]> jfr.im git - yt-dlp.git/blame_incremental - yt_dlp/YoutubeDL.py
Add `--extractor-args` to pass extractor-specific arguments
[yt-dlp.git] / yt_dlp / YoutubeDL.py
... / ...
CommitLineData
1#!/usr/bin/env python3
2# coding: utf-8
3
4from __future__ import absolute_import, unicode_literals
5
6import collections
7import contextlib
8import copy
9import datetime
10import errno
11import fileinput
12import io
13import itertools
14import json
15import locale
16import operator
17import os
18import platform
19import re
20import shutil
21import subprocess
22import sys
23import tempfile
24import time
25import tokenize
26import traceback
27import random
28
29from string import ascii_letters
30from zipimport import zipimporter
31
32from .compat import (
33 compat_basestring,
34 compat_cookiejar,
35 compat_get_terminal_size,
36 compat_kwargs,
37 compat_numeric_types,
38 compat_os_name,
39 compat_str,
40 compat_tokenize_tokenize,
41 compat_urllib_error,
42 compat_urllib_request,
43 compat_urllib_request_DataHandler,
44)
45from .utils import (
46 age_restricted,
47 args_to_str,
48 ContentTooShortError,
49 date_from_str,
50 DateRange,
51 DEFAULT_OUTTMPL,
52 determine_ext,
53 determine_protocol,
54 DOT_DESKTOP_LINK_TEMPLATE,
55 DOT_URL_LINK_TEMPLATE,
56 DOT_WEBLOC_LINK_TEMPLATE,
57 DownloadError,
58 encode_compat_str,
59 encodeFilename,
60 EntryNotInPlaylist,
61 error_to_compat_str,
62 ExistingVideoReached,
63 expand_path,
64 ExtractorError,
65 float_or_none,
66 format_bytes,
67 format_field,
68 STR_FORMAT_RE,
69 formatSeconds,
70 GeoRestrictedError,
71 HEADRequest,
72 int_or_none,
73 iri_to_uri,
74 ISO3166Utils,
75 LazyList,
76 locked_file,
77 make_dir,
78 make_HTTPS_handler,
79 MaxDownloadsReached,
80 network_exceptions,
81 orderedSet,
82 OUTTMPL_TYPES,
83 PagedList,
84 parse_filesize,
85 PerRequestProxyHandler,
86 platform_name,
87 PostProcessingError,
88 preferredencoding,
89 prepend_extension,
90 process_communicate_or_kill,
91 register_socks_protocols,
92 RejectedVideoReached,
93 render_table,
94 replace_extension,
95 SameFileError,
96 sanitize_filename,
97 sanitize_path,
98 sanitize_url,
99 sanitized_Request,
100 std_headers,
101 str_or_none,
102 strftime_or_none,
103 subtitles_filename,
104 ThrottledDownload,
105 to_high_limit_path,
106 traverse_obj,
107 try_get,
108 UnavailableVideoError,
109 url_basename,
110 version_tuple,
111 write_json_file,
112 write_string,
113 YoutubeDLCookieJar,
114 YoutubeDLCookieProcessor,
115 YoutubeDLHandler,
116 YoutubeDLRedirectHandler,
117)
118from .cache import Cache
119from .extractor import (
120 gen_extractor_classes,
121 get_info_extractor,
122 _LAZY_LOADER,
123 _PLUGIN_CLASSES
124)
125from .extractor.openload import PhantomJSwrapper
126from .downloader import (
127 get_suitable_downloader,
128 shorten_protocol_name
129)
130from .downloader.rtmp import rtmpdump_version
131from .postprocessor import (
132 get_postprocessor,
133 FFmpegFixupDurationPP,
134 FFmpegFixupM3u8PP,
135 FFmpegFixupM4aPP,
136 FFmpegFixupStretchedPP,
137 FFmpegFixupTimestampPP,
138 FFmpegMergerPP,
139 FFmpegPostProcessor,
140 MoveFilesAfterDownloadPP,
141)
142from .version import __version__
143
144if compat_os_name == 'nt':
145 import ctypes
146
147
148class YoutubeDL(object):
149 """YoutubeDL class.
150
151 YoutubeDL objects are the ones responsible of downloading the
152 actual video file and writing it to disk if the user has requested
153 it, among some other tasks. In most cases there should be one per
154 program. As, given a video URL, the downloader doesn't know how to
155 extract all the needed information, task that InfoExtractors do, it
156 has to pass the URL to one of them.
157
158 For this, YoutubeDL objects have a method that allows
159 InfoExtractors to be registered in a given order. When it is passed
160 a URL, the YoutubeDL object handles it to the first InfoExtractor it
161 finds that reports being able to handle it. The InfoExtractor extracts
162 all the information about the video or videos the URL refers to, and
163 YoutubeDL process the extracted information, possibly using a File
164 Downloader to download the video.
165
166 YoutubeDL objects accept a lot of parameters. In order not to saturate
167 the object constructor with arguments, it receives a dictionary of
168 options instead. These options are available through the params
169 attribute for the InfoExtractors to use. The YoutubeDL also
170 registers itself as the downloader in charge for the InfoExtractors
171 that are added to it, so this is a "mutual registration".
172
173 Available options:
174
175 username: Username for authentication purposes.
176 password: Password for authentication purposes.
177 videopassword: Password for accessing a video.
178 ap_mso: Adobe Pass multiple-system operator identifier.
179 ap_username: Multiple-system operator account username.
180 ap_password: Multiple-system operator account password.
181 usenetrc: Use netrc for authentication instead.
182 verbose: Print additional info to stdout.
183 quiet: Do not print messages to stdout.
184 no_warnings: Do not print out anything for warnings.
185 forceprint: A list of templates to force print
186 forceurl: Force printing final URL. (Deprecated)
187 forcetitle: Force printing title. (Deprecated)
188 forceid: Force printing ID. (Deprecated)
189 forcethumbnail: Force printing thumbnail URL. (Deprecated)
190 forcedescription: Force printing description. (Deprecated)
191 forcefilename: Force printing final filename. (Deprecated)
192 forceduration: Force printing duration. (Deprecated)
193 forcejson: Force printing info_dict as JSON.
194 dump_single_json: Force printing the info_dict of the whole playlist
195 (or video) as a single JSON line.
196 force_write_download_archive: Force writing download archive regardless
197 of 'skip_download' or 'simulate'.
198 simulate: Do not download the video files.
199 format: Video format code. see "FORMAT SELECTION" for more details.
200 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
201 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
202 extracting metadata even if the video is not actually
203 available for download (experimental)
204 format_sort: How to sort the video formats. see "Sorting Formats"
205 for more details.
206 format_sort_force: Force the given format_sort. see "Sorting Formats"
207 for more details.
208 allow_multiple_video_streams: Allow multiple video streams to be merged
209 into a single file
210 allow_multiple_audio_streams: Allow multiple audio streams to be merged
211 into a single file
212 paths: Dictionary of output paths. The allowed keys are 'home'
213 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
214 outtmpl: Dictionary of templates for output names. Allowed keys
215 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
216 A string a also accepted for backward compatibility
217 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
218 restrictfilenames: Do not allow "&" and spaces in file names
219 trim_file_name: Limit length of filename (extension excluded)
220 windowsfilenames: Force the filenames to be windows compatible
221 ignoreerrors: Do not stop on download errors
222 (Default True when running yt-dlp,
223 but False when directly accessing YoutubeDL class)
224 skip_playlist_after_errors: Number of allowed failures until the rest of
225 the playlist is skipped
226 force_generic_extractor: Force downloader to use the generic extractor
227 overwrites: Overwrite all video and metadata files if True,
228 overwrite only non-video files if None
229 and don't overwrite any file if False
230 playliststart: Playlist item to start at.
231 playlistend: Playlist item to end at.
232 playlist_items: Specific indices of playlist to download.
233 playlistreverse: Download playlist items in reverse order.
234 playlistrandom: Download playlist items in random order.
235 matchtitle: Download only matching titles.
236 rejecttitle: Reject downloads for matching titles.
237 logger: Log messages to a logging.Logger instance.
238 logtostderr: Log messages to stderr instead of stdout.
239 writedescription: Write the video description to a .description file
240 writeinfojson: Write the video description to a .info.json file
241 clean_infojson: Remove private fields from the infojson
242 writecomments: Extract video comments. This will not be written to disk
243 unless writeinfojson is also given
244 writeannotations: Write the video annotations to a .annotations.xml file
245 writethumbnail: Write the thumbnail image to a file
246 allow_playlist_files: Whether to write playlists' description, infojson etc
247 also to disk when using the 'write*' options
248 write_all_thumbnails: Write all thumbnail formats to files
249 writelink: Write an internet shortcut file, depending on the
250 current platform (.url/.webloc/.desktop)
251 writeurllink: Write a Windows internet shortcut file (.url)
252 writewebloclink: Write a macOS internet shortcut file (.webloc)
253 writedesktoplink: Write a Linux internet shortcut file (.desktop)
254 writesubtitles: Write the video subtitles to a file
255 writeautomaticsub: Write the automatically generated subtitles to a file
256 allsubtitles: Deprecated - Use subtitlelangs = ['all']
257 Downloads all the subtitles of the video
258 (requires writesubtitles or writeautomaticsub)
259 listsubtitles: Lists all available subtitles for the video
260 subtitlesformat: The format code for subtitles
261 subtitleslangs: List of languages of the subtitles to download (can be regex).
262 The list may contain "all" to refer to all the available
263 subtitles. The language can be prefixed with a "-" to
264 exclude it from the requested languages. Eg: ['all', '-live_chat']
265 keepvideo: Keep the video file after post-processing
266 daterange: A DateRange object, download only if the upload_date is in the range.
267 skip_download: Skip the actual download of the video file
268 cachedir: Location of the cache files in the filesystem.
269 False to disable filesystem cache.
270 noplaylist: Download single video instead of a playlist if in doubt.
271 age_limit: An integer representing the user's age in years.
272 Unsuitable videos for the given age are skipped.
273 min_views: An integer representing the minimum view count the video
274 must have in order to not be skipped.
275 Videos without view count information are always
276 downloaded. None for no limit.
277 max_views: An integer representing the maximum view count.
278 Videos that are more popular than that are not
279 downloaded.
280 Videos without view count information are always
281 downloaded. None for no limit.
282 download_archive: File name of a file where all downloads are recorded.
283 Videos already present in the file are not downloaded
284 again.
285 break_on_existing: Stop the download process after attempting to download a
286 file that is in the archive.
287 break_on_reject: Stop the download process when encountering a video that
288 has been filtered out.
289 cookiefile: File name where cookies should be read from and dumped to
290 nocheckcertificate:Do not verify SSL certificates
291 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
292 At the moment, this is only supported by YouTube.
293 proxy: URL of the proxy server to use
294 geo_verification_proxy: URL of the proxy to use for IP address verification
295 on geo-restricted sites.
296 socket_timeout: Time to wait for unresponsive hosts, in seconds
297 bidi_workaround: Work around buggy terminals without bidirectional text
298 support, using fridibi
299 debug_printtraffic:Print out sent and received HTTP traffic
300 include_ads: Download ads as well
301 default_search: Prepend this string if an input url is not valid.
302 'auto' for elaborate guessing
303 encoding: Use this encoding instead of the system-specified.
304 extract_flat: Do not resolve URLs, return the immediate result.
305 Pass in 'in_playlist' to only show this behavior for
306 playlist items.
307 postprocessors: A list of dictionaries, each with an entry
308 * key: The name of the postprocessor. See
309 yt_dlp/postprocessor/__init__.py for a list.
310 * when: When to run the postprocessor. Can be one of
311 pre_process|before_dl|post_process|after_move.
312 Assumed to be 'post_process' if not given
313 post_hooks: A list of functions that get called as the final step
314 for each video file, after all postprocessors have been
315 called. The filename will be passed as the only argument.
316 progress_hooks: A list of functions that get called on download
317 progress, with a dictionary with the entries
318 * status: One of "downloading", "error", or "finished".
319 Check this first and ignore unknown values.
320
321 If status is one of "downloading", or "finished", the
322 following properties may also be present:
323 * filename: The final filename (always present)
324 * tmpfilename: The filename we're currently writing to
325 * downloaded_bytes: Bytes on disk
326 * total_bytes: Size of the whole file, None if unknown
327 * total_bytes_estimate: Guess of the eventual file size,
328 None if unavailable.
329 * elapsed: The number of seconds since download started.
330 * eta: The estimated time in seconds, None if unknown
331 * speed: The download speed in bytes/second, None if
332 unknown
333 * fragment_index: The counter of the currently
334 downloaded video fragment.
335 * fragment_count: The number of fragments (= individual
336 files that will be merged)
337
338 Progress hooks are guaranteed to be called at least once
339 (with status "finished") if the download is successful.
340 merge_output_format: Extension to use when merging formats.
341 final_ext: Expected final extension; used to detect when the file was
342 already downloaded and converted. "merge_output_format" is
343 replaced by this extension when given
344 fixup: Automatically correct known faults of the file.
345 One of:
346 - "never": do nothing
347 - "warn": only emit a warning
348 - "detect_or_warn": check whether we can do anything
349 about it, warn otherwise (default)
350 source_address: Client-side IP address to bind to.
351 call_home: Boolean, true iff we are allowed to contact the
352 yt-dlp servers for debugging. (BROKEN)
353 sleep_interval_requests: Number of seconds to sleep between requests
354 during extraction
355 sleep_interval: Number of seconds to sleep before each download when
356 used alone or a lower bound of a range for randomized
357 sleep before each download (minimum possible number
358 of seconds to sleep) when used along with
359 max_sleep_interval.
360 max_sleep_interval:Upper bound of a range for randomized sleep before each
361 download (maximum possible number of seconds to sleep).
362 Must only be used along with sleep_interval.
363 Actual sleep time will be a random float from range
364 [sleep_interval; max_sleep_interval].
365 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
366 listformats: Print an overview of available video formats and exit.
367 list_thumbnails: Print a table of all thumbnails and exit.
368 match_filter: A function that gets called with the info_dict of
369 every video.
370 If it returns a message, the video is ignored.
371 If it returns None, the video is downloaded.
372 match_filter_func in utils.py is one example for this.
373 no_color: Do not emit color codes in output.
374 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
375 HTTP header
376 geo_bypass_country:
377 Two-letter ISO 3166-2 country code that will be used for
378 explicit geographic restriction bypassing via faking
379 X-Forwarded-For HTTP header
380 geo_bypass_ip_block:
381 IP range in CIDR notation that will be used similarly to
382 geo_bypass_country
383
384 The following options determine which downloader is picked:
385 external_downloader: A dictionary of protocol keys and the executable of the
386 external downloader to use for it. The allowed protocols
387 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
388 Set the value to 'native' to use the native downloader
389 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
390 or {'m3u8': 'ffmpeg'} instead.
391 Use the native HLS downloader instead of ffmpeg/avconv
392 if True, otherwise use ffmpeg/avconv if False, otherwise
393 use downloader suggested by extractor if None.
394 compat_opts: Compatibility options. See "Differences in default behavior".
