f'Non-uniform pixel ratio {stretched_ratio}',
3010 FFmpegFixupStretchedPP)
3011
3012 ffmpeg_fixup(
3013 (info_dict.get('requested_formats') is None
3014 and info_dict.get('container') == 'm4a_dash'
3015 and info_dict.get('ext') == 'm4a'),
3016 'writing DASH m4a. Only some players support this container',
3017 FFmpegFixupM4aPP)
3018
3019 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3020 downloader = downloader.__name__ if downloader else None
3021
3022 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3023 ffmpeg_fixup(downloader == 'HlsFD',
3024 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3025 FFmpegFixupM3u8PP)
3026 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3027 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3028
3029 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3030 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed duration detected', FFmpegFixupDurationPP)
3031
3032 fixup()
3033 try:
3034 info_dict = self.post_process(dl_filename, info_dict, files_to_move)
3035 except PostProcessingError as err:
3036 self.report_error('Postprocessing: %s' % str(err))
3037 return
3038 try:
3039 for ph in self._post_hooks:
3040 ph(info_dict['filepath'])
3041 except Exception as err:
3042 self.report_error('post hooks: %s' % str(err))
3043 return
3044 must_record_download_archive = True
3045
3046 if must_record_download_archive or self.params.get('force_write_download_archive', False):
3047 self.record_download_archive(info_dict)
3048 max_downloads = self.params.get('max_downloads')
3049 if max_downloads is not None and self._num_downloads >= int(max_downloads):
3050 raise MaxDownloadsReached()
3051
3052 def __download_wrapper(self, func):
3053 @functools.wraps(func)
3054 def wrapper(*args, **kwargs):
3055 try:
3056 res = func(*args, **kwargs)
3057 except UnavailableVideoError as e:
3058 self.report_error(e)
3059 except MaxDownloadsReached as e:
3060 self.to_screen(f'[info] {e}')
3061 raise
3062 except DownloadCancelled as e:
3063 self.to_screen(f'[info] {e}')
3064 if not self.params.get('break_per_url'):
3065 raise
3066 else:
3067 if self.params.get('dump_single_json', False):
3068 self.post_extract(res)
3069 self.to_stdout(json.dumps(self.sanitize_info(res)))
3070 return wrapper
3071
3072 def download(self, url_list):
3073 """Download a given list of URLs."""
3074 url_list = variadic(url_list) # Passing a single URL is a common mistake
3075 outtmpl = self.outtmpl_dict['default']
3076 if (len(url_list) > 1
3077 and outtmpl != '-'
3078 and '%' not in outtmpl
3079 and self.params.get('max_downloads') != 1):
3080 raise SameFileError(outtmpl)
3081
3082 for url in url_list:
3083 self.__download_wrapper(self.extract_info)(
3084 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3085
3086 return self._download_retcode
3087
3088 def download_with_info_file(self, info_filename):
3089 with contextlib.closing(fileinput.FileInput(
3090 [info_filename], mode='r',
3091 openhook=fileinput.hook_encoded('utf-8'))) as f:
3092 # FileInput doesn't have a read method, we can't call json.load
3093 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3094 try:
3095 self.__download_wrapper(self.process_ie_result)(info, download=True)
3096 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3097 if not isinstance(e, EntryNotInPlaylist):
3098 self.to_stderr('\r')
3099 webpage_url = info.get('webpage_url')
3100 if webpage_url is not None:
3101 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3102 return self.download([webpage_url])
3103 else:
3104 raise
3105 return self._download_retcode
3106
3107 @staticmethod
3108 def sanitize_info(info_dict, remove_private_keys=False):
3109 ''' Sanitize the infodict for converting to json '''
3110 if info_dict is None:
3111 return info_dict
3112 info_dict.setdefault('epoch', int(time.time()))
3113 remove_keys = {'__original_infodict'} # Always remove this since this may contain a copy of the entire dict
3114 keep_keys = ['_type'] # Always keep this to facilitate load-info-json
3115 if remove_private_keys:
3116 remove_keys |= {
3117 'requested_formats', 'requested_subtitles', 'requested_entries', 'entries',
3118 'filepath', 'infojson_filename', 'original_url', 'playlist_autonumber',
3119 }
3120 empty_values = (None, {}, [], set(), tuple())
3121 reject = lambda k, v: k not in keep_keys and (
3122 k.startswith('_') or k in remove_keys or v in empty_values)
3123 else:
3124 reject = lambda k, v: k in remove_keys
3125
3126 def filter_fn(obj):
3127 if isinstance(obj, dict):
