]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[youtube] Add `thirdParty` to agegate clients (#577)
[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_get_terminal_size,
35 compat_kwargs,
36 compat_numeric_types,
37 compat_os_name,
38 compat_shlex_quote,
39 compat_str,
40 compat_tokenize_tokenize,
41 compat_urllib_error,
42 compat_urllib_request,
43 compat_urllib_request_DataHandler,
44 )
45 from .cookies import load_cookies
46 from .utils import (
47 age_restricted,
48 args_to_str,
49 ContentTooShortError,
50 date_from_str,
51 DateRange,
52 DEFAULT_OUTTMPL,
53 determine_ext,
54 determine_protocol,
55 DOT_DESKTOP_LINK_TEMPLATE,
56 DOT_URL_LINK_TEMPLATE,
57 DOT_WEBLOC_LINK_TEMPLATE,
58 DownloadError,
59 encode_compat_str,
60 encodeFilename,
61 EntryNotInPlaylist,
62 error_to_compat_str,
63 ExistingVideoReached,
64 expand_path,
65 ExtractorError,
66 float_or_none,
67 format_bytes,
68 format_field,
69 STR_FORMAT_RE_TMPL,
70 STR_FORMAT_TYPES,
71 formatSeconds,
72 GeoRestrictedError,
73 HEADRequest,
74 int_or_none,
75 iri_to_uri,
76 ISO3166Utils,
77 LazyList,
78 locked_file,
79 make_dir,
80 make_HTTPS_handler,
81 MaxDownloadsReached,
82 network_exceptions,
83 orderedSet,
84 OUTTMPL_TYPES,
85 PagedList,
86 parse_filesize,
87 PerRequestProxyHandler,
88 platform_name,
89 PostProcessingError,
90 preferredencoding,
91 prepend_extension,
92 process_communicate_or_kill,
93 register_socks_protocols,
94 RejectedVideoReached,
95 render_table,
96 replace_extension,
97 SameFileError,
98 sanitize_filename,
99 sanitize_path,
100 sanitize_url,
101 sanitized_Request,
102 std_headers,
103 str_or_none,
104 strftime_or_none,
105 subtitles_filename,
106 ThrottledDownload,
107 to_high_limit_path,
108 traverse_obj,
109 try_get,
110 UnavailableVideoError,
111 url_basename,
112 variadic,
113 version_tuple,
114 write_json_file,
115 write_string,
116 YoutubeDLCookieProcessor,
117 YoutubeDLHandler,
118 YoutubeDLRedirectHandler,
119 )
120 from .cache import Cache
121 from .extractor import (
122 gen_extractor_classes,
123 get_info_extractor,
124 _LAZY_LOADER,
125 _PLUGIN_CLASSES
126 )
127 from .extractor.openload import PhantomJSwrapper
128 from .downloader import (
129 get_suitable_downloader,
130 shorten_protocol_name
131 )
132 from .downloader.rtmp import rtmpdump_version
133 from .postprocessor import (
134 get_postprocessor,
135 FFmpegFixupDurationPP,
136 FFmpegFixupM3u8PP,
137 FFmpegFixupM4aPP,
138 FFmpegFixupStretchedPP,
139 FFmpegFixupTimestampPP,
140 FFmpegMergerPP,
141 FFmpegPostProcessor,
142 MoveFilesAfterDownloadPP,
143 )
144 from .version import __version__
145
146 if compat_os_name == 'nt':
147 import ctypes
148
149
150 class YoutubeDL(object):
151 """YoutubeDL class.
152
153 YoutubeDL objects are the ones responsible of downloading the
154 actual video file and writing it to disk if the user has requested
155 it, among some other tasks. In most cases there should be one per
156 program. As, given a video URL, the downloader doesn't know how to
157 extract all the needed information, task that InfoExtractors do, it
158 has to pass the URL to one of them.
159
160 For this, YoutubeDL objects have a method that allows
161 InfoExtractors to be registered in a given order. When it is passed
162 a URL, the YoutubeDL object handles it to the first InfoExtractor it
163 finds that reports being able to handle it. The InfoExtractor extracts
164 all the information about the video or videos the URL refers to, and
165 YoutubeDL process the extracted information, possibly using a File
166 Downloader to download the video.
167
168 YoutubeDL objects accept a lot of parameters. In order not to saturate
169 the object constructor with arguments, it receives a dictionary of
170 options instead. These options are available through the params
171 attribute for the InfoExtractors to use. The YoutubeDL also
172 registers itself as the downloader in charge for the InfoExtractors
173 that are added to it, so this is a "mutual registration".
