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