395 Note that only format-sort, format-spec, no-live-chat,
396 no-attach-info-json, playlist-index, list-formats,
397 no-direct-merge, embed-thumbnail-atomicparsley,
398 no-youtube-unavailable-videos, no-youtube-channel-redirect,
399 works when used via the API
400
401 The following parameters are not used by YoutubeDL itself, they are used by
402 the downloader (see yt_dlp/downloader/common.py):
403 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
404 max_filesize, test, noresizebuffer, retries, continuedl, noprogress, consoletitle,
405 xattr_set_filesize, external_downloader_args, hls_use_mpegts, http_chunk_size.
406
407 The following options are used by the post processors:
408 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
409 otherwise prefer ffmpeg. (avconv support is deprecated)
410 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
411 to the binary or its containing directory.
412 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
413 and a list of additional command-line arguments for the
414 postprocessor/executable. The dict can also have "PP+EXE" keys
415 which are used when the given exe is used by the given PP.
416 Use 'default' as the name for arguments to passed to all PP
417
418 The following options are used by the extractors:
419 extractor_retries: Number of times to retry for known errors
420 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
421 hls_split_discontinuity: Split HLS playlists to different formats at
422 discontinuities such as ad breaks (default: False)
423 extractor_args: A dictionary of arguments to be passed to the extractors.
424 See "EXTRACTOR ARGUMENTS" for details.
425 Eg: {'youtube': {'skip': ['dash', 'hls']}}
426 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
427 If True (default), DASH manifests and related
428 data will be downloaded and processed by extractor.
429 You can reduce network I/O by disabling it if you don't
430 care about DASH. (only for youtube)
431 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
432 If True (default), HLS manifests and related
433 data will be downloaded and processed by extractor.
434 You can reduce network I/O by disabling it if you don't
435 care about HLS. (only for youtube)
436 """
437
438 _NUMERIC_FIELDS = set((
439 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
440 'timestamp', 'upload_year', 'upload_month', 'upload_day',
441 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
442 'average_rating', 'comment_count', 'age_limit',
443 'start_time', 'end_time',
444 'chapter_number', 'season_number', 'episode_number',
445 'track_number', 'disc_number', 'release_year',
446 'playlist_index',
447 ))
448
449 params = None
450 _ies = []
451 _pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
452 __prepare_filename_warned = False
453 _first_webpage_request = True
454 _download_retcode = None
455 _num_downloads = None
456 _playlist_level = 0
457 _playlist_urls = set()
458 _screen_file = None
459
460 def __init__(self, params=None, auto_init=True):
461 """Create a FileDownloader object with the given options."""
462 if params is None:
463 params = {}
464 self._ies = []
465 self._ies_instances = {}
466 self._pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
467 self.__prepare_filename_warned = False
468 self._first_webpage_request = True
469 self._post_hooks = []
470 self._progress_hooks = []
471 self._download_retcode = 0
472 self._num_downloads = 0
473 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
474 self._err_file = sys.stderr
475 self.params = {
476 # Default parameters
477 'nocheckcertificate': False,
478 }
479 self.params.update(params)
480 self.cache = Cache(self)
481
482 if sys.version_info < (3, 6):
483 self.report_warning(
484 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
485
486 def check_deprecated(param, option, suggestion):
487 if self.params.get(param) is not None:
488 self.report_warning('%s is deprecated. Use %s instead' % (option, suggestion))
489 return True
490 return False
491
492 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
493 if self.params.get('geo_verification_proxy') is None:
494 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
495
496 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
497 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
498 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
499
500 for msg in self.params.get('warnings', []):
501 self.report_warning(msg)
502
503 if self.params.get('final_ext'):
504 if self.params.get('merge_output_format'):
505 self.report_warning('--merge-output-format will be ignored since --remux-video or --recode-video is given')
506 self.params['merge_output_format'] = self.params['final_ext']
507
508 if 'overwrites' in self.params and self.params['overwrites'] is None:
509 del self.params['overwrites']
510
511 if params.get('bidi_workaround', False):
512 try:
513 import pty
514 master, slave = pty.openpty()
515 width = compat_get_terminal_size().columns
516 if width is None:
517 width_args = []
518 else:
519 width_args = ['-w', str(width)]
520 sp_kwargs = dict(
521 stdin=subprocess.PIPE,
522 stdout=slave,
523 stderr=self._err_file)
524 try:
525 self._output_process = subprocess.Popen(
526 ['bidiv'] + width_args, **sp_kwargs
527 )
528 except OSError:
529 self._output_process = subprocess.Popen(
530 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
531 self._output_channel = os.fdopen(master, 'rb')
532 except OSError as ose:
533 if ose.errno == errno.ENOENT:
534 self.report_warning('Could not find fribidi executable, ignoring --bidi-workaround . Make sure that fribidi is an executable file in one of the directories in your $PATH.')
535 else:
536 raise
537
538 if (sys.platform != 'win32'
539 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
540 and not params.get('restrictfilenames', False)):
541 # Unicode filesystem API will throw errors (#1474, #13027)
542 self.report_warning(
543 'Assuming --restrict-filenames since file system encoding '
544 'cannot encode all characters. '
545 'Set the LC_ALL environment variable to fix this.')
546 self.params['restrictfilenames'] = True
547
548 self.outtmpl_dict = self.parse_outtmpl()
549
550 # Creating format selector here allows us to catch syntax errors before the extraction
551 self.format_selector = (
552 None if self.params.get('format') is None
553 else self.build_format_selector(self.params['format']))
554
555 self._setup_opener()
556
557 """Preload the archive, if any is specified"""
558 def preload_download_archive(fn):
559 if fn is None:
560 return False
561 self.write_debug('Loading archive file %r\n' % fn)
562 try:
563 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
564 for line in archive_file:
565 self.archive.add(line.strip())
566 except IOError as ioe:
567 if ioe.errno != errno.ENOENT:
568 raise
569 return False
570 return True
571
572 self.archive = set()
573 preload_download_archive(self.params.get('download_archive'))
574
575 if auto_init:
576 self.print_debug_header()
577 self.add_default_info_extractors()
578
579 for pp_def_raw in self.params.get('postprocessors', []):
580 pp_def = dict(pp_def_raw)
581 when = pp_def.pop('when', 'post_process')
582 pp_class = get_postprocessor(pp_def.pop('key'))
583 pp = pp_class(self, **compat_kwargs(pp_def))
584 self.add_post_processor(pp, when=when)
585
586 for ph in self.params.get('post_hooks', []):
587 self.add_post_hook(ph)
588
589 for ph in self.params.get('progress_hooks', []):
590 self.add_progress_hook(ph)
591
592 register_socks_protocols()
593
594 def warn_if_short_id(self, argv):
595 # short YouTube ID starting with dash?
596 idxs = [
597 i for i, a in enumerate(argv)
598 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
599 if idxs:
600 correct_argv = (
601 ['yt-dlp']
602 + [a for i, a in enumerate(argv) if i not in idxs]
603 + ['--'] + [argv[i] for i in idxs]
604 )
605 self.report_warning(
606 'Long argument string detected. '
607 'Use -- to separate parameters and URLs, like this:\n%s\n' %
608 args_to_str(correct_argv))
609
610 def add_info_extractor(self, ie):
611 """Add an InfoExtractor object to the end of the list."""
612 self._ies.append(ie)
613 if not isinstance(ie, type):
614 self._ies_instances[ie.ie_key()] = ie
615 ie.set_downloader(self)
616
617 def get_info_extractor(self, ie_key):
618 """
619 Get an instance of an IE with name ie_key, it will try to get one from
620 the _ies list, if there's no instance it will create a new one and add
621 it to the extractor list.
622 """
623 ie = self._ies_instances.get(ie_key)
624 if ie is None:
625 ie = get_info_extractor(ie_key)()
626 self.add_info_extractor(ie)
627 return ie
628
629 def add_default_info_extractors(self):
630 """
631 Add the InfoExtractors returned by gen_extractors to the end of the list
632 """
633 for ie in gen_extractor_classes():
634 self.add_info_extractor(ie)
635
636 def add_post_processor(self, pp, when='post_process'):
637 """Add a PostProcessor object to the end of the chain."""
638 self._pps[when].append(pp)
639 pp.set_downloader(self)
640
641 def add_post_hook(self, ph):
642 """Add the post hook"""
643 self._post_hooks.append(ph)
644
645 def add_progress_hook(self, ph):
646 """Add the progress hook (currently only for the file downloader)"""
647 self._progress_hooks.append(ph)
648
649 def _bidi_workaround(self, message):
650 if not hasattr(self, '_output_channel'):
651 return message
652
653 assert hasattr(self, '_output_process')
654 assert isinstance(message, compat_str)
655 line_count = message.count('\n') + 1
656 self._output_process.stdin.write((message + '\n').encode('utf-8'))
657 self._output_process.stdin.flush()
658 res = ''.join(self._output_channel.readline().decode('utf-8')
659 for _ in range(line_count))
660 return res[:-len('\n')]
661
662 def _write_string(self, s, out=None):
663 write_string(s, out=out, encoding=self.params.get('encoding'))
664
665 def to_stdout(self, message, skip_eol=False, quiet=False):
666 """Print message to stdout"""
667 if self.params.get('logger'):
668 self.params['logger'].debug(message)
669 elif not quiet or self.params.get('verbose'):
670 self._write_string(
671 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
672 self._err_file if quiet else self._screen_file)
673
674 def to_stderr(self, message):
675 """Print message to stderr"""
676 assert isinstance(message, compat_str)
677 if self.params.get('logger'):
678 self.params['logger'].error(message)
679 else:
680 self._write_string('%s\n' % self._bidi_workaround(message), self._err_file)
681
682 def to_console_title(self, message):
683 if not self.params.get('consoletitle', False):
684 return
685 if compat_os_name == 'nt':
686 if ctypes.windll.kernel32.GetConsoleWindow():
687 # c_wchar_p() might not be necessary if `message` is
688 # already of type unicode()
689 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
690 elif 'TERM' in os.environ:
691 self._write_string('\033]0;%s\007' % message, self._screen_file)
692
693 def save_console_title(self):
694 if not self.params.get('consoletitle', False):
695 return
696 if self.params.get('simulate', False):
697 return
698 if compat_os_name != 'nt' and 'TERM' in os.environ:
699 # Save the title on stack
700 self._write_string('\033[22;0t', self._screen_file)
701
702 def restore_console_title(self):
703 if not self.params.get('consoletitle', False):
704 return
705 if self.params.get('simulate', False):
706 return
707 if compat_os_name != 'nt' and 'TERM' in os.environ:
708 # Restore the title from stack
709 self._write_string('\033[23;0t', self._screen_file)
710
711 def __enter__(self):
712 self.save_console_title()
713 return self
714
715 def __exit__(self, *args):
716 self.restore_console_title()
717
718 if self.params.get('cookiefile') is not None:
719 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
720
721 def trouble(self, message=None, tb=None):
722 """Determine action to take when a download problem appears.
723
724 Depending on if the downloader has been configured to ignore
725 download errors or not, this method may throw an exception or
726 not when errors are found, after printing the message.
727
728 tb, if given, is additional traceback information.
729 """
730 if message is not None:
731 self.to_stderr(message)
732 if self.params.get('verbose'):
733 if tb is None:
734 if sys.exc_info()[0]: # if .trouble has been called from an except block
735 tb = ''
736 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
737 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
738 tb += encode_compat_str(traceback.format_exc())
739 else:
740 tb_data = traceback.format_list(traceback.extract_stack())
741 tb = ''.join(tb_data)
742 if tb:
743 self.to_stderr(tb)
744 if not self.params.get('ignoreerrors', False):
745 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
746 exc_info = sys.exc_info()[1].exc_info
747 else:
748 exc_info = sys.exc_info()
749 raise DownloadError(message, exc_info)
750 self._download_retcode = 1
751
752 def to_screen(self, message, skip_eol=False):
753 """Print message to stdout if not in quiet mode"""
754 self.to_stdout(
755 message, skip_eol, quiet=self.params.get('quiet', False))
756
757 def report_warning(self, message):
758 '''
759 Print the message to stderr, it will be prefixed with 'WARNING:'
760 If stderr is a tty file the 'WARNING:' will be colored
761 '''
762 if self.params.get('logger') is not None:
763 self.params['logger'].warning(message)
764 else:
765 if self.params.get('no_warnings'):
766 return
767 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
768 _msg_header = '\033[0;33mWARNING:\033[0m'
769 else:
770 _msg_header = 'WARNING:'
771 warning_message = '%s %s' % (_msg_header, message)
772 self.to_stderr(warning_message)
773
774 def report_error(self, message, tb=None):
775 '''
776 Do the same as trouble, but prefixes the message with 'ERROR:', colored
777 in red if stderr is a tty file.
778 '''
779 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
780 _msg_header = '\033[0;31mERROR:\033[0m'
781 else:
782 _msg_header = 'ERROR:'
783 error_message = '%s %s' % (_msg_header, message)
784 self.trouble(error_message, tb)
785
786 def write_debug(self, message):
787 '''Log debug message or Print message to stderr'''
788 if not self.params.get('verbose', False):
789 return
790 message = '[debug] %s' % message
791 if self.params.get('logger'):
792 self.params['logger'].debug(message)
793 else:
794 self._write_string('%s\n' % message)
795
796 def report_file_already_downloaded(self, file_name):
797 """Report file has already been fully downloaded."""
798 try:
799 self.to_screen('[download] %s has already been downloaded' % file_name)
800 except UnicodeEncodeError:
801 self.to_screen('[download] The file has already been downloaded')
802
803 def report_file_delete(self, file_name):
804 """Report that existing file will be deleted."""
805 try:
806 self.to_screen('Deleting existing file %s' % file_name)
807 except UnicodeEncodeError:
808 self.to_screen('Deleting existing file')
809
810 def parse_outtmpl(self):
811 outtmpl_dict = self.params.get('outtmpl', {})
812 if not isinstance(outtmpl_dict, dict):
813 outtmpl_dict = {'default': outtmpl_dict}
814 outtmpl_dict.update({
815 k: v for k, v in DEFAULT_OUTTMPL.items()
816 if not outtmpl_dict.get(k)})
817 for key, val in outtmpl_dict.items():
818 if isinstance(val, bytes):
819 self.report_warning(
820 'Parameter outtmpl is bytes, but should be a unicode string. '
821 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
822 return outtmpl_dict
823
824 def get_output_path(self, dir_type='', filename=None):
825 paths = self.params.get('paths', {})
826 assert isinstance(paths, dict)
827 path = os.path.join(
828 expand_path(paths.get('home', '').strip()),
829 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
830 filename or '')
831
832 # Temporary fix for #4787
833 # 'Treat' all problem characters by passing filename through preferredencoding
834 # to workaround encoding issues with subprocess on python2 @ Windows
835 if sys.version_info < (3, 0) and sys.platform == 'win32':
836 path = encodeFilename(path, True).decode(preferredencoding())
837 return sanitize_path(path, force=self.params.get('windowsfilenames'))
838
839 @staticmethod
840 def validate_outtmpl(tmpl):
841 ''' @return None or Exception object '''
842 try:
843 re.sub(
844 STR_FORMAT_RE.format(''),
845 lambda mobj: ('%' if not mobj.group('has_key') else '') + mobj.group(0),
846 tmpl
847 ) % collections.defaultdict(int)
848 return None
849 except ValueError as err:
850 return err
851
852 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=None):
853 """ Make the template and info_dict suitable for substitution (outtmpl % info_dict)"""
854 info_dict = dict(info_dict)
855 na = self.params.get('outtmpl_na_placeholder', 'NA')
856
857 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
858 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
859 if info_dict.get('duration', None) is not None
860 else None)
861 info_dict['epoch'] = int(time.time())
862 info_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
863 if info_dict.get('resolution') is None:
864 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
865
866 # For fields playlist_index and autonumber convert all occurrences
867 # of %(field)s to %(field)0Nd for backward compatibility
868 field_size_compat_map = {
869 'playlist_index': len(str(info_dict.get('_last_playlist_index') or '')),
870 'autonumber': self.params.get('autonumber_size') or 5,
871 }
872
873 TMPL_DICT = {}
874 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE.format('[^)]*'))
875 MATH_FUNCTIONS = {
876 '+': float.__add__,
877 '-': float.__sub__,
878 }
879 # Field is of the form key1.key2...