174
175 Available options:
176
177 username: Username for authentication purposes.
178 password: Password for authentication purposes.
179 videopassword: Password for accessing a video.
180 ap_mso: Adobe Pass multiple-system operator identifier.
181 ap_username: Multiple-system operator account username.
182 ap_password: Multiple-system operator account password.
183 usenetrc: Use netrc for authentication instead.
184 verbose: Print additional info to stdout.
185 quiet: Do not print messages to stdout.
186 no_warnings: Do not print out anything for warnings.
187 forceprint: A list of templates to force print
188 forceurl: Force printing final URL. (Deprecated)
189 forcetitle: Force printing title. (Deprecated)
190 forceid: Force printing ID. (Deprecated)
191 forcethumbnail: Force printing thumbnail URL. (Deprecated)
192 forcedescription: Force printing description. (Deprecated)
193 forcefilename: Force printing final filename. (Deprecated)
194 forceduration: Force printing duration. (Deprecated)
195 forcejson: Force printing info_dict as JSON.
196 dump_single_json: Force printing the info_dict of the whole playlist
197 (or video) as a single JSON line.
198 force_write_download_archive: Force writing download archive regardless
199 of 'skip_download' or 'simulate'.
200 simulate: Do not download the video files.
201 format: Video format code. see "FORMAT SELECTION" for more details.
202 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
203 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
204 extracting metadata even if the video is not actually
205 available for download (experimental)
206 format_sort: How to sort the video formats. see "Sorting Formats"
207 for more details.
208 format_sort_force: Force the given format_sort. see "Sorting Formats"
209 for more details.
210 allow_multiple_video_streams: Allow multiple video streams to be merged
211 into a single file
212 allow_multiple_audio_streams: Allow multiple audio streams to be merged
213 into a single file
214 check_formats Whether to test if the formats are downloadable.
215 Can be True (check all), False (check none)
216 or None (check only if requested by extractor)
217 paths: Dictionary of output paths. The allowed keys are 'home'
218 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
219 outtmpl: Dictionary of templates for output names. Allowed keys
220 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
221 A string a also accepted for backward compatibility
222 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
223 restrictfilenames: Do not allow "&" and spaces in file names
224 trim_file_name: Limit length of filename (extension excluded)
225 windowsfilenames: Force the filenames to be windows compatible
226 ignoreerrors: Do not stop on download errors
227 (Default True when running yt-dlp,
228 but False when directly accessing YoutubeDL class)
229 skip_playlist_after_errors: Number of allowed failures until the rest of
230 the playlist is skipped
231 force_generic_extractor: Force downloader to use the generic extractor
232 overwrites: Overwrite all video and metadata files if True,
233 overwrite only non-video files if None
234 and don't overwrite any file if False
235 playliststart: Playlist item to start at.
236 playlistend: Playlist item to end at.
237 playlist_items: Specific indices of playlist to download.
238 playlistreverse: Download playlist items in reverse order.
239 playlistrandom: Download playlist items in random order.
240 matchtitle: Download only matching titles.
241 rejecttitle: Reject downloads for matching titles.
242 logger: Log messages to a logging.Logger instance.
243 logtostderr: Log messages to stderr instead of stdout.
244 writedescription: Write the video description to a .description file
245 writeinfojson: Write the video description to a .info.json file
246 clean_infojson: Remove private fields from the infojson
247 writecomments: Extract video comments. This will not be written to disk
248 unless writeinfojson is also given
249 writeannotations: Write the video annotations to a .annotations.xml file
250 writethumbnail: Write the thumbnail image to a file
251 allow_playlist_files: Whether to write playlists' description, infojson etc
252 also to disk when using the 'write*' options
253 write_all_thumbnails: Write all thumbnail formats to files
254 writelink: Write an internet shortcut file, depending on the
255 current platform (.url/.webloc/.desktop)
256 writeurllink: Write a Windows internet shortcut file (.url)
257 writewebloclink: Write a macOS internet shortcut file (.webloc)
258 writedesktoplink: Write a Linux internet shortcut file (.desktop)
259 writesubtitles: Write the video subtitles to a file
260 writeautomaticsub: Write the automatically generated subtitles to a file
261 allsubtitles: Deprecated - Use subtitleslangs = ['all']
262 Downloads all the subtitles of the video
263 (requires writesubtitles or writeautomaticsub)
264 listsubtitles: Lists all available subtitles for the video
265 subtitlesformat: The format code for subtitles
266 subtitleslangs: List of languages of the subtitles to download (can be regex).