3128 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3129 elif isinstance(obj, (list, tuple, set, LazyList)):
3130 return list(map(filter_fn, obj))
3131 elif obj is None or isinstance(obj, (str, int, float, bool)):
3132 return obj
3133 else:
3134 return repr(obj)
3135
3136 return filter_fn(info_dict)
3137
3138 @staticmethod
3139 def filter_requested_info(info_dict, actually_filter=True):
3140 ''' Alias of sanitize_info for backward compatibility '''
3141 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3142
3143 def run_pp(self, pp, infodict):
3144 files_to_delete = []
3145 if '__files_to_move' not in infodict:
3146 infodict['__files_to_move'] = {}
3147 try:
3148 files_to_delete, infodict = pp.run(infodict)
3149 except PostProcessingError as e:
3150 # Must be True and not 'only_download'
3151 if self.params.get('ignoreerrors') is True:
3152 self.report_error(e)
3153 return infodict
3154 raise
3155
3156 if not files_to_delete:
3157 return infodict
3158 if self.params.get('keepvideo', False):
3159 for f in files_to_delete:
3160 infodict['__files_to_move'].setdefault(f, '')
3161 else:
3162 for old_filename in set(files_to_delete):
3163 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
3164 try:
3165 os.remove(encodeFilename(old_filename))
3166 except (IOError, OSError):
3167 self.report_warning('Unable to remove downloaded original file')
3168 if old_filename in infodict['__files_to_move']:
3169 del infodict['__files_to_move'][old_filename]
3170 return infodict
3171
3172 @staticmethod
3173 def post_extract(info_dict):
3174 def actual_post_extract(info_dict):
3175 if info_dict.get('_type') in ('playlist', 'multi_video'):
3176 for video_dict in info_dict.get('entries', {}):
3177 actual_post_extract(video_dict or {})
3178 return
3179
3180 post_extractor = info_dict.get('__post_extractor') or (lambda: {})
3181 extra = post_extractor().items()
3182 info_dict.update(extra)
3183 info_dict.pop('__post_extractor', None)
3184
3185 original_infodict = info_dict.get('__original_infodict') or {}
3186 original_infodict.update(extra)
3187 original_infodict.pop('__post_extractor', None)
3188
3189 actual_post_extract(info_dict or {})
3190
3191 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3192 info = dict(ie_info)
3193 info['__files_to_move'] = files_to_move or {}
3194 for pp in self._pps[key]:
3195 info = self.run_pp(pp, info)
3196 return info, info.pop('__files_to_move', None)
3197
3198 def post_process(self, filename, ie_info, files_to_move=None):
3199 """Run all the postprocessors on the given file."""
3200 info = dict(ie_info)
3201 info['filepath'] = filename
3202 info['__files_to_move'] = files_to_move or {}
3203
3204 for pp in ie_info.get('__postprocessors', []) + self._pps['post_process']:
3205 info = self.run_pp(pp, info)
3206 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3207 del info['__files_to_move']
3208 for pp in self._pps['after_move']:
3209 info = self.run_pp(pp, info)
3210 return info
3211
3212 def _make_archive_id(self, info_dict):
3213 video_id = info_dict.get('id')
3214 if not video_id:
3215 return
3216 # Future-proof against any change in case
3217 # and backwards compatibility with prior versions
3218 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3219 if extractor is None:
3220 url = str_or_none(info_dict.get('url'))
3221 if not url:
3222 return
3223 # Try to find matching extractor for the URL and take its ie_key
3224 for ie_key, ie in self._ies.items():
3225 if ie.suitable(url):
3226 extractor = ie_key
3227 break
3228 else:
3229 return
3230 return '%s %s' % (extractor.lower(), video_id)
3231
3232 def in_download_archive(self, info_dict):
3233 fn = self.params.get('download_archive')
3234 if fn is None:
3235 return False
3236
3237 vid_id = self._make_archive_id(info_dict)
3238 if not vid_id:
3239 return False # Incomplete video information
3240
3241 return vid_id in self.archive
3242
3243 def record_download_archive(self, info_dict):
3244 fn = self.params.get('download_archive')
3245 if fn is None:
3246 return
3247 vid_id = self._make_archive_id(info_dict)
3248 assert vid_id
3249 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3250 archive_file.write(vid_id + '\n')
3251 self.archive.add(vid_id)
3252
3253 @staticmethod
3254 def format_resolution(format, default='unknown'):
3255 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3256 return 'audio only'
3257 if format.get('resolution') is not None:
3258 return format['resolution']