880 # where keys (except first) can be string, int or slice
881 FIELD_RE = r'\w+(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
882 MATH_FIELD_RE = r'''{field}|{num}'''.format(field=FIELD_RE, num=r'-?\d+(?:.\d+)?')
883 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
884 INTERNAL_FORMAT_RE = re.compile(r'''(?x)
885 (?P<negate>-)?
886 (?P<fields>{field})
887 (?P<maths>(?:{math_op}{math_field})*)
888 (?:>(?P<strf_format>.+?))?
889 (?:\|(?P<default>.*?))?
890 $'''.format(field=FIELD_RE, math_op=MATH_OPERATORS_RE, math_field=MATH_FIELD_RE))
891
892 get_key = lambda k: traverse_obj(
893 info_dict, k.split('.'), is_user_input=True, traverse_string=True)
894
895 def get_value(mdict):
896 # Object traversal
897 value = get_key(mdict['fields'])
898 # Negative
899 if mdict['negate']:
900 value = float_or_none(value)
901 if value is not None:
902 value *= -1
903 # Do maths
904 offset_key = mdict['maths']
905 if offset_key:
906 value = float_or_none(value)
907 operator = None
908 while offset_key:
909 item = re.match(
910 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
911 offset_key).group(0)
912 offset_key = offset_key[len(item):]
913 if operator is None:
914 operator = MATH_FUNCTIONS[item]
915 continue
916 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
917 offset = float_or_none(item)
918 if offset is None:
919 offset = float_or_none(get_key(item))
920 try:
921 value = operator(value, multiplier * offset)
922 except (TypeError, ZeroDivisionError):
923 return None
924 operator = None
925 # Datetime formatting
926 if mdict['strf_format']:
927 value = strftime_or_none(value, mdict['strf_format'])
928
929 return value
930
931 def create_key(outer_mobj):
932 if not outer_mobj.group('has_key'):
933 return '%{}'.format(outer_mobj.group(0))
934
935 key = outer_mobj.group('key')
936 fmt = outer_mobj.group('format')
937 mobj = re.match(INTERNAL_FORMAT_RE, key)
938 if mobj is None:
939 value, default, mobj = None, na, {'fields': ''}
940 else:
941 mobj = mobj.groupdict()
942 default = mobj['default'] if mobj['default'] is not None else na
943 value = get_value(mobj)
944
945 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
946 fmt = '0{:d}d'.format(field_size_compat_map[key])
947
948 value = default if value is None else value
949
950 if fmt == 'c':
951 value = compat_str(value)
952 if value is None:
953 value, fmt = default, 's'
954 else:
955 value = value[0]
956 elif fmt[-1] not in 'rs': # numeric
957 value = float_or_none(value)
958 if value is None:
959 value, fmt = default, 's'
960 if sanitize:
961 if fmt[-1] == 'r':
962 # If value is an object, sanitize might convert it to a string
963 # So we convert it to repr first
964 value, fmt = repr(value), '%ss' % fmt[:-1]
965 if fmt[-1] in 'csr':
966 value = sanitize(mobj['fields'].split('.')[-1], value)
967 key += '\0%s' % fmt
968 TMPL_DICT[key] = value
969 return '%({key}){fmt}'.format(key=key, fmt=fmt)
970
971 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
972
973 def _prepare_filename(self, info_dict, tmpl_type='default'):
974 try:
975 sanitize = lambda k, v: sanitize_filename(
976 compat_str(v),
977 restricted=self.params.get('restrictfilenames'),
978 is_id=(k == 'id' or k.endswith('_id')))
979 outtmpl = self.outtmpl_dict.get(tmpl_type, self.outtmpl_dict['default'])
980 outtmpl, template_dict = self.prepare_outtmpl(outtmpl, info_dict, sanitize)
981
982 # expand_path translates '%%' into '%' and '$$' into '$'
983 # correspondingly that is not what we want since we need to keep
984 # '%%' intact for template dict substitution step. Working around
985 # with boundary-alike separator hack.
986 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
987 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
988
989 # outtmpl should be expand_path'ed before template dict substitution
990 # because meta fields may contain env variables we don't want to
991 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
992 # title "Hello $PATH", we don't want `$PATH` to be expanded.
993 filename = expand_path(outtmpl).replace(sep, '') % template_dict
994
995 force_ext = OUTTMPL_TYPES.get(tmpl_type)
996 if force_ext is not None:
997 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
998
999 # https://github.com/blackjack4494/youtube-dlc/issues/85
1000 trim_file_name = self.params.get('trim_file_name', False)
1001 if trim_file_name:
1002 fn_groups = filename.rsplit('.')
1003 ext = fn_groups[-1]
1004 sub_ext = ''
1005 if len(fn_groups) > 2:
1006 sub_ext = fn_groups[-2]
1007 filename = '.'.join(filter(None, [fn_groups[0][:trim_file_name], sub_ext, ext]))
1008
1009 return filename
1010 except ValueError as err:
1011 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1012 return None
1013
1014 def prepare_filename(self, info_dict, dir_type='', warn=False):
1015 """Generate the output filename."""
1016
1017 filename = self._prepare_filename(info_dict, dir_type or 'default')
1018
1019 if warn and not self.__prepare_filename_warned:
1020 if not self.params.get('paths'):
1021 pass
1022 elif filename == '-':
1023 self.report_warning('--paths is ignored when an outputting to stdout')
1024 elif os.path.isabs(filename):
1025 self.report_warning('--paths is ignored since an absolute path is given in output template')
1026 self.__prepare_filename_warned = True
1027 if filename == '-' or not filename:
1028 return filename
1029
1030 return self.get_output_path(dir_type, filename)
1031
1032 def _match_entry(self, info_dict, incomplete=False, silent=False):
1033 """ Returns None if the file should be downloaded """
1034
1035 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1036
1037 def check_filter():
1038 if 'title' in info_dict:
1039 # This can happen when we're just evaluating the playlist
1040 title = info_dict['title']
1041 matchtitle = self.params.get('matchtitle', False)
1042 if matchtitle:
1043 if not re.search(matchtitle, title, re.IGNORECASE):
1044 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1045 rejecttitle = self.params.get('rejecttitle', False)
1046 if rejecttitle:
1047 if re.search(rejecttitle, title, re.IGNORECASE):
1048 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1049 date = info_dict.get('upload_date')
1050 if date is not None:
1051 dateRange = self.params.get('daterange', DateRange())
1052 if date not in dateRange:
1053 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
1054 view_count = info_dict.get('view_count')
1055 if view_count is not None:
1056 min_views = self.params.get('min_views')
1057 if min_views is not None and view_count < min_views:
1058 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1059 max_views = self.params.get('max_views')
1060 if max_views is not None and view_count > max_views:
1061 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1062 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1063 return 'Skipping "%s" because it is age restricted' % video_title
1064
1065 if not incomplete:
1066 match_filter = self.params.get('match_filter')
1067 if match_filter is not None:
1068 ret = match_filter(info_dict)
1069 if ret is not None:
1070 return ret
1071 return None
1072
1073 if self.in_download_archive(info_dict):
1074 reason = '%s has already been recorded in the archive' % video_title
1075 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1076 else:
1077 reason = check_filter()
1078 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1079 if reason is not None:
1080 if not silent:
1081 self.to_screen('[download] ' + reason)
1082 if self.params.get(break_opt, False):
1083 raise break_err()
1084 return reason
1085
1086 @staticmethod
1087 def add_extra_info(info_dict, extra_info):
1088 '''Set the keys from extra_info in info dict if they are missing'''
1089 for key, value in extra_info.items():
1090 info_dict.setdefault(key, value)
1091
1092 def extract_info(self, url, download=True, ie_key=None, extra_info={},
1093 process=True, force_generic_extractor=False):
1094 """
1095 Return a list with a dictionary for each video extracted.
1096
1097 Arguments:
1098 url -- URL to extract
1099
1100 Keyword arguments:
1101 download -- whether to download videos during extraction
1102 ie_key -- extractor key hint
1103 extra_info -- dictionary containing the extra values to add to each result
1104 process -- whether to resolve all unresolved references (URLs, playlist items),
1105 must be True for download to work.
1106 force_generic_extractor -- force using the generic extractor
1107 """
1108
1109 if not ie_key and force_generic_extractor:
1110 ie_key = 'Generic'
1111
1112 if ie_key:
1113 ies = [self.get_info_extractor(ie_key)]
1114 else:
1115 ies = self._ies
1116
1117 for ie in ies:
1118 if not ie.suitable(url):
1119 continue
1120
1121 ie_key = ie.ie_key()
1122 ie = self.get_info_extractor(ie_key)
1123 if not ie.working():
1124 self.report_warning('The program functionality for this site has been marked as broken, '
1125 'and will probably not work.')
1126
1127 try:
1128 temp_id = str_or_none(
1129 ie.extract_id(url) if callable(getattr(ie, 'extract_id', None))
1130 else ie._match_id(url))
1131 except (AssertionError, IndexError, AttributeError):
1132 temp_id = None
1133 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1134 self.to_screen("[%s] %s: has already been recorded in archive" % (
1135 ie_key, temp_id))
1136 break
1137 return self.__extract_info(url, ie, download, extra_info, process)
1138 else:
1139 self.report_error('no suitable InfoExtractor for URL %s' % url)
1140
1141 def __handle_extraction_exceptions(func):
1142 def wrapper(self, *args, **kwargs):
1143 try:
1144 return func(self, *args, **kwargs)
1145 except GeoRestrictedError as e:
1146 msg = e.msg
1147 if e.countries:
1148 msg += '\nThis video is available in %s.' % ', '.join(
1149 map(ISO3166Utils.short2full, e.countries))
1150 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1151 self.report_error(msg)
1152 except ExtractorError as e: # An error we somewhat expected
1153 self.report_error(compat_str(e), e.format_traceback())
1154 except ThrottledDownload:
1155 self.to_stderr('\r')
1156 self.report_warning('The download speed is below throttle limit. Re-extracting data')
1157 return wrapper(self, *args, **kwargs)
1158 except (MaxDownloadsReached, ExistingVideoReached, RejectedVideoReached):
1159 raise
1160 except Exception as e:
1161 if self.params.get('ignoreerrors', False):
1162 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
1163 else:
1164 raise
1165 return wrapper
1166
1167 @__handle_extraction_exceptions
1168 def __extract_info(self, url, ie, download, extra_info, process):
1169 ie_result = ie.extract(url)
1170 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1171 return
1172 if isinstance(ie_result, list):
1173 # Backwards compatibility: old IE result format
1174 ie_result = {
1175 '_type': 'compat_list',
1176 'entries': ie_result,
1177 }
1178 self.add_default_extra_info(ie_result, ie, url)
1179 if process:
1180 return self.process_ie_result(ie_result, download, extra_info)
1181 else:
1182 return ie_result
1183
1184 def add_default_extra_info(self, ie_result, ie, url):
1185 if url is not None:
1186 self.add_extra_info(ie_result, {
1187 'webpage_url': url,
1188 'original_url': url,
1189 'webpage_url_basename': url_basename(url),
1190 })
1191 if ie is not None:
1192 self.add_extra_info(ie_result, {
1193 'extractor': ie.IE_NAME,
1194 'extractor_key': ie.ie_key(),
1195 })
1196
1197 def process_ie_result(self, ie_result, download=True, extra_info={}):
1198 """
1199 Take the result of the ie(may be modified) and resolve all unresolved
1200 references (URLs, playlist items).
1201
1202 It will also download the videos if 'download'.
1203 Returns the resolved ie_result.
1204 """
1205 result_type = ie_result.get('_type', 'video')
1206
1207 if result_type in ('url', 'url_transparent'):
1208 ie_result['url'] = sanitize_url(ie_result['url'])
1209 extract_flat = self.params.get('extract_flat', False)
1210 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1211 or extract_flat is True):
1212 info_copy = ie_result.copy()
1213 self.add_extra_info(info_copy, extra_info)
1214 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1215 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1216 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1217 return ie_result
1218
1219 if result_type == 'video':
1220 self.add_extra_info(ie_result, extra_info)
1221 ie_result = self.process_video_result(ie_result, download=download)
1222 additional_urls = (ie_result or {}).get('additional_urls')
1223 if additional_urls:
1224 # TODO: Improve MetadataFromFieldPP to allow setting a list
1225 if isinstance(additional_urls, compat_str):
1226 additional_urls = [additional_urls]
1227 self.to_screen(
1228 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1229 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1230 ie_result['additional_entries'] = [
1231 self.extract_info(
1232 url, download, extra_info,
1233 force_generic_extractor=self.params.get('force_generic_extractor'))
1234 for url in additional_urls
1235 ]
1236 return ie_result
1237 elif result_type == 'url':
1238 # We have to add extra_info to the results because it may be
1239 # contained in a playlist
1240 return self.extract_info(
1241 ie_result['url'], download,
1242 ie_key=ie_result.get('ie_key'),
1243 extra_info=extra_info)
1244 elif result_type == 'url_transparent':
1245 # Use the information from the embedding page
1246 info = self.extract_info(
1247 ie_result['url'], ie_key=ie_result.get('ie_key'),
1248 extra_info=extra_info, download=False, process=False)
1249
1250 # extract_info may return None when ignoreerrors is enabled and
1251 # extraction failed with an error, don't crash and return early
1252 # in this case
1253 if not info:
1254 return info
1255
1256 force_properties = dict(
1257 (k, v) for k, v in ie_result.items() if v is not None)
1258 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1259 if f in force_properties:
1260 del force_properties[f]
1261 new_result = info.copy()
1262 new_result.update(force_properties)
1263
1264 # Extracted info may not be a video result (i.e.