267 The list may contain "all" to refer to all the available
268 subtitles. The language can be prefixed with a "-" to
269 exclude it from the requested languages. Eg: ['all', '-live_chat']
270 keepvideo: Keep the video file after post-processing
271 daterange: A DateRange object, download only if the upload_date is in the range.
272 skip_download: Skip the actual download of the video file
273 cachedir: Location of the cache files in the filesystem.
274 False to disable filesystem cache.
275 noplaylist: Download single video instead of a playlist if in doubt.
276 age_limit: An integer representing the user's age in years.
277 Unsuitable videos for the given age are skipped.
278 min_views: An integer representing the minimum view count the video
279 must have in order to not be skipped.
280 Videos without view count information are always
281 downloaded. None for no limit.
282 max_views: An integer representing the maximum view count.
283 Videos that are more popular than that are not
284 downloaded.
285 Videos without view count information are always
286 downloaded. None for no limit.
287 download_archive: File name of a file where all downloads are recorded.
288 Videos already present in the file are not downloaded
289 again.
290 break_on_existing: Stop the download process after attempting to download a
291 file that is in the archive.
292 break_on_reject: Stop the download process when encountering a video that
293 has been filtered out.
294 cookiefile: File name where cookies should be read from and dumped to
295 cookiesfrombrowser: A tuple containing the name of the browser and the profile
296 name/path from where cookies are loaded.
297 Eg: ('chrome', ) or (vivaldi, 'default')
298 nocheckcertificate:Do not verify SSL certificates
299 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
300 At the moment, this is only supported by YouTube.
301 proxy: URL of the proxy server to use
302 geo_verification_proxy: URL of the proxy to use for IP address verification
303 on geo-restricted sites.
304 socket_timeout: Time to wait for unresponsive hosts, in seconds
305 bidi_workaround: Work around buggy terminals without bidirectional text
306 support, using fridibi
307 debug_printtraffic:Print out sent and received HTTP traffic
308 include_ads: Download ads as well
309 default_search: Prepend this string if an input url is not valid.
310 'auto' for elaborate guessing
311 encoding: Use this encoding instead of the system-specified.
312 extract_flat: Do not resolve URLs, return the immediate result.
313 Pass in 'in_playlist' to only show this behavior for
314 playlist items.
315 postprocessors: A list of dictionaries, each with an entry
316 * key: The name of the postprocessor. See
317 yt_dlp/postprocessor/__init__.py for a list.
318 * when: When to run the postprocessor. Can be one of
319 pre_process|before_dl|post_process|after_move.
320 Assumed to be 'post_process' if not given
321 post_hooks: A list of functions that get called as the final step
322 for each video file, after all postprocessors have been
323 called. The filename will be passed as the only argument.
324 progress_hooks: A list of functions that get called on download
325 progress, with a dictionary with the entries
326 * status: One of "downloading", "error", or "finished".
327 Check this first and ignore unknown values.
328 * info_dict: The extracted info_dict
329
330 If status is one of "downloading", or "finished", the
331 following properties may also be present:
332 * filename: The final filename (always present)
333 * tmpfilename: The filename we're currently writing to
334 * downloaded_bytes: Bytes on disk
335 * total_bytes: Size of the whole file, None if unknown
336 * total_bytes_estimate: Guess of the eventual file size,
337 None if unavailable.
338 * elapsed: The number of seconds since download started.
339 * eta: The estimated time in seconds, None if unknown
340 * speed: The download speed in bytes/second, None if
341 unknown
342 * fragment_index: The counter of the currently
343 downloaded video fragment.
344 * fragment_count: The number of fragments (= individual
345 files that will be merged)
346
347 Progress hooks are guaranteed to be called at least once
348 (with status "finished") if the download is successful.
349 merge_output_format: Extension to use when merging formats.
350 final_ext: Expected final extension; used to detect when the file was
351 already downloaded and converted. "merge_output_format" is
352 replaced by this extension when given
353 fixup: Automatically correct known faults of the file.
354 One of:
355 - "never": do nothing
356 - "warn": only emit a warning
357 - "detect_or_warn": check whether we can do anything
358 about it, warn otherwise (default)
359 source_address: Client-side IP address to bind to.