3259 if format.get('width') and format.get('height'):
3260 return '%dx%d' % (format['width'], format['height'])
3261 elif format.get('height'):
3262 return '%sp' % format['height']
3263 elif format.get('width'):
3264 return '%dx?' % format['width']
3265 return default
3266
3267 def _format_note(self, fdict):
3268 res = ''
3269 if fdict.get('ext') in ['f4f', 'f4m']:
3270 res += '(unsupported)'
3271 if fdict.get('language'):
3272 if res:
3273 res += ' '
3274 res += '[%s]' % fdict['language']
3275 if fdict.get('format_note') is not None:
3276 if res:
3277 res += ' '
3278 res += fdict['format_note']
3279 if fdict.get('tbr') is not None:
3280 if res:
3281 res += ', '
3282 res += '%4dk' % fdict['tbr']
3283 if fdict.get('container') is not None:
3284 if res:
3285 res += ', '
3286 res += '%s container' % fdict['container']
3287 if (fdict.get('vcodec') is not None
3288 and fdict.get('vcodec') != 'none'):
3289 if res:
3290 res += ', '
3291 res += fdict['vcodec']
3292 if fdict.get('vbr') is not None:
3293 res += '@'
3294 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3295 res += 'video@'
3296 if fdict.get('vbr') is not None:
3297 res += '%4dk' % fdict['vbr']
3298 if fdict.get('fps') is not None:
3299 if res:
3300 res += ', '
3301 res += '%sfps' % fdict['fps']
3302 if fdict.get('acodec') is not None:
3303 if res:
3304 res += ', '
3305 if fdict['acodec'] == 'none':
3306 res += 'video only'
3307 else:
3308 res += '%-5s' % fdict['acodec']
3309 elif fdict.get('abr') is not None:
3310 if res:
3311 res += ', '
3312 res += 'audio'
3313 if fdict.get('abr') is not None:
3314 res += '@%3dk' % fdict['abr']
3315 if fdict.get('asr') is not None:
3316 res += ' (%5dHz)' % fdict['asr']
3317 if fdict.get('filesize') is not None:
3318 if res:
3319 res += ', '
3320 res += format_bytes(fdict['filesize'])
3321 elif fdict.get('filesize_approx') is not None:
3322 if res:
3323 res += ', '
3324 res += '~' + format_bytes(fdict['filesize_approx'])
3325 return res
3326
3327 def _list_format_headers(self, *headers):
3328 if self.params.get('listformats_table', True) is not False:
3329 return [self._format_screen(header, self.Styles.HEADERS) for header in headers]
3330 return headers
3331
3332 def list_formats(self, info_dict):
3333 formats = info_dict.get('formats', [info_dict])
3334 new_format = self.params.get('listformats_table', True) is not False
3335 if new_format:
3336 delim = self._format_screen('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3337 table = [
3338 [
3339 self._format_screen(format_field(f, 'format_id'), self.Styles.ID),
3340 format_field(f, 'ext'),
3341 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3342 format_field(f, 'fps', '\t%d'),
3343 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3344 delim,
3345 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3346 format_field(f, 'tbr', '\t%dk'),
3347 shorten_protocol_name(f.get('protocol', '').replace('native', 'n')),
3348 delim,
3349 format_field(f, 'vcodec', default='unknown').replace(
3350 'none',
3351 'images' if f.get('acodec') == 'none'
3352 else self._format_screen('audio only', self.Styles.SUPPRESS)),
3353 format_field(f, 'vbr', '\t%dk'),
3354 format_field(f, 'acodec', default='unknown').replace(
3355 'none',
3356 '' if f.get('vcodec') == 'none'
3357 else self._format_screen('video only', self.Styles.SUPPRESS)),
3358 format_field(f, 'abr', '\t%dk'),
3359 format_field(f, 'asr', '\t%dHz'),
3360 join_nonempty(
3361 self._format_screen('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3362 format_field(f, 'language', '[%s]'),
3363 join_nonempty(
3364 format_field(f, 'format_note'),
3365 format_field(f, 'container', ignore=(None, f.get('ext'))),
3366 delim=', '),
3367 delim=' '),
3368 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3369 header_line = self._list_format_headers(
3370 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3371 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3372 else:
3373 table = [
3374 [
3375 format_field(f, 'format_id'),
3376 format_field(f, 'ext'),
3377 self.format_resolution(f),
3378 self._format_note(f)]
3379 for f in formats
3380 if f.get('preference') is None or f['preference'] >= -1000]
3381 header_line = ['format code', 'extension', 'resolution', 'note']
3382
3383 self.to_screen(
3384 '[info] Available formats for %s:' % info_dict['id'])
3385 self.to_stdout(render_table(
3386 header_line, table,
3387 extra_gap=(0 if new_format else 1),
3388 hide_empty=new_format,
3389 delim=new_format and self._format_screen('\u2500', self.Styles.DELIM, '-', test_encoding=True)))