1265 # info.get('_type', 'video') != video) but rather an url or
1266 # url_transparent. In such cases outer metadata (from ie_result)
1267 # should be propagated to inner one (info). For this to happen
1268 # _type of info should be overridden with url_transparent. This
1269 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1270 if new_result.get('_type') == 'url':
1271 new_result['_type'] = 'url_transparent'
1272
1273 return self.process_ie_result(
1274 new_result, download=download, extra_info=extra_info)
1275 elif result_type in ('playlist', 'multi_video'):
1276 # Protect from infinite recursion due to recursively nested playlists
1277 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1278 webpage_url = ie_result['webpage_url']
1279 if webpage_url in self._playlist_urls:
1280 self.to_screen(
1281 '[download] Skipping already downloaded playlist: %s'
1282 % ie_result.get('title') or ie_result.get('id'))
1283 return
1284
1285 self._playlist_level += 1
1286 self._playlist_urls.add(webpage_url)
1287 self._sanitize_thumbnails(ie_result)
1288 try:
1289 return self.__process_playlist(ie_result, download)
1290 finally:
1291 self._playlist_level -= 1
1292 if not self._playlist_level:
1293 self._playlist_urls.clear()
1294 elif result_type == 'compat_list':
1295 self.report_warning(
1296 'Extractor %s returned a compat_list result. '
1297 'It needs to be updated.' % ie_result.get('extractor'))
1298
1299 def _fixup(r):
1300 self.add_extra_info(
1301 r,
1302 {
1303 'extractor': ie_result['extractor'],
1304 'webpage_url': ie_result['webpage_url'],
1305 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1306 'extractor_key': ie_result['extractor_key'],
1307 }
1308 )
1309 return r
1310 ie_result['entries'] = [
1311 self.process_ie_result(_fixup(r), download, extra_info)
1312 for r in ie_result['entries']
1313 ]
1314 return ie_result
1315 else:
1316 raise Exception('Invalid result type: %s' % result_type)
1317
1318 def _ensure_dir_exists(self, path):
1319 return make_dir(path, self.report_error)
1320
1321 def __process_playlist(self, ie_result, download):
1322 # We process each entry in the playlist
1323 playlist = ie_result.get('title') or ie_result.get('id')
1324 self.to_screen('[download] Downloading playlist: %s' % playlist)
1325
1326 if 'entries' not in ie_result:
1327 raise EntryNotInPlaylist()
1328 incomplete_entries = bool(ie_result.get('requested_entries'))
1329 if incomplete_entries:
1330 def fill_missing_entries(entries, indexes):
1331 ret = [None] * max(*indexes)
1332 for i, entry in zip(indexes, entries):
1333 ret[i - 1] = entry
1334 return ret
1335 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1336
1337 playlist_results = []
1338
1339 playliststart = self.params.get('playliststart', 1)
1340 playlistend = self.params.get('playlistend')
1341 # For backwards compatibility, interpret -1 as whole list
1342 if playlistend == -1:
1343 playlistend = None
1344
1345 playlistitems_str = self.params.get('playlist_items')
1346 playlistitems = None
1347 if playlistitems_str is not None:
1348 def iter_playlistitems(format):
1349 for string_segment in format.split(','):
1350 if '-' in string_segment:
1351 start, end = string_segment.split('-')
1352 for item in range(int(start), int(end) + 1):
1353 yield int(item)
1354 else:
1355 yield int(string_segment)
1356 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1357
1358 ie_entries = ie_result['entries']
1359 msg = (
1360 'Downloading %d videos' if not isinstance(ie_entries, list)
1361 else 'Collected %d videos; downloading %%d of them' % len(ie_entries))
1362 if not isinstance(ie_entries, (list, PagedList)):
1363 ie_entries = LazyList(ie_entries)
1364
1365 entries = []
1366 for i in playlistitems or itertools.count(playliststart):
1367 if playlistitems is None and playlistend is not None and playlistend < i:
1368 break
1369 entry = None
1370 try:
1371 entry = ie_entries[i - 1]
1372 if entry is None:
1373 raise EntryNotInPlaylist()
1374 except (IndexError, EntryNotInPlaylist):
1375 if incomplete_entries:
1376 raise EntryNotInPlaylist()
1377 elif not playlistitems:
1378 break
1379 entries.append(entry)
1380 try:
1381 if entry is not None:
1382 self._match_entry(entry, incomplete=True, silent=True)
1383 except (ExistingVideoReached, RejectedVideoReached):
1384 break
1385 ie_result['entries'] = entries
1386
1387 # Save playlist_index before re-ordering
1388 entries = [
1389 ((playlistitems[i - 1] if playlistitems else i), entry)
1390 for i, entry in enumerate(entries, 1)
1391 if entry is not None]
1392 n_entries = len(entries)
1393
1394 if not playlistitems and (playliststart or playlistend):
1395 playlistitems = list(range(playliststart, playliststart + n_entries))
1396 ie_result['requested_entries'] = playlistitems
1397
1398 if self.params.get('allow_playlist_files', True):
1399 ie_copy = {
1400 'playlist': playlist,
1401 'playlist_id': ie_result.get('id'),
1402 'playlist_title': ie_result.get('title'),
1403 'playlist_uploader': ie_result.get('uploader'),
1404 'playlist_uploader_id': ie_result.get('uploader_id'),
1405 'playlist_index': 0,
1406 }
1407 ie_copy.update(dict(ie_result))
1408
1409 if self.params.get('writeinfojson', False):
1410 infofn = self.prepare_filename(ie_copy, 'pl_infojson')
1411 if not self._ensure_dir_exists(encodeFilename(infofn)):
1412 return
1413 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
1414 self.to_screen('[info] Playlist metadata is already present')
1415 else:
1416 self.to_screen('[info] Writing playlist metadata as JSON to: ' + infofn)
1417 try:
1418 write_json_file(self.filter_requested_info(ie_result, self.params.get('clean_infojson', True)), infofn)
1419 except (OSError, IOError):
1420 self.report_error('Cannot write playlist metadata to JSON file ' + infofn)
1421
1422 # TODO: This should be passed to ThumbnailsConvertor if necessary
1423 self._write_thumbnails(ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1424
1425 if self.params.get('writedescription', False):
1426 descfn = self.prepare_filename(ie_copy, 'pl_description')
1427 if not self._ensure_dir_exists(encodeFilename(descfn)):
1428 return
1429 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
1430 self.to_screen('[info] Playlist description is already present')
1431 elif ie_result.get('description') is None:
1432 self.report_warning('There\'s no playlist description to write.')
1433 else:
1434 try:
1435 self.to_screen('[info] Writing playlist description to: ' + descfn)
1436 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1437 descfile.write(ie_result['description'])
1438 except (OSError, IOError):
1439 self.report_error('Cannot write playlist description file ' + descfn)
1440 return
1441
1442 if self.params.get('playlistreverse', False):
1443 entries = entries[::-1]
1444 if self.params.get('playlistrandom', False):
1445 random.shuffle(entries)
1446
1447 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1448
1449 self.to_screen('[%s] playlist %s: %s' % (ie_result['extractor'], playlist, msg % n_entries))
1450 failures = 0
1451 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1452 for i, entry_tuple in enumerate(entries, 1):
1453 playlist_index, entry = entry_tuple
1454 if 'playlist_index' in self.params.get('compat_options', []):
1455 playlist_index = playlistitems[i - 1] if playlistitems else i
1456 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1457 # This __x_forwarded_for_ip thing is a bit ugly but requires
1458 # minimal changes
1459 if x_forwarded_for:
1460 entry['__x_forwarded_for_ip'] = x_forwarded_for
1461 extra = {
1462 'n_entries': n_entries,
1463 '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1464 'playlist_index': playlist_index,
1465 'playlist_autonumber': i,
1466 'playlist': playlist,
1467 'playlist_id': ie_result.get('id'),
1468 'playlist_title': ie_result.get('title'),
1469 'playlist_uploader': ie_result.get('uploader'),
1470 'playlist_uploader_id': ie_result.get('uploader_id'),
1471 'extractor': ie_result['extractor'],
1472 'webpage_url': ie_result['webpage_url'],
1473 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1474 'extractor_key': ie_result['extractor_key'],
1475 }
1476
1477 if self._match_entry(entry, incomplete=True) is not None:
1478 continue
1479
1480 entry_result = self.__process_iterable_entry(entry, download, extra)
1481 if not entry_result:
1482 failures += 1
1483 if failures >= max_failures:
1484 self.report_error(
1485 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1486 break
1487 # TODO: skip failed (empty) entries?
1488 playlist_results.append(entry_result)
1489 ie_result['entries'] = playlist_results
1490 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1491 return ie_result
1492
1493 @__handle_extraction_exceptions
1494 def __process_iterable_entry(self, entry, download, extra_info):
1495 return self.process_ie_result(
1496 entry, download=download, extra_info=extra_info)
1497
1498 def _build_format_filter(self, filter_spec):
1499 " Returns a function to filter the formats according to the filter_spec "
1500
1501 OPERATORS = {
1502 '<': operator.lt,
1503 '<=': operator.le,
1504 '>': operator.gt,
1505 '>=': operator.ge,
1506 '=': operator.eq,
1507 '!=': operator.ne,
1508 }
1509 operator_rex = re.compile(r'''(?x)\s*
1510 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1511 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1512 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1513 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1514 m = operator_rex.fullmatch(filter_spec)
1515 if m:
1516 try:
1517 comparison_value = int(m.group('value'))
1518 except ValueError:
1519 comparison_value = parse_filesize(m.group('value'))
1520 if comparison_value is None:
1521 comparison_value = parse_filesize(m.group('value') + 'B')
1522 if comparison_value is None:
1523 raise ValueError(
1524 'Invalid value %r in format specification %r' % (
1525 m.group('value'), filter_spec))
1526 op = OPERATORS[m.group('op')]
1527
1528 if not m:
1529 STR_OPERATORS = {
1530 '=': operator.eq,
1531 '^=': lambda attr, value: attr.startswith(value),
1532 '$=': lambda attr, value: attr.endswith(value),
1533 '*=': lambda attr, value: value in attr,
1534 }
1535 str_operator_rex = re.compile(r'''(?x)\s*
1536 (?P<key>[a-zA-Z0-9._-]+)\s*
1537 (?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1538 (?P<value>[a-zA-Z0-9._-]+)\s*
1539 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1540 m = str_operator_rex.fullmatch(filter_spec)
1541 if m:
1542 comparison_value = m.group('value')
1543 str_op = STR_OPERATORS[m.group('op')]
1544 if m.group('negation'):
1545 op = lambda attr, value: not str_op(attr, value)
1546 else:
1547 op = str_op
1548
1549 if not m:
1550 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1551
1552 def _filter(f):
1553 actual_value = f.get(m.group('key'))
1554 if actual_value is None:
1555 return m.group('none_inclusive')
1556 return op(actual_value, comparison_value)
1557 return _filter
1558
1559 def _default_format_spec(self, info_dict, download=True):
1560
1561 def can_merge():
1562 merger = FFmpegMergerPP(self)
1563 return merger.available and merger.can_merge()
1564
1565 prefer_best = (
1566 not self.params.get('simulate', False)
1567 and download
1568 and (
1569 not can_merge()
1570 or info_dict.get('is_live', False)
1571 or self.outtmpl_dict['default'] == '-'))
1572 compat = (
1573 prefer_best
1574 or self.params.get('allow_multiple_audio_streams', False)
1575 or 'format-spec' in self.params.get('compat_opts', []))
1576
1577 return (
1578 'best/bestvideo+bestaudio' if prefer_best
1579 else 'bestvideo*+bestaudio/best' if not compat
1580 else 'bestvideo+bestaudio/best')
1581
1582 def build_format_selector(self, format_spec):
1583 def syntax_error(note, start):
1584 message = (
1585 'Invalid format specification: '
1586 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1587 return SyntaxError(message)
1588
1589 PICKFIRST = 'PICKFIRST'
1590 MERGE = 'MERGE'
1591 SINGLE = 'SINGLE'
1592 GROUP = 'GROUP'
1593 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1594
1595 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1596 'video': self.params.get('allow_multiple_video_streams', False)}
1597
1598 check_formats = self.params.get('check_formats')
1599
1600 def _parse_filter(tokens):
1601 filter_parts = []
1602 for type, string, start, _, _ in tokens:
1603 if type == tokenize.OP and string == ']':
1604 return ''.join(filter_parts)
1605 else:
1606 filter_parts.append(string)
1607
1608 def _remove_unused_ops(tokens):
1609 # Remove operators that we don't use and join them with the surrounding strings
1610 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1611 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1612 last_string, last_start, last_end, last_line = None, None, None, None
1613 for type, string, start, end, line in tokens:
1614 if type == tokenize.OP and string == '[':
1615 if last_string:
1616 yield tokenize.NAME, last_string, last_start, last_end, last_line
1617 last_string = None
1618 yield type, string, start, end, line
1619 # everything inside brackets will be handled by _parse_filter
1620 for type, string, start, end, line in tokens:
1621 yield type, string, start, end, line
1622 if type == tokenize.OP and string == ']':
1623 break
1624 elif type == tokenize.OP and string in ALLOWED_OPS:
1625 if last_string:
1626 yield tokenize.NAME, last_string, last_start, last_end, last_line
1627 last_string = None
1628 yield type, string, start, end, line
1629 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1630 if not last_string:
1631 last_string = string
1632 last_start = start
1633 last_end = end
1634 else:
1635 last_string += string
1636 if last_string:
1637 yield tokenize.NAME, last_string, last_start, last_end, last_line
1638
1639 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1640 selectors = []
1641 current_selector = None
1642 for type, string, start, _, _ in tokens:
1643 # ENCODING is only defined in python 3.x
1644 if type == getattr(tokenize, 'ENCODING', None):
1645 continue
1646 elif type in [tokenize.NAME, tokenize.NUMBER]:
1647 current_selector = FormatSelector(SINGLE, string, [])
1648 elif type == tokenize.OP:
1649 if string == ')':
1650 if not inside_group:
1651 # ')' will be handled by the parentheses group
1652 tokens.restore_last_token()
1653 break
1654 elif inside_merge and string in ['/', ',']:
1655 tokens.restore_last_token()
1656 break
1657 elif inside_choice and string == ',':
1658 tokens.restore_last_token()
1659 break
1660 elif string == ',':
1661 if not current_selector:
1662 raise syntax_error('"," must follow a format selector', start)
1663 selectors.append(current_selector)