360 call_home: Boolean, true iff we are allowed to contact the
361 yt-dlp servers for debugging. (BROKEN)
362 sleep_interval_requests: Number of seconds to sleep between requests
363 during extraction
364 sleep_interval: Number of seconds to sleep before each download when
365 used alone or a lower bound of a range for randomized
366 sleep before each download (minimum possible number
367 of seconds to sleep) when used along with
368 max_sleep_interval.
369 max_sleep_interval:Upper bound of a range for randomized sleep before each
370 download (maximum possible number of seconds to sleep).
371 Must only be used along with sleep_interval.
372 Actual sleep time will be a random float from range
373 [sleep_interval; max_sleep_interval].
374 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
375 listformats: Print an overview of available video formats and exit.
376 list_thumbnails: Print a table of all thumbnails and exit.
377 match_filter: A function that gets called with the info_dict of
378 every video.
379 If it returns a message, the video is ignored.
380 If it returns None, the video is downloaded.
381 match_filter_func in utils.py is one example for this.
382 no_color: Do not emit color codes in output.
383 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
384 HTTP header
385 geo_bypass_country:
386 Two-letter ISO 3166-2 country code that will be used for
387 explicit geographic restriction bypassing via faking
388 X-Forwarded-For HTTP header
389 geo_bypass_ip_block:
390 IP range in CIDR notation that will be used similarly to
391 geo_bypass_country
392
393 The following options determine which downloader is picked:
394 external_downloader: A dictionary of protocol keys and the executable of the
395 external downloader to use for it. The allowed protocols
396 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
397 Set the value to 'native' to use the native downloader
398 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
399 or {'m3u8': 'ffmpeg'} instead.
400 Use the native HLS downloader instead of ffmpeg/avconv
401 if True, otherwise use ffmpeg/avconv if False, otherwise
402 use downloader suggested by extractor if None.
403 compat_opts: Compatibility options. See "Differences in default behavior".
404 The following options do not work when used through the API:
405 filename, abort-on-error, multistreams, no-live-chat,
406 no-playlist-metafiles. Refer __init__.py for their implementation
407
408 The following parameters are not used by YoutubeDL itself, they are used by
409 the downloader (see yt_dlp/downloader/common.py):
410 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
411 max_filesize, test, noresizebuffer, retries, continuedl, noprogress, consoletitle,
412 xattr_set_filesize, external_downloader_args, hls_use_mpegts, http_chunk_size.
413
414 The following options are used by the post processors:
415 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
416 otherwise prefer ffmpeg. (avconv support is deprecated)
417 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
418 to the binary or its containing directory.
419 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
420 and a list of additional command-line arguments for the
421 postprocessor/executable. The dict can also have "PP+EXE" keys
422 which are used when the given exe is used by the given PP.
423 Use 'default' as the name for arguments to passed to all PP
424
425 The following options are used by the extractors:
426 extractor_retries: Number of times to retry for known errors
427 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
428 hls_split_discontinuity: Split HLS playlists to different formats at
429 discontinuities such as ad breaks (default: False)
430 extractor_args: A dictionary of arguments to be passed to the extractors.
431 See "EXTRACTOR ARGUMENTS" for details.
432 Eg: {'youtube': {'skip': ['dash', 'hls']}}
433 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
434 If True (default), DASH manifests and related
435 data will be downloaded and processed by extractor.
436 You can reduce network I/O by disabling it if you don't
437 care about DASH. (only for youtube)
438 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
439 If True (default), HLS manifests and related
440 data will be downloaded and processed by extractor.
441 You can reduce network I/O by disabling it if you don't
442 care about HLS. (only for youtube)
443 """
444
445 _NUMERIC_FIELDS = set((
446 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
447 'timestamp', 'upload_year', 'upload_month', 'upload_day',
448 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
449 'average_rating', 'comment_count', 'age_limit',
450 'start_time', 'end_time',
451 'chapter_number', 'season_number', 'episode_number',
452 'track_number', 'disc_number', 'release_year',
453 'playlist_index',
454 ))
455
456 params = None
457 _ies = []
458 _pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
459 _printed_messages = set()
460 _first_webpage_request = True
461 _download_retcode = None
462 _num_downloads = None
463 _playlist_level = 0
464 _playlist_urls = set()
465 _screen_file = None
466
467 def __init__(self, params=None, auto_init=True):
468 """Create a FileDownloader object with the given options."""