3390
3391 def list_thumbnails(self, info_dict):
3392 thumbnails = list(info_dict.get('thumbnails'))
3393 if not thumbnails:
3394 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
3395 return
3396
3397 self.to_screen(
3398 '[info] Thumbnails for %s:' % info_dict['id'])
3399 self.to_stdout(render_table(
3400 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3401 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
3402
3403 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3404 if not subtitles:
3405 self.to_screen('%s has no %s' % (video_id, name))
3406 return
3407 self.to_screen(
3408 'Available %s for %s:' % (name, video_id))
3409
3410 def _row(lang, formats):
3411 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3412 if len(set(names)) == 1:
3413 names = [] if names[0] == 'unknown' else names[:1]
3414 return [lang, ', '.join(names), ', '.join(exts)]
3415
3416 self.to_stdout(render_table(
3417 self._list_format_headers('Language', 'Name', 'Formats'),
3418 [_row(lang, formats) for lang, formats in subtitles.items()],
3419 hide_empty=True))
3420
3421 def urlopen(self, req):
3422 """ Start an HTTP download """
3423 if isinstance(req, compat_basestring):
3424 req = sanitized_Request(req)
3425 return self._opener.open(req, timeout=self._socket_timeout)
3426
3427 def print_debug_header(self):
3428 if not self.params.get('verbose'):
3429 return
3430
3431 def get_encoding(stream):
3432 ret = getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__)
3433 if not supports_terminal_sequences(stream):
3434 from .compat import WINDOWS_VT_MODE
3435 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3436 return ret
3437
3438 encoding_str = 'Encodings: locale %s, fs %s, out %s, err %s, pref %s' % (
3439 locale.getpreferredencoding(),
3440 sys.getfilesystemencoding(),
3441 get_encoding(self._screen_file), get_encoding(self._err_file),
3442 self.get_encoding())
3443
3444 logger = self.params.get('logger')
3445 if logger:
3446 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3447 write_debug(encoding_str)
3448 else:
3449 write_string(f'[debug] {encoding_str}\n', encoding=None)
3450 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3451
3452 source = detect_variant()
3453 write_debug(join_nonempty(
3454 'yt-dlp version', __version__,
3455 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3456 '' if source == 'unknown' else f'({source})',
3457 delim=' '))
3458 if not _LAZY_LOADER:
3459 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3460 write_debug('Lazy loading extractors is forcibly disabled')
3461 else:
3462 write_debug('Lazy loading extractors is disabled')
3463 if plugin_extractors or plugin_postprocessors:
3464 write_debug('Plugins: %s' % [
3465 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3466 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3467 if self.params.get('compat_opts'):
3468 write_debug('Compatibility options: %s' % ', '.join(self.params.get('compat_opts')))
3469
3470 if source == 'source':
3471 try:
3472 sp = Popen(
3473 ['git', 'rev-parse', '--short', 'HEAD'],
3474 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3475 cwd=os.path.dirname(os.path.abspath(__file__)))
3476 out, err = sp.communicate_or_kill()
3477 out = out.decode().strip()
3478 if re.match('[0-9a-f]+', out):
3479 write_debug('Git HEAD: %s' % out)
3480 except Exception:
3481 try:
3482 sys.exc_clear()
3483 except Exception:
3484 pass
3485
3486 def python_implementation():
3487 impl_name = platform.python_implementation()
3488 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3489 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3490 return impl_name
3491
3492 write_debug('Python version %s (%s %s) - %s' % (
3493 platform.python_version(),
3494 python_implementation(),
3495 platform.architecture()[0],
3496 platform_name()))
3497
3498 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3499 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3500 if ffmpeg_features:
3501 exe_versions['ffmpeg'] += ' (%s)' % ','.join(ffmpeg_features)
3502
3503 exe_versions['rtmpdump'] = rtmpdump_version()
3504 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3505 exe_str = ', '.join(
3506 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3507 ) or 'none'
3508 write_debug('exe versions: %s' % exe_str)
3509
3510 from .downloader.websocket import has_websockets
3511 from .postprocessor.embedthumbnail import has_mutagen
3512 from .cookies import SQLITE_AVAILABLE, KEYRING_AVAILABLE