1664 current_selector = None
1665 elif string == '/':
1666 if not current_selector:
1667 raise syntax_error('"/" must follow a format selector', start)
1668 first_choice = current_selector
1669 second_choice = _parse_format_selection(tokens, inside_choice=True)
1670 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1671 elif string == '[':
1672 if not current_selector:
1673 current_selector = FormatSelector(SINGLE, 'best', [])
1674 format_filter = _parse_filter(tokens)
1675 current_selector.filters.append(format_filter)
1676 elif string == '(':
1677 if current_selector:
1678 raise syntax_error('Unexpected "("', start)
1679 group = _parse_format_selection(tokens, inside_group=True)
1680 current_selector = FormatSelector(GROUP, group, [])
1681 elif string == '+':
1682 if not current_selector:
1683 raise syntax_error('Unexpected "+"', start)
1684 selector_1 = current_selector
1685 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1686 if not selector_2:
1687 raise syntax_error('Expected a selector', start)
1688 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1689 else:
1690 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1691 elif type == tokenize.ENDMARKER:
1692 break
1693 if current_selector:
1694 selectors.append(current_selector)
1695 return selectors
1696
1697 def _merge(formats_pair):
1698 format_1, format_2 = formats_pair
1699
1700 formats_info = []
1701 formats_info.extend(format_1.get('requested_formats', (format_1,)))
1702 formats_info.extend(format_2.get('requested_formats', (format_2,)))
1703
1704 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
1705 get_no_more = {'video': False, 'audio': False}
1706 for (i, fmt_info) in enumerate(formats_info):
1707 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
1708 formats_info.pop(i)
1709 continue
1710 for aud_vid in ['audio', 'video']:
1711 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
1712 if get_no_more[aud_vid]:
1713 formats_info.pop(i)
1714 get_no_more[aud_vid] = True
1715
1716 if len(formats_info) == 1:
1717 return formats_info[0]
1718
1719 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
1720 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
1721
1722 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
1723 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
1724
1725 output_ext = self.params.get('merge_output_format')
1726 if not output_ext:
1727 if the_only_video:
1728 output_ext = the_only_video['ext']
1729 elif the_only_audio and not video_fmts:
1730 output_ext = the_only_audio['ext']
1731 else:
1732 output_ext = 'mkv'
1733
1734 new_dict = {
1735 'requested_formats': formats_info,
1736 'format': '+'.join(fmt_info.get('format') for fmt_info in formats_info),
1737 'format_id': '+'.join(fmt_info.get('format_id') for fmt_info in formats_info),
1738 'ext': output_ext,
1739 }
1740
1741 if the_only_video:
1742 new_dict.update({
1743 'width': the_only_video.get('width'),
1744 'height': the_only_video.get('height'),
1745 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
1746 'fps': the_only_video.get('fps'),
1747 'vcodec': the_only_video.get('vcodec'),
1748 'vbr': the_only_video.get('vbr'),
1749 'stretched_ratio': the_only_video.get('stretched_ratio'),
1750 })
1751
1752 if the_only_audio:
1753 new_dict.update({
1754 'acodec': the_only_audio.get('acodec'),
1755 'abr': the_only_audio.get('abr'),
1756 })
1757
1758 return new_dict
1759
1760 def _check_formats(formats):
1761 for f in formats:
1762 self.to_screen('[info] Testing format %s' % f['format_id'])
1763 temp_file = tempfile.NamedTemporaryFile(
1764 suffix='.tmp', delete=False,
1765 dir=self.get_output_path('temp') or None)
1766 temp_file.close()
1767 try:
1768 dl, _ = self.dl(temp_file.name, f, test=True)
1769 except (ExtractorError, IOError, OSError, ValueError) + network_exceptions:
1770 dl = False
1771 finally:
1772 if os.path.exists(temp_file.name):
1773 try:
1774 os.remove(temp_file.name)
1775 except OSError:
1776 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1777 if dl:
1778 yield f
1779 else:
1780 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1781
1782 def _build_selector_function(selector):
1783 if isinstance(selector, list): # ,
1784 fs = [_build_selector_function(s) for s in selector]
1785
1786 def selector_function(ctx):
1787 for f in fs:
1788 for format in f(ctx):
1789 yield format
1790 return selector_function
1791
1792 elif selector.type == GROUP: # ()
1793 selector_function = _build_selector_function(selector.selector)
1794
1795 elif selector.type == PICKFIRST: # /
1796 fs = [_build_selector_function(s) for s in selector.selector]
1797
1798 def selector_function(ctx):
1799 for f in fs:
1800 picked_formats = list(f(ctx))
1801 if picked_formats:
1802 return picked_formats
1803 return []
1804
1805 elif selector.type == SINGLE: # atom
1806 format_spec = selector.selector or 'best'
1807
1808 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
1809 if format_spec == 'all':
1810 def selector_function(ctx):
1811 formats = list(ctx['formats'])
1812 if check_formats:
1813 formats = _check_formats(formats)
1814 for f in formats:
1815 yield f
1816 elif format_spec == 'mergeall':
1817 def selector_function(ctx):
1818 formats = ctx['formats']
1819 if check_formats:
1820 formats = list(_check_formats(formats))
1821 if not formats:
1822 return
1823 merged_format = formats[-1]
1824 for f in formats[-2::-1]:
1825 merged_format = _merge((merged_format, f))
1826 yield merged_format
1827
1828 else:
1829 format_fallback, format_reverse, format_idx = False, True, 1
1830 mobj = re.match(
1831 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
1832 format_spec)
1833 if mobj is not None:
1834 format_idx = int_or_none(mobj.group('n'), default=1)
1835 format_reverse = mobj.group('bw')[0] == 'b'
1836 format_type = (mobj.group('type') or [None])[0]
1837 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
1838 format_modified = mobj.group('mod') is not None
1839
1840 format_fallback = not format_type and not format_modified # for b, w
1841 _filter_f = (
1842 (lambda f: f.get('%scodec' % format_type) != 'none')
1843 if format_type and format_modified # bv*, ba*, wv*, wa*
1844 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
1845 if format_type # bv, ba, wv, wa
1846 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
1847 if not format_modified # b, w
1848 else lambda f: True) # b*, w*
1849 filter_f = lambda f: _filter_f(f) and (
1850 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
1851 else:
1852 filter_f = ((lambda f: f.get('ext') == format_spec)
1853 if format_spec in ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] # extension
1854 else (lambda f: f.get('format_id') == format_spec)) # id
1855
1856 def selector_function(ctx):
1857 formats = list(ctx['formats'])
1858 if not formats:
1859 return
1860 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
1861 if format_fallback and ctx['incomplete_formats'] and not matches:
1862 # for extractors with incomplete formats (audio only (soundcloud)
1863 # or video only (imgur)) best/worst will fallback to
1864 # best/worst {video,audio}-only format
1865 matches = formats
1866 if format_reverse:
1867 matches = matches[::-1]
1868 if check_formats:
1869 matches = list(itertools.islice(_check_formats(matches), format_idx))
1870 n = len(matches)
1871 if -n <= format_idx - 1 < n:
1872 yield matches[format_idx - 1]
1873
1874 elif selector.type == MERGE: # +
1875 selector_1, selector_2 = map(_build_selector_function, selector.selector)
1876
1877 def selector_function(ctx):
1878 for pair in itertools.product(
1879 selector_1(copy.deepcopy(ctx)), selector_2(copy.deepcopy(ctx))):
1880 yield _merge(pair)
1881
1882 filters = [self._build_format_filter(f) for f in selector.filters]
1883
1884 def final_selector(ctx):
1885 ctx_copy = copy.deepcopy(ctx)
1886 for _filter in filters:
1887 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1888 return selector_function(ctx_copy)
1889 return final_selector
1890
1891 stream = io.BytesIO(format_spec.encode('utf-8'))
1892 try:
1893 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1894 except tokenize.TokenError:
1895 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1896
1897 class TokenIterator(object):
1898 def __init__(self, tokens):
1899 self.tokens = tokens
1900 self.counter = 0
1901
1902 def __iter__(self):
1903 return self
1904
1905 def __next__(self):
1906 if self.counter >= len(self.tokens):
1907 raise StopIteration()
1908 value = self.tokens[self.counter]
1909 self.counter += 1
1910 return value
1911
1912 next = __next__
1913
1914 def restore_last_token(self):
1915 self.counter -= 1
1916
1917 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1918 return _build_selector_function(parsed_selector)
1919
1920 def _calc_headers(self, info_dict):
1921 res = std_headers.copy()
1922
1923 add_headers = info_dict.get('http_headers')
1924 if add_headers:
1925 res.update(add_headers)
1926
1927 cookies = self._calc_cookies(info_dict)
1928 if cookies:
1929 res['Cookie'] = cookies
1930
1931 if 'X-Forwarded-For' not in res:
1932 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1933 if x_forwarded_for_ip:
1934 res['X-Forwarded-For'] = x_forwarded_for_ip
1935
1936 return res
1937
1938 def _calc_cookies(self, info_dict):
1939 pr = sanitized_Request(info_dict['url'])
1940 self.cookiejar.add_cookie_header(pr)
1941 return pr.get_header('Cookie')
1942
1943 def _sanitize_thumbnails(self, info_dict):
1944 thumbnails = info_dict.get('thumbnails')
1945 if thumbnails is None:
1946 thumbnail = info_dict.get('thumbnail')
1947 if thumbnail:
1948 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1949 if thumbnails:
1950 thumbnails.sort(key=lambda t: (
1951 t.get('preference') if t.get('preference') is not None else -1,
1952 t.get('width') if t.get('width') is not None else -1,
1953 t.get('height') if t.get('height') is not None else -1,
1954 t.get('id') if t.get('id') is not None else '',
1955 t.get('url')))
1956
1957 def test_thumbnail(t):
1958 self.to_screen('[info] Testing thumbnail %s' % t['id'])
1959 try:
1960 self.urlopen(HEADRequest(t['url']))
1961 except network_exceptions as err:
1962 self.to_screen('[info] Unable to connect to thumbnail %s URL "%s" - %s. Skipping...' % (
1963 t['id'], t['url'], error_to_compat_str(err)))
1964 return False
1965 return True
1966
1967 for i, t in enumerate(thumbnails):
1968 if t.get('id') is None:
1969 t['id'] = '%d' % i
1970 if t.get('width') and t.get('height'):
1971 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1972 t['url'] = sanitize_url(t['url'])
1973 if self.params.get('check_formats'):
1974 info_dict['thumbnails'] = reversed(LazyList(filter(test_thumbnail, thumbnails[::-1])))
1975
1976 def process_video_result(self, info_dict, download=True):
1977 assert info_dict.get('_type', 'video') == 'video'
1978
1979 if 'id' not in info_dict:
1980 raise ExtractorError('Missing "id" field in extractor result')
1981 if 'title' not in info_dict:
1982 raise ExtractorError('Missing "title" field in extractor result')
1983
1984 def report_force_conversion(field, field_not, conversion):
1985 self.report_warning(
1986 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
1987 % (field, field_not, conversion))
1988
1989 def sanitize_string_field(info, string_field):
1990 field = info.get(string_field)
1991 if field is None or isinstance(field, compat_str):
1992 return
1993 report_force_conversion(string_field, 'a string', 'string')
1994 info[string_field] = compat_str(field)
1995
1996 def sanitize_numeric_fields(info):
1997 for numeric_field in self._NUMERIC_FIELDS:
1998 field = info.get(numeric_field)
1999 if field is None or isinstance(field, compat_numeric_types):
2000 continue
2001 report_force_conversion(numeric_field, 'numeric', 'int')
2002 info[numeric_field] = int_or_none(field)
2003
2004 sanitize_string_field(info_dict, 'id')
2005 sanitize_numeric_fields(info_dict)
2006
2007 if 'playlist' not in info_dict:
2008 # It isn't part of a playlist
2009 info_dict['playlist'] = None
2010 info_dict['playlist_index'] = None
2011
2012 self._sanitize_thumbnails(info_dict)
2013
2014 if self.params.get('list_thumbnails'):
2015 self.list_thumbnails(info_dict)
2016 return
2017
2018 thumbnail = info_dict.get('thumbnail')
2019 thumbnails = info_dict.get('thumbnails')
2020 if thumbnail:
2021 info_dict['thumbnail'] = sanitize_url(thumbnail)
2022 elif thumbnails:
2023 info_dict['thumbnail'] = thumbnails[-1]['url']
2024
2025 if 'display_id' not in info_dict and 'id' in info_dict:
2026 info_dict['display_id'] = info_dict['id']
2027
2028 for ts_key, date_key in (
2029 ('timestamp', 'upload_date'),
2030 ('release_timestamp', 'release_date'),
2031 ):
2032 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2033 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2034 # see http://bugs.python.org/issue1646728)
2035 try:
2036 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2037 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2038 except (ValueError, OverflowError, OSError):
2039 pass
2040
2041 # Auto generate title fields corresponding to the *_number fields when missing
2042 # in order to always have clean titles. This is very common for TV series.
2043 for field in ('chapter', 'season', 'episode'):
2044 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2045 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2046
2047 for cc_kind in ('subtitles', 'automatic_captions'):
2048 cc = info_dict.get(cc_kind)
2049 if cc:
2050 for _, subtitle in cc.items():
2051 for subtitle_format in subtitle:
2052 if subtitle_format.get('url'):
2053 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2054 if subtitle_format.get('ext') is None:
2055 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2056
2057 automatic_captions = info_dict.get('automatic_captions')
2058 subtitles = info_dict.get('subtitles')
2059
2060 if self.params.get('listsubtitles', False):
2061 if 'automatic_captions' in info_dict:
2062 self.list_subtitles(
2063 info_dict['id'], automatic_captions, 'automatic captions')
2064 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2065 return
2066
2067 info_dict['requested_subtitles'] = self.process_subtitles(
2068 info_dict['id'], subtitles, automatic_captions)
2069
2070 # We now pick which formats have to be downloaded
2071 if info_dict.get('formats') is None:
2072 # There's only one format available
2073 formats = [info_dict]
2074 else:
2075 formats = info_dict['formats']
2076
2077 if not formats:
2078 if not self.params.get('ignore_no_formats_error'):
2079 raise ExtractorError('No video formats found!')
2080 else:
2081 self.report_warning('No video formats found!')