469 if params is None:
470 params = {}
471 self._ies = []
472 self._ies_instances = {}
473 self._pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
474 self._printed_messages = set()
475 self._first_webpage_request = True
476 self._post_hooks = []
477 self._progress_hooks = []
478 self._download_retcode = 0
479 self._num_downloads = 0
480 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
481 self._err_file = sys.stderr
482 self.params = {
483 # Default parameters
484 'nocheckcertificate': False,
485 }
486 self.params.update(params)
487 self.cache = Cache(self)
488
489 if sys.version_info < (3, 6):
490 self.report_warning(
491 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
492
493 def check_deprecated(param, option, suggestion):
494 if self.params.get(param) is not None:
495 self.report_warning('%s is deprecated. Use %s instead' % (option, suggestion))
496 return True
497 return False
498
499 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
500 if self.params.get('geo_verification_proxy') is None:
501 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
502
503 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
504 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
505 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
506
507 for msg in self.params.get('warnings', []):
508 self.report_warning(msg)
509
510 if self.params.get('final_ext'):
511 if self.params.get('merge_output_format'):
512 self.report_warning('--merge-output-format will be ignored since --remux-video or --recode-video is given')
513 self.params['merge_output_format'] = self.params['final_ext']
514
515 if 'overwrites' in self.params and self.params['overwrites'] is None:
516 del self.params['overwrites']
517
518 if params.get('bidi_workaround', False):
519 try:
520 import pty
521 master, slave = pty.openpty()
522 width = compat_get_terminal_size().columns
523 if width is None:
524 width_args = []
525 else:
526 width_args = ['-w', str(width)]
527 sp_kwargs = dict(
528 stdin=subprocess.PIPE,
529 stdout=slave,
530 stderr=self._err_file)
531 try:
532 self._output_process = subprocess.Popen(
533 ['bidiv'] + width_args, **sp_kwargs
534 )
535 except OSError:
536 self._output_process = subprocess.Popen(
537 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
538 self._output_channel = os.fdopen(master, 'rb')
539 except OSError as ose:
540 if ose.errno == errno.ENOENT:
541 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.')
542 else:
543 raise
544
545 if (sys.platform != 'win32'
546 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
547 and not params.get('restrictfilenames', False)):
548 # Unicode filesystem API will throw errors (#1474, #13027)
549 self.report_warning(
550 'Assuming --restrict-filenames since file system encoding '
551 'cannot encode all characters. '
552 'Set the LC_ALL environment variable to fix this.')
553 self.params['restrictfilenames'] = True
554
555 self.outtmpl_dict = self.parse_outtmpl()
556
557 # Creating format selector here allows us to catch syntax errors before the extraction
558 self.format_selector = (
559 None if self.params.get('format') is None
560 else self.build_format_selector(self.params['format']))
561
562 self._setup_opener()
563
564 """Preload the archive, if any is specified"""
565 def preload_download_archive(fn):
566 if fn is None:
567 return False
568 self.write_debug('Loading archive file %r\n' % fn)
569 try:
570 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
571 for line in archive_file:
572 self.archive.add(line.strip())
573 except IOError as ioe:
574 if ioe.errno != errno.ENOENT:
575 raise
576 return False
577 return True
578
579 self.archive = set()
580 preload_download_archive(self.params.get('download_archive'))
581
582 if auto_init:
583 self.print_debug_header()
584 self.add_default_info_extractors()
585
586 for pp_def_raw in self.params.get('postprocessors', []):
587 pp_def = dict(pp_def_raw)
588 when = pp_def.pop('when', 'post_process')
589 pp_class = get_postprocessor(pp_def.pop('key'))
590 pp = pp_class(self, **compat_kwargs(pp_def))
591 self.add_post_processor(pp, when=when)
592
593 for ph in self.params.get('post_hooks', []):
594 self.add_post_hook(ph)
595
596 for ph in self.params.get('progress_hooks', []):
597 self.add_progress_hook(ph)
598
599 register_socks_protocols()
600
601 def warn_if_short_id(self, argv):
602 # short YouTube ID starting with dash?