3513
3514 lib_str = join_nonempty(
3515 compat_pycrypto_AES and compat_pycrypto_AES.__name__.split('.')[0],
3516 KEYRING_AVAILABLE and 'keyring',
3517 has_mutagen and 'mutagen',
3518 SQLITE_AVAILABLE and 'sqlite',
3519 has_websockets and 'websockets',
3520 delim=', ') or 'none'
3521 write_debug('Optional libraries: %s' % lib_str)
3522
3523 proxy_map = {}
3524 for handler in self._opener.handlers:
3525 if hasattr(handler, 'proxies'):
3526 proxy_map.update(handler.proxies)
3527 write_debug(f'Proxy map: {proxy_map}')
3528
3529 # Not implemented
3530 if False and self.params.get('call_home'):
3531 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3532 write_debug('Public IP address: %s' % ipaddr)
3533 latest_version = self.urlopen(
3534 'https://yt-dl.org/latest/version').read().decode('utf-8')
3535 if version_tuple(latest_version) > version_tuple(__version__):
3536 self.report_warning(
3537 'You are using an outdated version (newest version: %s)! '
3538 'See https://yt-dl.org/update if you need help updating.' %
3539 latest_version)
3540
3541 def _setup_opener(self):
3542 timeout_val = self.params.get('socket_timeout')
3543 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3544
3545 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3546 opts_cookiefile = self.params.get('cookiefile')
3547 opts_proxy = self.params.get('proxy')
3548
3549 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3550
3551 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3552 if opts_proxy is not None:
3553 if opts_proxy == '':
3554 proxies = {}
3555 else:
3556 proxies = {'http': opts_proxy, 'https': opts_proxy}
3557 else:
3558 proxies = compat_urllib_request.getproxies()
3559 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3560 if 'http' in proxies and 'https' not in proxies:
3561 proxies['https'] = proxies['http']
3562 proxy_handler = PerRequestProxyHandler(proxies)
3563
3564 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3565 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3566 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3567 redirect_handler = YoutubeDLRedirectHandler()
3568 data_handler = compat_urllib_request_DataHandler()
3569
3570 # When passing our own FileHandler instance, build_opener won't add the
3571 # default FileHandler and allows us to disable the file protocol, which
3572 # can be used for malicious purposes (see
3573 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3574 file_handler = compat_urllib_request.FileHandler()
3575
3576 def file_open(*args, **kwargs):
3577 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3578 file_handler.file_open = file_open
3579
3580 opener = compat_urllib_request.build_opener(
3581 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3582
3583 # Delete the default user-agent header, which would otherwise apply in
3584 # cases where our custom HTTP handler doesn't come into play
3585 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3586 opener.addheaders = []
3587 self._opener = opener
3588
3589 def encode(self, s):
3590 if isinstance(s, bytes):
3591 return s # Already encoded
3592
3593 try:
3594 return s.encode(self.get_encoding())
3595 except UnicodeEncodeError as err:
3596 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3597 raise
3598
3599 def get_encoding(self):
3600 encoding = self.params.get('encoding')
3601 if encoding is None:
3602 encoding = preferredencoding()
3603 return encoding
3604
3605 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3606 ''' Write infojson and returns True = written, False = skip, None = error '''
3607 if overwrite is None:
3608 overwrite = self.params.get('overwrites', True)
3609 if not self.params.get('writeinfojson'):
3610 return False
3611 elif not infofn:
3612 self.write_debug(f'Skipping writing {label} infojson')
3613 return False
3614 elif not self._ensure_dir_exists(infofn):
3615 return None
3616 elif not overwrite and os.path.exists(infofn):
3617 self.to_screen(f'[info] {label.title()} metadata is already present')
3618 else:
3619 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3620 try:
3621 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3622 except (OSError, IOError):
3623 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3624 return None
3625 return True
3626
3627 def _write_description(self, label, ie_result, descfn):
3628 ''' Write description and returns True = written, False = skip, None = error '''
3629 if not self.params.get('writedescription'):
3630 return False