2082
2083 def is_wellformed(f):
2084 url = f.get('url')
2085 if not url:
2086 self.report_warning(
2087 '"url" field is missing or empty - skipping format, '
2088 'there is an error in extractor')
2089 return False
2090 if isinstance(url, bytes):
2091 sanitize_string_field(f, 'url')
2092 return True
2093
2094 # Filter out malformed formats for better extraction robustness
2095 formats = list(filter(is_wellformed, formats))
2096
2097 formats_dict = {}
2098
2099 # We check that all the formats have the format and format_id fields
2100 for i, format in enumerate(formats):
2101 sanitize_string_field(format, 'format_id')
2102 sanitize_numeric_fields(format)
2103 format['url'] = sanitize_url(format['url'])
2104 if not format.get('format_id'):
2105 format['format_id'] = compat_str(i)
2106 else:
2107 # Sanitize format_id from characters used in format selector expression
2108 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2109 format_id = format['format_id']
2110 if format_id not in formats_dict:
2111 formats_dict[format_id] = []
2112 formats_dict[format_id].append(format)
2113
2114 # Make sure all formats have unique format_id
2115 for format_id, ambiguous_formats in formats_dict.items():
2116 if len(ambiguous_formats) > 1:
2117 for i, format in enumerate(ambiguous_formats):
2118 format['format_id'] = '%s-%d' % (format_id, i)
2119
2120 for i, format in enumerate(formats):
2121 if format.get('format') is None:
2122 format['format'] = '{id} - {res}{note}'.format(
2123 id=format['format_id'],
2124 res=self.format_resolution(format),
2125 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
2126 )
2127 # Automatically determine file extension if missing
2128 if format.get('ext') is None:
2129 format['ext'] = determine_ext(format['url']).lower()
2130 # Automatically determine protocol if missing (useful for format
2131 # selection purposes)
2132 if format.get('protocol') is None:
2133 format['protocol'] = determine_protocol(format)
2134 # Add HTTP headers, so that external programs can use them from the
2135 # json output
2136 full_format_info = info_dict.copy()
2137 full_format_info.update(format)
2138 format['http_headers'] = self._calc_headers(full_format_info)
2139 # Remove private housekeeping stuff
2140 if '__x_forwarded_for_ip' in info_dict:
2141 del info_dict['__x_forwarded_for_ip']
2142
2143 # TODO Central sorting goes here
2144
2145 if formats and formats[0] is not info_dict:
2146 # only set the 'formats' fields if the original info_dict list them
2147 # otherwise we end up with a circular reference, the first (and unique)
2148 # element in the 'formats' field in info_dict is info_dict itself,
2149 # which can't be exported to json
2150 info_dict['formats'] = formats
2151
2152 info_dict, _ = self.pre_process(info_dict)
2153
2154 if self.params.get('listformats'):
2155 if not info_dict.get('formats'):
2156 raise ExtractorError('No video formats found', expected=True)
2157 self.list_formats(info_dict)
2158 return
2159
2160 format_selector = self.format_selector
2161 if format_selector is None:
2162 req_format = self._default_format_spec(info_dict, download=download)
2163 self.write_debug('Default format spec: %s' % req_format)
2164 format_selector = self.build_format_selector(req_format)
2165
2166 # While in format selection we may need to have an access to the original
2167 # format set in order to calculate some metrics or do some processing.
2168 # For now we need to be able to guess whether original formats provided
2169 # by extractor are incomplete or not (i.e. whether extractor provides only
2170 # video-only or audio-only formats) for proper formats selection for
2171 # extractors with such incomplete formats (see
2172 # https://github.com/ytdl-org/youtube-dl/pull/5556).
2173 # Since formats may be filtered during format selection and may not match
2174 # the original formats the results may be incorrect. Thus original formats
2175 # or pre-calculated metrics should be passed to format selection routines
2176 # as well.
2177 # We will pass a context object containing all necessary additional data
2178 # instead of just formats.
2179 # This fixes incorrect format selection issue (see
2180 # https://github.com/ytdl-org/youtube-dl/issues/10083).
2181 incomplete_formats = (
2182 # All formats are video-only or
2183 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2184 # all formats are audio-only
2185 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
2186
2187 ctx = {
2188 'formats': formats,
2189 'incomplete_formats': incomplete_formats,
2190 }
2191
2192 formats_to_download = list(format_selector(ctx))
2193 if not formats_to_download:
2194 if not self.params.get('ignore_no_formats_error'):
2195 raise ExtractorError('Requested format is not available', expected=True)
2196 else:
2197 self.report_warning('Requested format is not available')
2198 # Process what we can, even without any available formats.
2199 self.process_info(dict(info_dict))
2200 elif download:
2201 self.to_screen(
2202 '[info] %s: Downloading %d format(s): %s' % (
2203 info_dict['id'], len(formats_to_download),
2204 ", ".join([f['format_id'] for f in formats_to_download])))
2205 for fmt in formats_to_download:
2206 new_info = dict(info_dict)
2207 # Save a reference to the original info_dict so that it can be modified in process_info if needed
2208 new_info['__original_infodict'] = info_dict
2209 new_info.update(fmt)
2210 self.process_info(new_info)
2211 # We update the info dict with the best quality format (backwards compatibility)
2212 if formats_to_download:
2213 info_dict.update(formats_to_download[-1])
2214 return info_dict
2215
2216 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2217 """Select the requested subtitles and their format"""
2218 available_subs = {}
2219 if normal_subtitles and self.params.get('writesubtitles'):
2220 available_subs.update(normal_subtitles)
2221 if automatic_captions and self.params.get('writeautomaticsub'):
2222 for lang, cap_info in automatic_captions.items():
2223 if lang not in available_subs:
2224 available_subs[lang] = cap_info
2225
2226 if (not self.params.get('writesubtitles') and not
2227 self.params.get('writeautomaticsub') or not
2228 available_subs):
2229 return None
2230
2231 all_sub_langs = available_subs.keys()
2232 if self.params.get('allsubtitles', False):
2233 requested_langs = all_sub_langs
2234 elif self.params.get('subtitleslangs', False):
2235 requested_langs = set()
2236 for lang in self.params.get('subtitleslangs'):
2237 if lang == 'all':
2238 requested_langs.update(all_sub_langs)
2239 continue
2240 discard = lang[0] == '-'
2241 if discard:
2242 lang = lang[1:]
2243 current_langs = filter(re.compile(lang + '$').match, all_sub_langs)
2244 if discard:
2245 for lang in current_langs:
2246 requested_langs.discard(lang)
2247 else:
2248 requested_langs.update(current_langs)
2249 elif 'en' in available_subs:
2250 requested_langs = ['en']
2251 else:
2252 requested_langs = [list(all_sub_langs)[0]]
2253 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2254
2255 formats_query = self.params.get('subtitlesformat', 'best')
2256 formats_preference = formats_query.split('/') if formats_query else []
2257 subs = {}
2258 for lang in requested_langs:
2259 formats = available_subs.get(lang)
2260 if formats is None:
2261 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
2262 continue
2263 for ext in formats_preference:
2264 if ext == 'best':
2265 f = formats[-1]
2266 break
2267 matches = list(filter(lambda f: f['ext'] == ext, formats))
2268 if matches:
2269 f = matches[-1]
2270 break
2271 else:
2272 f = formats[-1]
2273 self.report_warning(
2274 'No subtitle format found matching "%s" for language %s, '
2275 'using %s' % (formats_query, lang, f['ext']))
2276 subs[lang] = f
2277 return subs
2278
2279 def __forced_printings(self, info_dict, filename, incomplete):
2280 def print_mandatory(field, actual_field=None):
2281 if actual_field is None:
2282 actual_field = field
2283 if (self.params.get('force%s' % field, False)
2284 and (not incomplete or info_dict.get(actual_field) is not None)):
2285 self.to_stdout(info_dict[actual_field])
2286
2287 def print_optional(field):
2288 if (self.params.get('force%s' % field, False)
2289 and info_dict.get(field) is not None):
2290 self.to_stdout(info_dict[field])
2291
2292 info_dict = info_dict.copy()
2293 if filename is not None:
2294 info_dict['filename'] = filename
2295 if info_dict.get('requested_formats') is not None:
2296 # For RTMP URLs, also include the playpath
2297 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2298 elif 'url' in info_dict:
2299 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2300
2301 for tmpl in self.params.get('forceprint', []):
2302 if re.match(r'\w+$', tmpl):
2303 tmpl = '%({})s'.format(tmpl)
2304 tmpl, info_copy = self.prepare_outtmpl(tmpl, info_dict)
2305 self.to_stdout(tmpl % info_copy)
2306
2307 print_mandatory('title')
2308 print_mandatory('id')
2309 print_mandatory('url', 'urls')
2310 print_optional('thumbnail')
2311 print_optional('description')
2312 print_optional('filename')
2313 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
2314 self.to_stdout(formatSeconds(info_dict['duration']))
2315 print_mandatory('format')
2316
2317 if self.params.get('forcejson', False):
2318 self.post_extract(info_dict)
2319 self.to_stdout(json.dumps(info_dict, default=repr))
2320
2321 def dl(self, name, info, subtitle=False, test=False):
2322
2323 if test:
2324 verbose = self.params.get('verbose')
2325 params = {
2326 'test': True,
2327 'quiet': not verbose,
2328 'verbose': verbose,
2329 'noprogress': not verbose,
2330 'nopart': True,
2331 'skip_unavailable_fragments': False,
2332 'keep_fragments': False,
2333 'overwrites': True,
2334 '_no_ytdl_file': True,
2335 }
2336 else:
2337 params = self.params
2338 fd = get_suitable_downloader(info, params)(self, params)
2339 if not test:
2340 for ph in self._progress_hooks:
2341 fd.add_progress_hook(ph)
2342 urls = '", "'.join([f['url'] for f in info.get('requested_formats', [])] or [info['url']])
2343 self.write_debug('Invoking downloader on "%s"' % urls)
2344 new_info = dict(info)
2345 if new_info.get('http_headers') is None:
2346 new_info['http_headers'] = self._calc_headers(new_info)
2347 return fd.download(name, new_info, subtitle)
2348
2349 def process_info(self, info_dict):
2350 """Process a single resolved IE result."""
2351
2352 assert info_dict.get('_type', 'video') == 'video'
2353
2354 info_dict.setdefault('__postprocessors', [])
2355
2356 max_downloads = self.params.get('max_downloads')
2357 if max_downloads is not None:
2358 if self._num_downloads >= int(max_downloads):
2359 raise MaxDownloadsReached()
2360
2361 # TODO: backward compatibility, to be removed
2362 info_dict['fulltitle'] = info_dict['title']
2363
2364 if 'format' not in info_dict and 'ext' in info_dict:
2365 info_dict['format'] = info_dict['ext']
2366
2367 if self._match_entry(info_dict) is not None:
2368 return
2369
2370 self.post_extract(info_dict)
2371 self._num_downloads += 1
2372
2373 # info_dict['_filename'] needs to be set for backward compatibility
2374 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2375 temp_filename = self.prepare_filename(info_dict, 'temp')
2376 files_to_move = {}
2377
2378 # Forced printings
2379 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2380
2381 if self.params.get('simulate', False):
2382 if self.params.get('force_write_download_archive', False):
2383 self.record_download_archive(info_dict)
2384
2385 # Do nothing else if in simulate mode
2386 return
2387
2388 if full_filename is None:
2389 return
2390
2391 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2392 return
2393 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2394 return
2395
2396 if self.params.get('writedescription', False):
2397 descfn = self.prepare_filename(info_dict, 'description')
2398 if not self._ensure_dir_exists(encodeFilename(descfn)):
2399 return
2400 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
2401 self.to_screen('[info] Video description is already present')
2402 elif info_dict.get('description') is None:
2403 self.report_warning('There\'s no description to write.')
2404 else:
2405 try:
2406 self.to_screen('[info] Writing video description to: ' + descfn)
2407 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
2408 descfile.write(info_dict['description'])
2409 except (OSError, IOError):
2410 self.report_error('Cannot write description file ' + descfn)
2411 return
2412
2413 if self.params.get('writeannotations', False):
2414 annofn = self.prepare_filename(info_dict, 'annotation')
2415 if not self._ensure_dir_exists(encodeFilename(annofn)):
2416 return
2417 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2418 self.to_screen('[info] Video annotations are already present')
2419 elif not info_dict.get('annotations'):
2420 self.report_warning('There are no annotations to write.')
2421 else:
2422 try:
2423 self.to_screen('[info] Writing video annotations to: ' + annofn)
2424 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2425 annofile.write(info_dict['annotations'])
2426 except (KeyError, TypeError):
2427 self.report_warning('There are no annotations to write.')
2428 except (OSError, IOError):
2429 self.report_error('Cannot write annotations file: ' + annofn)
2430 return
2431
2432 subtitles_are_requested = any([self.params.get('writesubtitles', False),
2433 self.params.get('writeautomaticsub')])
2434
2435 if subtitles_are_requested and info_dict.get('requested_subtitles'):
2436 # subtitles download errors are already managed as troubles in relevant IE
2437 # that way it will silently go on when used with unsupporting IE
2438 subtitles = info_dict['requested_subtitles']
2439 # ie = self.get_info_extractor(info_dict['extractor_key'])
2440 for sub_lang, sub_info in subtitles.items():
2441 sub_format = sub_info['ext']
2442 sub_filename = subtitles_filename(temp_filename, sub_lang, sub_format, info_dict.get('ext'))
2443 sub_filename_final = subtitles_filename(
2444 self.prepare_filename(info_dict, 'subtitle'), sub_lang, sub_format, info_dict.get('ext'))
2445 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(sub_filename)):
2446 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
2447 sub_info['filepath'] = sub_filename
2448 files_to_move[sub_filename] = sub_filename_final
2449 else:
2450 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
2451 if sub_info.get('data') is not None:
2452 try:
2453 # Use newline='' to prevent conversion of newline characters
2454 # See https://github.com/ytdl-org/youtube-dl/issues/10268
2455 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
2456 subfile.write(sub_info['data'])
2457 sub_info['filepath'] = sub_filename
2458 files_to_move[sub_filename] = sub_filename_final
2459 except (OSError, IOError):
2460 self.report_error('Cannot write subtitles file ' + sub_filename)
2461 return
2462 else:
2463 try:
2464 self.dl(sub_filename, sub_info.copy(), subtitle=True)
2465 sub_info['filepath'] = sub_filename
2466 files_to_move[sub_filename] = sub_filename_final
2467 except (ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
2468 self.report_warning('Unable to download subtitle for "%s": %s' %
2469 (sub_lang, error_to_compat_str(err)))
2470 continue
2471
2472 if self.params.get('writeinfojson', False):
2473 infofn = self.prepare_filename(info_dict, 'infojson')
2474 if not self._ensure_dir_exists(encodeFilename(infofn)):
2475 return
2476 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
2477 self.to_screen('[info] Video metadata is already present')
2478 else:
2479 self.to_screen('[info] Writing video metadata as JSON to: ' + infofn)
2480 try:
2481 write_json_file(self.filter_requested_info(info_dict, self.params.get('clean_infojson', True)), infofn)
2482 except (OSError, IOError):
2483 self.report_error('Cannot write video metadata to JSON file ' + infofn)
2484 return
2485 info_dict['__infojson_filename'] = infofn
2486
2487 for thumb_ext in self._write_thumbnails(info_dict, temp_filename):
2488 thumb_filename_temp = replace_extension(temp_filename, thumb_ext, info_dict.get('ext'))
2489 thumb_filename = replace_extension(
2490 self.prepare_filename(info_dict, 'thumbnail'), thumb_ext, info_dict.get('ext'))
2491 files_to_move[thumb_filename_temp] = thumb_filename
2492
2493 # Write internet shortcut files
2494 url_link = webloc_link = desktop_link = False
2495 if self.params.get('writelink', False):
2496 if sys.platform == "darwin": # macOS.