603 idxs = [
604 i for i, a in enumerate(argv)
605 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
606 if idxs:
607 correct_argv = (
608 ['yt-dlp']
609 + [a for i, a in enumerate(argv) if i not in idxs]
610 + ['--'] + [argv[i] for i in idxs]
611 )
612 self.report_warning(
613 'Long argument string detected. '
614 'Use -- to separate parameters and URLs, like this:\n%s\n' %
615 args_to_str(correct_argv))
616
617 def add_info_extractor(self, ie):
618 """Add an InfoExtractor object to the end of the list."""
619 self._ies.append(ie)
620 if not isinstance(ie, type):
621 self._ies_instances[ie.ie_key()] = ie
622 ie.set_downloader(self)
623
624 def get_info_extractor(self, ie_key):
625 """
626 Get an instance of an IE with name ie_key, it will try to get one from
627 the _ies list, if there's no instance it will create a new one and add
628 it to the extractor list.
629 """
630 ie = self._ies_instances.get(ie_key)
631 if ie is None:
632 ie = get_info_extractor(ie_key)()
633 self.add_info_extractor(ie)
634 return ie
635
636 def add_default_info_extractors(self):
637 """
638 Add the InfoExtractors returned by gen_extractors to the end of the list
639 """
640 for ie in gen_extractor_classes():
641 self.add_info_extractor(ie)
642
643 def add_post_processor(self, pp, when='post_process'):
644 """Add a PostProcessor object to the end of the chain."""
645 self._pps[when].append(pp)
646 pp.set_downloader(self)
647
648 def add_post_hook(self, ph):
649 """Add the post hook"""
650 self._post_hooks.append(ph)
651
652 def add_progress_hook(self, ph):
653 """Add the progress hook (currently only for the file downloader)"""
654 self._progress_hooks.append(ph)
655
656 def _bidi_workaround(self, message):
657 if not hasattr(self, '_output_channel'):
658 return message
659
660 assert hasattr(self, '_output_process')
661 assert isinstance(message, compat_str)
662 line_count = message.count('\n') + 1
663 self._output_process.stdin.write((message + '\n').encode('utf-8'))
664 self._output_process.stdin.flush()
665 res = ''.join(self._output_channel.readline().decode('utf-8')
666 for _ in range(line_count))
667 return res[:-len('\n')]
668
669 def _write_string(self, message, out=None, only_once=False):
670 if only_once:
671 if message in self._printed_messages:
672 return
673 self._printed_messages.add(message)
674 write_string(message, out=out, encoding=self.params.get('encoding'))
675
676 def to_stdout(self, message, skip_eol=False, quiet=False):
677 """Print message to stdout"""
678 if self.params.get('logger'):
679 self.params['logger'].debug(message)
680 elif not quiet or self.params.get('verbose'):
681 self._write_string(
682 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
683 self._err_file if quiet else self._screen_file)
684
685 def to_stderr(self, message, only_once=False):
686 """Print message to stderr"""
687 assert isinstance(message, compat_str)
688 if self.params.get('logger'):
689 self.params['logger'].error(message)
690 else:
691 self._write_string('%s\n' % self._bidi_workaround(message), self._err_file, only_once=only_once)
692
693 def to_console_title(self, message):
694 if not self.params.get('consoletitle', False):
695 return
696 if compat_os_name == 'nt':
697 if ctypes.windll.kernel32.GetConsoleWindow():
698 # c_wchar_p() might not be necessary if `message` is
699 # already of type unicode()
700 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
701 elif 'TERM' in os.environ:
702 self._write_string('\033]0;%s\007' % message, self._screen_file)
703
704 def save_console_title(self):
705 if not self.params.get('consoletitle', False):
706 return
707 if self.params.get('simulate', False):
708 return
709 if compat_os_name != 'nt' and 'TERM' in os.environ:
710 # Save the title on stack
711 self._write_string('\033[22;0t', self._screen_file)
712
713 def restore_console_title(self):
714 if not self.params.get('consoletitle', False):
715 return
716 if self.params.get('simulate', False):
717 return
718 if compat_os_name != 'nt' and 'TERM' in os.environ:
719 # Restore the title from stack
720 self._write_string('\033[23;0t', self._screen_file)
721
722 def __enter__(self):
723 self.save_console_title()
724 return self
725
726 def __exit__(self, *args):
727 self.restore_console_title()
728
729 if self.params.get('cookiefile') is not None:
730 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
731
732 def trouble(self, message=None, tb=None):
733 """Determine action to take when a download problem appears.
734
735 Depending on if the downloader has been configured to ignore
736 download errors or not, this method may throw an exception or
737 not when errors are found, after printing the message.