3631 elif not descfn:
3632 self.write_debug(f'Skipping writing {label} description')
3633 return False
3634 elif not self._ensure_dir_exists(descfn):
3635 return None
3636 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3637 self.to_screen(f'[info] {label.title()} description is already present')
3638 elif ie_result.get('description') is None:
3639 self.report_warning(f'There\'s no {label} description to write')
3640 return False
3641 else:
3642 try:
3643 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3644 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3645 descfile.write(ie_result['description'])
3646 except (OSError, IOError):
3647 self.report_error(f'Cannot write {label} description file {descfn}')
3648 return None
3649 return True
3650
3651 def _write_subtitles(self, info_dict, filename):
3652 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3653 ret = []
3654 subtitles = info_dict.get('requested_subtitles')
3655 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3656 # subtitles download errors are already managed as troubles in relevant IE
3657 # that way it will silently go on when used with unsupporting IE
3658 return ret
3659
3660 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3661 if not sub_filename_base:
3662 self.to_screen('[info] Skipping writing video subtitles')
3663 return ret
3664 for sub_lang, sub_info in subtitles.items():
3665 sub_format = sub_info['ext']
3666 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3667 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3668 if not self.params.get('overwrites', True) and os.path.exists(sub_filename):
3669 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3670 sub_info['filepath'] = sub_filename
3671 ret.append((sub_filename, sub_filename_final))
3672 continue
3673
3674 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3675 if sub_info.get('data') is not None:
3676 try:
3677 # Use newline='' to prevent conversion of newline characters
3678 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3679 with io.open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3680 subfile.write(sub_info['data'])
3681 sub_info['filepath'] = sub_filename
3682 ret.append((sub_filename, sub_filename_final))
3683 continue
3684 except (OSError, IOError):
3685 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3686 return None
3687
3688 try:
3689 sub_copy = sub_info.copy()
3690 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3691 self.dl(sub_filename, sub_copy, subtitle=True)
3692 sub_info['filepath'] = sub_filename
3693 ret.append((sub_filename, sub_filename_final))
3694 except (ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3695 self.report_warning(f'Unable to download video subtitles for {sub_lang!r}: {err}')
3696 continue
3697 return ret
3698
3699 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3700 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3701 write_all = self.params.get('write_all_thumbnails', False)
3702 thumbnails, ret = [], []
3703 if write_all or self.params.get('writethumbnail', False):
3704 thumbnails = info_dict.get('thumbnails') or []
3705 multiple = write_all and len(thumbnails) > 1
3706
3707 if thumb_filename_base is None:
3708 thumb_filename_base = filename
3709 if thumbnails and not thumb_filename_base:
3710 self.write_debug(f'Skipping writing {label} thumbnail')
3711 return ret
3712
3713 for idx, t in list(enumerate(thumbnails))[::-1]:
3714 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3715 thumb_display_id = f'{label} thumbnail {t["id"]}'
3716 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3717 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3718
3719 if not self.params.get('overwrites', True) and os.path.exists(thumb_filename):
3720 ret.append((thumb_filename, thumb_filename_final))
3721 t['filepath'] = thumb_filename
3722 self.to_screen('[info] %s is already present' % (
3723 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3724 else:
3725 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3726 try:
3727 uf = self.urlopen(t['url'])
3728 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3729 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3730 shutil.copyfileobj(uf, thumbf)
3731 ret.append((thumb_filename, thumb_filename_final))
3732 t['filepath'] = thumb_filename
3733 except network_exceptions as err:
3734 thumbnails.pop(idx)
3735 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3736 if ret and not write_all:
3737 break
3738 return ret