2497 webloc_link = True
2498 elif sys.platform.startswith("linux"):
2499 desktop_link = True
2500 else: # if sys.platform in ['win32', 'cygwin']:
2501 url_link = True
2502 if self.params.get('writeurllink', False):
2503 url_link = True
2504 if self.params.get('writewebloclink', False):
2505 webloc_link = True
2506 if self.params.get('writedesktoplink', False):
2507 desktop_link = True
2508
2509 if url_link or webloc_link or desktop_link:
2510 if 'webpage_url' not in info_dict:
2511 self.report_error('Cannot write internet shortcut file because the "webpage_url" field is missing in the media information')
2512 return
2513 ascii_url = iri_to_uri(info_dict['webpage_url'])
2514
2515 def _write_link_file(extension, template, newline, embed_filename):
2516 linkfn = replace_extension(full_filename, extension, info_dict.get('ext'))
2517 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2518 self.to_screen('[info] Internet shortcut is already present')
2519 else:
2520 try:
2521 self.to_screen('[info] Writing internet shortcut to: ' + linkfn)
2522 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', newline=newline) as linkfile:
2523 template_vars = {'url': ascii_url}
2524 if embed_filename:
2525 template_vars['filename'] = linkfn[:-(len(extension) + 1)]
2526 linkfile.write(template % template_vars)
2527 except (OSError, IOError):
2528 self.report_error('Cannot write internet shortcut ' + linkfn)
2529 return False
2530 return True
2531
2532 if url_link:
2533 if not _write_link_file('url', DOT_URL_LINK_TEMPLATE, '\r\n', embed_filename=False):
2534 return
2535 if webloc_link:
2536 if not _write_link_file('webloc', DOT_WEBLOC_LINK_TEMPLATE, '\n', embed_filename=False):
2537 return
2538 if desktop_link:
2539 if not _write_link_file('desktop', DOT_DESKTOP_LINK_TEMPLATE, '\n', embed_filename=True):
2540 return
2541
2542 try:
2543 info_dict, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2544 except PostProcessingError as err:
2545 self.report_error('Preprocessing: %s' % str(err))
2546 return
2547
2548 must_record_download_archive = False
2549 if self.params.get('skip_download', False):
2550 info_dict['filepath'] = temp_filename
2551 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2552 info_dict['__files_to_move'] = files_to_move
2553 info_dict = self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict)
2554 else:
2555 # Download
2556 try:
2557
2558 def existing_file(*filepaths):
2559 ext = info_dict.get('ext')
2560 final_ext = self.params.get('final_ext', ext)
2561 existing_files = []
2562 for file in orderedSet(filepaths):
2563 if final_ext != ext:
2564 converted = replace_extension(file, final_ext, ext)
2565 if os.path.exists(encodeFilename(converted)):
2566 existing_files.append(converted)
2567 if os.path.exists(encodeFilename(file)):
2568 existing_files.append(file)
2569
2570 if not existing_files or self.params.get('overwrites', False):
2571 for file in orderedSet(existing_files):
2572 self.report_file_delete(file)
2573 os.remove(encodeFilename(file))
2574 return None
2575
2576 self.report_file_already_downloaded(existing_files[0])
2577 info_dict['ext'] = os.path.splitext(existing_files[0])[1][1:]
2578 return existing_files[0]
2579
2580 success = True
2581 if info_dict.get('requested_formats') is not None:
2582
2583 def compatible_formats(formats):
2584 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2585 video_formats = [format for format in formats if format.get('vcodec') != 'none']
2586 audio_formats = [format for format in formats if format.get('acodec') != 'none']
2587 if len(video_formats) > 2 or len(audio_formats) > 2:
2588 return False
2589
2590 # Check extension
2591 exts = set(format.get('ext') for format in formats)
2592 COMPATIBLE_EXTS = (
2593 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
2594 set(('webm',)),
2595 )
2596 for ext_sets in COMPATIBLE_EXTS:
2597 if ext_sets.issuperset(exts):
2598 return True
2599 # TODO: Check acodec/vcodec
2600 return False
2601
2602 requested_formats = info_dict['requested_formats']
2603 old_ext = info_dict['ext']
2604 if self.params.get('merge_output_format') is None:
2605 if not compatible_formats(requested_formats):
2606 info_dict['ext'] = 'mkv'
2607 self.report_warning(
2608 'Requested formats are incompatible for merge and will be merged into mkv.')
2609 if (info_dict['ext'] == 'webm'
2610 and self.params.get('writethumbnail', False)
2611 and info_dict.get('thumbnails')):
2612 info_dict['ext'] = 'mkv'
2613 self.report_warning(
2614 'webm doesn\'t support embedding a thumbnail, mkv will be used.')
2615
2616 def correct_ext(filename):
2617 filename_real_ext = os.path.splitext(filename)[1][1:]
2618 filename_wo_ext = (
2619 os.path.splitext(filename)[0]
2620 if filename_real_ext == old_ext
2621 else filename)
2622 return '%s.%s' % (filename_wo_ext, info_dict['ext'])
2623
2624 # Ensure filename always has a correct extension for successful merge
2625 full_filename = correct_ext(full_filename)
2626 temp_filename = correct_ext(temp_filename)
2627 dl_filename = existing_file(full_filename, temp_filename)
2628 info_dict['__real_download'] = False
2629
2630 _protocols = set(determine_protocol(f) for f in requested_formats)
2631 if len(_protocols) == 1:
2632 info_dict['protocol'] = _protocols.pop()
2633 directly_mergable = (
2634 'no-direct-merge' not in self.params.get('compat_opts', [])
2635 and info_dict.get('protocol') is not None # All requested formats have same protocol
2636 and not self.params.get('allow_unplayable_formats')
2637 and get_suitable_downloader(info_dict, self.params).__name__ == 'FFmpegFD')
2638 if directly_mergable:
2639 info_dict['url'] = requested_formats[0]['url']
2640 # Treat it as a single download
2641 dl_filename = existing_file(full_filename, temp_filename)
2642 if dl_filename is None:
2643 success, real_download = self.dl(temp_filename, info_dict)
2644 info_dict['__real_download'] = real_download
2645 else:
2646 downloaded = []
2647 merger = FFmpegMergerPP(self)
2648 if self.params.get('allow_unplayable_formats'):
2649 self.report_warning(
2650 'You have requested merging of multiple formats '
2651 'while also allowing unplayable formats to be downloaded. '
2652 'The formats won\'t be merged to prevent data corruption.')
2653 elif not merger.available:
2654 self.report_warning(
2655 'You have requested merging of multiple formats but ffmpeg is not installed. '
2656 'The formats won\'t be merged.')
2657
2658 if dl_filename is None:
2659 for f in requested_formats:
2660 new_info = dict(info_dict)
2661 del new_info['requested_formats']
2662 new_info.update(f)
2663 fname = prepend_extension(
2664 self.prepare_filename(new_info, 'temp'),
2665 'f%s' % f['format_id'], new_info['ext'])
2666 if not self._ensure_dir_exists(fname):
2667 return
2668 downloaded.append(fname)
2669 partial_success, real_download = self.dl(fname, new_info)
2670 info_dict['__real_download'] = info_dict['__real_download'] or real_download
2671 success = success and partial_success
2672 if merger.available and not self.params.get('allow_unplayable_formats'):
2673 info_dict['__postprocessors'].append(merger)
2674 info_dict['__files_to_merge'] = downloaded
2675 # Even if there were no downloads, it is being merged only now
2676 info_dict['__real_download'] = True
2677 else:
2678 for file in downloaded:
2679 files_to_move[file] = None
2680 else:
2681 # Just a single file
2682 dl_filename = existing_file(full_filename, temp_filename)
2683 if dl_filename is None:
2684 success, real_download = self.dl(temp_filename, info_dict)
2685 info_dict['__real_download'] = real_download
2686
2687 dl_filename = dl_filename or temp_filename
2688 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2689
2690 except network_exceptions as err:
2691 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
2692 return
2693 except (OSError, IOError) as err:
2694 raise UnavailableVideoError(err)
2695 except (ContentTooShortError, ) as err:
2696 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
2697 return
2698
2699 if success and full_filename != '-':
2700
2701 def fixup():
2702 do_fixup = True
2703 fixup_policy = self.params.get('fixup')
2704 vid = info_dict['id']
2705
2706 if fixup_policy in ('ignore', 'never'):
2707 return
2708 elif fixup_policy == 'warn':
2709 do_fixup = False
2710 elif fixup_policy != 'force':
2711 assert fixup_policy in ('detect_or_warn', None)
2712 if not info_dict.get('__real_download'):
2713 do_fixup = False
2714
2715 def ffmpeg_fixup(cndn, msg, cls):
2716 if not cndn:
2717 return
2718 if not do_fixup:
2719 self.report_warning(f'{vid}: {msg}')
2720 return
2721 pp = cls(self)
2722 if pp.available:
2723 info_dict['__postprocessors'].append(pp)
2724 else:
2725 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
2726
2727 stretched_ratio = info_dict.get('stretched_ratio')
2728 ffmpeg_fixup(
2729 stretched_ratio not in (1, None),
2730 f'Non-uniform pixel ratio {stretched_ratio}',
2731 FFmpegFixupStretchedPP)
2732
2733 ffmpeg_fixup(
2734 (info_dict.get('requested_formats') is None
2735 and info_dict.get('container') == 'm4a_dash'
2736 and info_dict.get('ext') == 'm4a'),
2737 'writing DASH m4a. Only some players support this container',
2738 FFmpegFixupM4aPP)
2739
2740 downloader = (get_suitable_downloader(info_dict, self.params).__name__
2741 if 'protocol' in info_dict else None)
2742 ffmpeg_fixup(downloader == 'HlsFD', 'malformed AAC bitstream detected', FFmpegFixupM3u8PP)
2743 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'malformed timestamps detected', FFmpegFixupTimestampPP)
2744 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'malformed duration detected', FFmpegFixupDurationPP)
2745
2746 fixup()
2747 try:
2748 info_dict = self.post_process(dl_filename, info_dict, files_to_move)
2749 except PostProcessingError as err:
2750 self.report_error('Postprocessing: %s' % str(err))
2751 return
2752 try:
2753 for ph in self._post_hooks:
2754 ph(info_dict['filepath'])
2755 except Exception as err:
2756 self.report_error('post hooks: %s' % str(err))
2757 return
2758 must_record_download_archive = True
2759
2760 if must_record_download_archive or self.params.get('force_write_download_archive', False):
2761 self.record_download_archive(info_dict)
2762 max_downloads = self.params.get('max_downloads')
2763 if max_downloads is not None and self._num_downloads >= int(max_downloads):
2764 raise MaxDownloadsReached()
2765
2766 def download(self, url_list):
2767 """Download a given list of URLs."""
2768 outtmpl = self.outtmpl_dict['default']
2769 if (len(url_list) > 1
2770 and outtmpl != '-'
2771 and '%' not in outtmpl
2772 and self.params.get('max_downloads') != 1):
2773 raise SameFileError(outtmpl)
2774
2775 for url in url_list:
2776 try:
2777 # It also downloads the videos
2778 res = self.extract_info(
2779 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2780 except UnavailableVideoError:
2781 self.report_error('unable to download video')
2782 except MaxDownloadsReached:
2783 self.to_screen('[info] Maximum number of downloaded files reached')
2784 raise
2785 except ExistingVideoReached:
2786 self.to_screen('[info] Encountered a file that is already in the archive, stopping due to --break-on-existing')
2787 raise
2788 except RejectedVideoReached:
2789 self.to_screen('[info] Encountered a file that did not match filter, stopping due to --break-on-reject')
2790 raise
2791 else:
2792 if self.params.get('dump_single_json', False):
2793 self.post_extract(res)
2794 self.to_stdout(json.dumps(res, default=repr))
2795
2796 return self._download_retcode
2797
2798 def download_with_info_file(self, info_filename):
2799 with contextlib.closing(fileinput.FileInput(
2800 [info_filename], mode='r',
2801 openhook=fileinput.hook_encoded('utf-8'))) as f:
2802 # FileInput doesn't have a read method, we can't call json.load
2803 info = self.filter_requested_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
2804 try:
2805 self.process_ie_result(info, download=True)
2806 except (DownloadError, EntryNotInPlaylist):
2807 webpage_url = info.get('webpage_url')
2808 if webpage_url is not None:
2809 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2810 return self.download([webpage_url])
2811 else:
2812 raise
2813 return self._download_retcode
2814
2815 @staticmethod
2816 def filter_requested_info(info_dict, actually_filter=True):
2817 remove_keys = ['__original_infodict'] # Always remove this since this may contain a copy of the entire dict
2818 keep_keys = ['_type'], # Always keep this to facilitate load-info-json
2819 if actually_filter:
2820 remove_keys += ('requested_formats', 'requested_subtitles', 'requested_entries', 'filepath', 'entries', 'original_url')
2821 empty_values = (None, {}, [], set(), tuple())
2822 reject = lambda k, v: k not in keep_keys and (
2823 k.startswith('_') or k in remove_keys or v in empty_values)
2824 else:
2825 info_dict['epoch'] = int(time.time())
2826 reject = lambda k, v: k in remove_keys
2827 filter_fn = lambda obj: (
2828 list(map(filter_fn, obj)) if isinstance(obj, (LazyList, list, tuple, set))
2829 else obj if not isinstance(obj, dict)
2830 else dict((k, filter_fn(v)) for k, v in obj.items() if not reject(k, v)))
2831 return filter_fn(info_dict)
2832
2833 def run_pp(self, pp, infodict):
2834 files_to_delete = []
2835 if '__files_to_move' not in infodict:
2836 infodict['__files_to_move'] = {}
2837 files_to_delete, infodict = pp.run(infodict)
2838 if not files_to_delete:
2839 return infodict
2840
2841 if self.params.get('keepvideo', False):
2842 for f in files_to_delete:
2843 infodict['__files_to_move'].setdefault(f, '')
2844 else:
2845 for old_filename in set(files_to_delete):
2846 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2847 try:
2848 os.remove(encodeFilename(old_filename))
2849 except (IOError, OSError):
2850 self.report_warning('Unable to remove downloaded original file')
2851 if old_filename in infodict['__files_to_move']:
2852 del infodict['__files_to_move'][old_filename]
2853 return infodict
2854
2855 @staticmethod
2856 def post_extract(info_dict):
2857 def actual_post_extract(info_dict):
2858 if info_dict.get('_type') in ('playlist', 'multi_video'):
2859 for video_dict in info_dict.get('entries', {}):
2860 actual_post_extract(video_dict or {})
2861 return
2862
2863 post_extractor = info_dict.get('__post_extractor') or (lambda: {})
2864 extra = post_extractor().items()
2865 info_dict.update(extra)
2866 info_dict.pop('__post_extractor', None)
2867
2868 original_infodict = info_dict.get('__original_infodict') or {}
2869 original_infodict.update(extra)
2870 original_infodict.pop('__post_extractor', None)
2871
2872 actual_post_extract(info_dict or {})
2873
2874 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
2875 info = dict(ie_info)
2876 info['__files_to_move'] = files_to_move or {}
2877 for pp in self._pps[key]:
2878 info = self.run_pp(pp, info)
2879 return info, info.pop('__files_to_move', None)
2880
2881 def post_process(self, filename, ie_info, files_to_move=None):
2882 """Run all the postprocessors on the given file."""