738
739 tb, if given, is additional traceback information.
740 """
741 if message is not None:
742 self.to_stderr(message)
743 if self.params.get('verbose'):
744 if tb is None:
745 if sys.exc_info()[0]: # if .trouble has been called from an except block
746 tb = ''
747 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
748 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
749 tb += encode_compat_str(traceback.format_exc())
750 else:
751 tb_data = traceback.format_list(traceback.extract_stack())
752 tb = ''.join(tb_data)
753 if tb:
754 self.to_stderr(tb)
755 if not self.params.get('ignoreerrors', False):
756 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
757 exc_info = sys.exc_info()[1].exc_info
758 else:
759 exc_info = sys.exc_info()
760 raise DownloadError(message, exc_info)
761 self._download_retcode = 1
762
763 def to_screen(self, message, skip_eol=False):
764 """Print message to stdout if not in quiet mode"""
765 self.to_stdout(
766 message, skip_eol, quiet=self.params.get('quiet', False))
767
768 def report_warning(self, message, only_once=False):
769 '''
770 Print the message to stderr, it will be prefixed with 'WARNING:'
771 If stderr is a tty file the 'WARNING:' will be colored
772 '''
773 if self.params.get('logger') is not None:
774 self.params['logger'].warning(message)
775 else:
776 if self.params.get('no_warnings'):
777 return
778 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
779 _msg_header = '\033[0;33mWARNING:\033[0m'
780 else:
781 _msg_header = 'WARNING:'
782 warning_message = '%s %s' % (_msg_header, message)
783 self.to_stderr(warning_message, only_once)
784
785 def report_error(self, message, tb=None):
786 '''
787 Do the same as trouble, but prefixes the message with 'ERROR:', colored
788 in red if stderr is a tty file.
789 '''
790 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
791 _msg_header = '\033[0;31mERROR:\033[0m'
792 else:
793 _msg_header = 'ERROR:'
794 error_message = '%s %s' % (_msg_header, message)
795 self.trouble(error_message, tb)
796
797 def write_debug(self, message, only_once=False):
798 '''Log debug message or Print message to stderr'''
799 if not self.params.get('verbose', False):
800 return
801 message = '[debug] %s' % message
802 if self.params.get('logger'):
803 self.params['logger'].debug(message)
804 else:
805 self.to_stderr(message, only_once)
806
807 def report_file_already_downloaded(self, file_name):
808 """Report file has already been fully downloaded."""
809 try:
810 self.to_screen('[download] %s has already been downloaded' % file_name)
811 except UnicodeEncodeError:
812 self.to_screen('[download] The file has already been downloaded')
813
814 def report_file_delete(self, file_name):
815 """Report that existing file will be deleted."""
816 try:
817 self.to_screen('Deleting existing file %s' % file_name)
818 except UnicodeEncodeError:
819 self.to_screen('Deleting existing file')
820
821 def parse_outtmpl(self):
822 outtmpl_dict = self.params.get('outtmpl', {})
823 if not isinstance(outtmpl_dict, dict):
824 outtmpl_dict = {'default': outtmpl_dict}
825 outtmpl_dict.update({
826 k: v for k, v in DEFAULT_OUTTMPL.items()
827 if not outtmpl_dict.get(k)})
828 for key, val in outtmpl_dict.items():
829 if isinstance(val, bytes):
830 self.report_warning(
831 'Parameter outtmpl is bytes, but should be a unicode string. '
832 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
833 return outtmpl_dict
834
835 def get_output_path(self, dir_type='', filename=None):
836 paths = self.params.get('paths', {})
837 assert isinstance(paths, dict)
838 path = os.path.join(
839 expand_path(paths.get('home', '').strip()),
840 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
841 filename or '')
842
843 # Temporary fix for #4787
844 # 'Treat' all problem characters by passing filename through preferredencoding
845 # to workaround encoding issues with subprocess on python2 @ Windows
846 if sys.version_info < (3, 0) and sys.platform == 'win32':
847 path = encodeFilename(path, True).decode(preferredencoding())
848 return sanitize_path(path, force=self.params.get('windowsfilenames'))
849
850 @staticmethod
851 def _outtmpl_expandpath(outtmpl):
852 # expand_path translates '%%' into '%' and '$$' into '$'
853 # correspondingly that is not what we want since we need to keep
854 # '%%' intact for template dict substitution step. Working around
855 # with boundary-alike separator hack.