2883 info = dict(ie_info)
2884 info['filepath'] = filename
2885 info['__files_to_move'] = files_to_move or {}
2886
2887 for pp in ie_info.get('__postprocessors', []) + self._pps['post_process']:
2888 info = self.run_pp(pp, info)
2889 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
2890 del info['__files_to_move']
2891 for pp in self._pps['after_move']:
2892 info = self.run_pp(pp, info)
2893 return info
2894
2895 def _make_archive_id(self, info_dict):
2896 video_id = info_dict.get('id')
2897 if not video_id:
2898 return
2899 # Future-proof against any change in case
2900 # and backwards compatibility with prior versions
2901 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
2902 if extractor is None:
2903 url = str_or_none(info_dict.get('url'))
2904 if not url:
2905 return
2906 # Try to find matching extractor for the URL and take its ie_key
2907 for ie in self._ies:
2908 if ie.suitable(url):
2909 extractor = ie.ie_key()
2910 break
2911 else:
2912 return
2913 return '%s %s' % (extractor.lower(), video_id)
2914
2915 def in_download_archive(self, info_dict):
2916 fn = self.params.get('download_archive')
2917 if fn is None:
2918 return False
2919
2920 vid_id = self._make_archive_id(info_dict)
2921 if not vid_id:
2922 return False # Incomplete video information
2923
2924 return vid_id in self.archive
2925
2926 def record_download_archive(self, info_dict):
2927 fn = self.params.get('download_archive')
2928 if fn is None:
2929 return
2930 vid_id = self._make_archive_id(info_dict)
2931 assert vid_id
2932 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2933 archive_file.write(vid_id + '\n')
2934 self.archive.add(vid_id)
2935
2936 @staticmethod
2937 def format_resolution(format, default='unknown'):
2938 if format.get('vcodec') == 'none':
2939 if format.get('acodec') == 'none':
2940 return 'images'
2941 return 'audio only'
2942 if format.get('resolution') is not None:
2943 return format['resolution']
2944 if format.get('width') and format.get('height'):
2945 res = '%dx%d' % (format['width'], format['height'])
2946 elif format.get('height'):
2947 res = '%sp' % format['height']
2948 elif format.get('width'):
2949 res = '%dx?' % format['width']
2950 else:
2951 res = default
2952 return res
2953
2954 def _format_note(self, fdict):
2955 res = ''
2956 if fdict.get('ext') in ['f4f', 'f4m']:
2957 res += '(unsupported) '
2958 if fdict.get('language'):
2959 if res:
2960 res += ' '
2961 res += '[%s] ' % fdict['language']
2962 if fdict.get('format_note') is not None:
2963 res += fdict['format_note'] + ' '
2964 if fdict.get('tbr') is not None:
2965 res += '%4dk ' % fdict['tbr']
2966 if fdict.get('container') is not None:
2967 if res:
2968 res += ', '
2969 res += '%s container' % fdict['container']
2970 if (fdict.get('vcodec') is not None
2971 and fdict.get('vcodec') != 'none'):
2972 if res:
2973 res += ', '
2974 res += fdict['vcodec']
2975 if fdict.get('vbr') is not None:
2976 res += '@'
2977 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2978 res += 'video@'
2979 if fdict.get('vbr') is not None:
2980 res += '%4dk' % fdict['vbr']
2981 if fdict.get('fps') is not None:
2982 if res:
2983 res += ', '
2984 res += '%sfps' % fdict['fps']
2985 if fdict.get('acodec') is not None:
2986 if res:
2987 res += ', '
2988 if fdict['acodec'] == 'none':
2989 res += 'video only'
2990 else:
2991 res += '%-5s' % fdict['acodec']
2992 elif fdict.get('abr') is not None:
2993 if res:
2994 res += ', '
2995 res += 'audio'
2996 if fdict.get('abr') is not None:
2997 res += '@%3dk' % fdict['abr']
2998 if fdict.get('asr') is not None:
2999 res += ' (%5dHz)' % fdict['asr']
3000 if fdict.get('filesize') is not None:
3001 if res:
3002 res += ', '
3003 res += format_bytes(fdict['filesize'])
3004 elif fdict.get('filesize_approx') is not None:
3005 if res:
3006 res += ', '
3007 res += '~' + format_bytes(fdict['filesize_approx'])
3008 return res
3009
3010 def _format_note_table(self, f):
3011 def join_fields(*vargs):
3012 return ', '.join((val for val in vargs if val != ''))
3013
3014 return join_fields(
3015 'UNSUPPORTED' if f.get('ext') in ('f4f', 'f4m') else '',
3016 format_field(f, 'language', '[%s]'),
3017 format_field(f, 'format_note'),
3018 format_field(f, 'container', ignore=(None, f.get('ext'))),
3019 format_field(f, 'asr', '%5dHz'))
3020
3021 def list_formats(self, info_dict):
3022 formats = info_dict.get('formats', [info_dict])
3023 new_format = (
3024 'list-formats' not in self.params.get('compat_opts', [])
3025 and self.params.get('list_formats_as_table', True) is not False)
3026 if new_format:
3027 table = [
3028 [
3029 format_field(f, 'format_id'),
3030 format_field(f, 'ext'),
3031 self.format_resolution(f),
3032 format_field(f, 'fps', '%d'),
3033 '|',
3034 format_field(f, 'filesize', ' %s', func=format_bytes) + format_field(f, 'filesize_approx', '~%s', func=format_bytes),
3035 format_field(f, 'tbr', '%4dk'),
3036 shorten_protocol_name(f.get('protocol', '').replace("native", "n")),
3037 '|',
3038 format_field(f, 'vcodec', default='unknown').replace('none', ''),
3039 format_field(f, 'vbr', '%4dk'),
3040 format_field(f, 'acodec', default='unknown').replace('none', ''),
3041 format_field(f, 'abr', '%3dk'),
3042 format_field(f, 'asr', '%5dHz'),
3043 self._format_note_table(f)]
3044 for f in formats
3045 if f.get('preference') is None or f['preference'] >= -1000]
3046 header_line = ['ID', 'EXT', 'RESOLUTION', 'FPS', '|', ' FILESIZE', ' TBR', 'PROTO',
3047 '|', 'VCODEC', ' VBR', 'ACODEC', ' ABR', ' ASR', 'NOTE']
3048 else:
3049 table = [
3050 [
3051 format_field(f, 'format_id'),
3052 format_field(f, 'ext'),
3053 self.format_resolution(f),
3054 self._format_note(f)]
3055 for f in formats
3056 if f.get('preference') is None or f['preference'] >= -1000]
3057 header_line = ['format code', 'extension', 'resolution', 'note']
3058
3059 self.to_screen(
3060 '[info] Available formats for %s:\n%s' % (info_dict['id'], render_table(
3061 header_line,
3062 table,
3063 delim=new_format,
3064 extraGap=(0 if new_format else 1),
3065 hideEmpty=new_format)))
3066
3067 def list_thumbnails(self, info_dict):
3068 thumbnails = list(info_dict.get('thumbnails'))
3069 if not thumbnails:
3070 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
3071 return
3072
3073 self.to_screen(
3074 '[info] Thumbnails for %s:' % info_dict['id'])
3075 self.to_screen(render_table(
3076 ['ID', 'width', 'height', 'URL'],
3077 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
3078
3079 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3080 if not subtitles:
3081 self.to_screen('%s has no %s' % (video_id, name))
3082 return
3083 self.to_screen(
3084 'Available %s for %s:' % (name, video_id))
3085
3086 def _row(lang, formats):
3087 exts, names = zip(*((f['ext'], f.get('name', 'unknown')) for f in reversed(formats)))
3088 if len(set(names)) == 1:
3089 names = [] if names[0] == 'unknown' else names[:1]
3090 return [lang, ', '.join(names), ', '.join(exts)]
3091
3092 self.to_screen(render_table(
3093 ['Language', 'Name', 'Formats'],
3094 [_row(lang, formats) for lang, formats in subtitles.items()],
3095 hideEmpty=True))
3096
3097 def urlopen(self, req):
3098 """ Start an HTTP download """
3099 if isinstance(req, compat_basestring):
3100 req = sanitized_Request(req)
3101 return self._opener.open(req, timeout=self._socket_timeout)
3102
3103 def print_debug_header(self):
3104 if not self.params.get('verbose'):
3105 return
3106
3107 if type('') is not compat_str:
3108 # Python 2.6 on SLES11 SP1 (https://github.com/ytdl-org/youtube-dl/issues/3326)
3109 self.report_warning(
3110 'Your Python is broken! Update to a newer and supported version')
3111
3112 stdout_encoding = getattr(
3113 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
3114 encoding_str = (
3115 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
3116 locale.getpreferredencoding(),
3117 sys.getfilesystemencoding(),
3118 stdout_encoding,
3119 self.get_encoding()))
3120 write_string(encoding_str, encoding=None)
3121
3122 source = (
3123 '(exe)' if hasattr(sys, 'frozen')
3124 else '(zip)' if isinstance(globals().get('__loader__'), zipimporter)
3125 else '(source)' if os.path.basename(sys.argv[0]) == '__main__.py'
3126 else '')
3127 self._write_string('[debug] yt-dlp version %s %s\n' % (__version__, source))
3128 if _LAZY_LOADER:
3129 self._write_string('[debug] Lazy loading extractors enabled\n')
3130 if _PLUGIN_CLASSES:
3131 self._write_string(
3132 '[debug] Plugin Extractors: %s\n' % [ie.ie_key() for ie in _PLUGIN_CLASSES])
3133 if self.params.get('compat_opts'):
3134 self._write_string(
3135 '[debug] Compatibility options: %s\n' % ', '.join(self.params.get('compat_opts')))
3136 try:
3137 sp = subprocess.Popen(
3138 ['git', 'rev-parse', '--short', 'HEAD'],
3139 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3140 cwd=os.path.dirname(os.path.abspath(__file__)))
3141 out, err = process_communicate_or_kill(sp)
3142 out = out.decode().strip()
3143 if re.match('[0-9a-f]+', out):
3144 self._write_string('[debug] Git HEAD: %s\n' % out)
3145 except Exception:
3146 try:
3147 sys.exc_clear()
3148 except Exception:
3149 pass
3150
3151 def python_implementation():
3152 impl_name = platform.python_implementation()
3153 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3154 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3155 return impl_name
3156
3157 self._write_string('[debug] Python version %s (%s %s) - %s\n' % (
3158 platform.python_version(),
3159 python_implementation(),
3160 platform.architecture()[0],
3161 platform_name()))
3162
3163 exe_versions = FFmpegPostProcessor.get_versions(self)
3164 exe_versions['rtmpdump'] = rtmpdump_version()
3165 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3166 exe_str = ', '.join(
3167 '%s %s' % (exe, v)
3168 for exe, v in sorted(exe_versions.items())
3169 if v
3170 )
3171 if not exe_str:
3172 exe_str = 'none'
3173 self._write_string('[debug] exe versions: %s\n' % exe_str)
3174
3175 proxy_map = {}
3176 for handler in self._opener.handlers:
3177 if hasattr(handler, 'proxies'):
3178 proxy_map.update(handler.proxies)
3179 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
3180
3181 if self.params.get('call_home', False):
3182 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3183 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
3184 return
3185 latest_version = self.urlopen(
3186 'https://yt-dl.org/latest/version').read().decode('utf-8')
3187 if version_tuple(latest_version) > version_tuple(__version__):
3188 self.report_warning(
3189 'You are using an outdated version (newest version: %s)! '
3190 'See https://yt-dl.org/update if you need help updating.' %
3191 latest_version)
3192
3193 def _setup_opener(self):
3194 timeout_val = self.params.get('socket_timeout')
3195 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
3196
3197 opts_cookiefile = self.params.get('cookiefile')
3198 opts_proxy = self.params.get('proxy')
3199
3200 if opts_cookiefile is None:
3201 self.cookiejar = compat_cookiejar.CookieJar()
3202 else:
3203 opts_cookiefile = expand_path(opts_cookiefile)
3204 self.cookiejar = YoutubeDLCookieJar(opts_cookiefile)
3205 if os.access(opts_cookiefile, os.R_OK):
3206 self.cookiejar.load(ignore_discard=True, ignore_expires=True)
3207
3208 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3209 if opts_proxy is not None:
3210 if opts_proxy == '':
3211 proxies = {}
3212 else:
3213 proxies = {'http': opts_proxy, 'https': opts_proxy}
3214 else:
3215 proxies = compat_urllib_request.getproxies()
3216 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3217 if 'http' in proxies and 'https' not in proxies:
3218 proxies['https'] = proxies['http']
3219 proxy_handler = PerRequestProxyHandler(proxies)
3220
3221 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3222 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3223 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3224 redirect_handler = YoutubeDLRedirectHandler()
3225 data_handler = compat_urllib_request_DataHandler()
3226
3227 # When passing our own FileHandler instance, build_opener won't add the
3228 # default FileHandler and allows us to disable the file protocol, which
3229 # can be used for malicious purposes (see
3230 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3231 file_handler = compat_urllib_request.FileHandler()
3232
3233 def file_open(*args, **kwargs):
3234 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3235 file_handler.file_open = file_open
3236
3237 opener = compat_urllib_request.build_opener(
3238 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3239
3240 # Delete the default user-agent header, which would otherwise apply in
3241 # cases where our custom HTTP handler doesn't come into play
3242 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3243 opener.addheaders = []
3244 self._opener = opener
3245
3246 def encode(self, s):
3247 if isinstance(s, bytes):
3248 return s # Already encoded
3249
3250 try:
3251 return s.encode(self.get_encoding())
3252 except UnicodeEncodeError as err:
3253 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3254 raise
3255
3256 def get_encoding(self):
3257 encoding = self.params.get('encoding')
3258 if encoding is None:
3259 encoding = preferredencoding()
3260 return encoding
3261
3262 def _write_thumbnails(self, info_dict, filename): # return the extensions
3263 write_all = self.params.get('write_all_thumbnails', False)
3264 thumbnails = []
3265 if write_all or self.params.get('writethumbnail', False):
3266 thumbnails = info_dict.get('thumbnails') or []
3267 multiple = write_all and len(thumbnails) > 1
3268
3269 ret = []
3270 for t in thumbnails[::1 if write_all else -1]:
3271 thumb_ext = determine_ext(t['url'], 'jpg')
3272 suffix = '%s.' % t['id'] if multiple else ''
3273 thumb_display_id = '%s ' % t['id'] if multiple else ''
3274 thumb_filename = replace_extension(filename, suffix + thumb_ext, info_dict.get('ext'))
3275
3276 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(thumb_filename)):
3277 ret.append(suffix + thumb_ext)
3278 t['filepath'] = thumb_filename
3279 self.to_screen('[%s] %s: Thumbnail %sis already present' %
3280 (info_dict['extractor'], info_dict['id'], thumb_display_id))
3281 else:
3282 self.to_screen('[%s] %s: Downloading thumbnail %s ...' %
3283 (info_dict['extractor'], info_dict['id'], thumb_display_id))
3284 try:
3285 uf = self.urlopen(t['url'])
3286 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3287 shutil.copyfileobj(uf, thumbf)
3288 ret.append(suffix + thumb_ext)
3289 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
3290 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
3291 t['filepath'] = thumb_filename
3292 except network_exceptions as err:
3293 self.report_warning('Unable to download thumbnail "%s": %s' %
3294 (t['url'], error_to_compat_str(err)))
3295 if ret and not write_all:
3296 break
3297 return ret