856 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
857 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
858
859 # outtmpl should be expand_path'ed before template dict substitution
860 # because meta fields may contain env variables we don't want to
861 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
862 # title "Hello $PATH", we don't want `$PATH` to be expanded.
863 return expand_path(outtmpl).replace(sep, '')
864
865 @staticmethod
866 def escape_outtmpl(outtmpl):
867 ''' Escape any remaining strings like %s, %abc% etc. '''
868 return re.sub(
869 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
870 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
871 outtmpl)
872
873 @classmethod
874 def validate_outtmpl(cls, outtmpl):
875 ''' @return None or Exception object '''
876 outtmpl = re.sub(
877 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljq]'),
878 lambda mobj: f'{mobj.group(0)[:-1]}s',
879 cls._outtmpl_expandpath(outtmpl))
880 try:
881 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
882 return None
883 except ValueError as err:
884 return err
885
886 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=None):
887 """ Make the template and info_dict suitable for substitution : ydl.outtmpl_escape(outtmpl) % info_dict """
888 info_dict = dict(info_dict)
889 na = self.params.get('outtmpl_na_placeholder', 'NA')
890
891 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
892 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
893 if info_dict.get('duration', None) is not None
894 else None)
895 info_dict['epoch'] = int(time.time())
896 info_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
897 if info_dict.get('resolution') is None:
898 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
899
900 # For fields playlist_index and autonumber convert all occurrences
901 # of %(field)s to %(field)0Nd for backward compatibility
902 field_size_compat_map = {
903 'playlist_index': len(str(info_dict.get('_last_playlist_index') or '')),
904 'autonumber': self.params.get('autonumber_size') or 5,
905 }
906
907 TMPL_DICT = {}
908 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljq]'))
909 MATH_FUNCTIONS = {
910 '+': float.__add__,
911 '-': float.__sub__,
912 }
913 # Field is of the form key1.key2...
914 # where keys (except first) can be string, int or slice
915 FIELD_RE = r'\w+(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
916 MATH_FIELD_RE = r'''{field}|{num}'''.format(field=FIELD_RE, num=r'-?\d+(?:.\d+)?')
917 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
918 INTERNAL_FORMAT_RE = re.compile(r'''(?x)
919 (?P<negate>-)?
920 (?P<fields>{field})
921 (?P<maths>(?:{math_op}{math_field})*)
922 (?:>(?P<strf_format>.+?))?
923 (?:\|(?P<default>.*?))?
924 $'''.format(field=FIELD_RE, math_op=MATH_OPERATORS_RE, math_field=MATH_FIELD_RE))
925
926 get_key = lambda k: traverse_obj(
927 info_dict, k.split('.'), is_user_input=True, traverse_string=True)
928
929 def get_value(mdict):
930 # Object traversal
931 value = get_key(mdict['fields'])
932 # Negative
933 if mdict['negate']:
934 value = float_or_none(value)
935 if value is not None:
936 value *= -1
937 # Do maths
938 offset_key = mdict['maths']
939 if offset_key:
940 value = float_or_none(value)
941 operator = None
942 while offset_key:
943 item = re.match(
944 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
945 offset_key).group(0)
946 offset_key = offset_key[len(item):]
947 if operator is None:
948 operator = MATH_FUNCTIONS[item]
949 continue
950 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
951 offset = float_or_none(item)
952 if offset is None:
953 offset = float_or_none(get_key(item))
954 try:
955 value = operator(value, multiplier * offset)
956 except (TypeError, ZeroDivisionError):
957 return None
958 operator = None
959 # Datetime formatting
960 if mdict['strf_format']:
961 value = strftime_or_none(value, mdict['strf_format'])
962
963 return value
964
965 def create_key(outer_mobj):
966 if not outer_mobj.group('has_key'):
967 return f'%{outer_mobj.group(0)}'
968
969 prefix = outer_mobj.group('prefix')
970 key = outer_mobj.group('key')
971 original_fmt = fmt = outer_mobj.group('format')
972 mobj = re.match(INTERNAL_FORMAT_RE, key)
973 if mobj is None:
974 value, default, mobj = None, na, {'fields': ''}
975 else:
976 mobj = mobj.groupdict()
977 default = mobj['default'] if mobj['default'] is not None else na
978 value = get_value(mobj)
979
980 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
981 fmt = '0{:d}d'.format(field_size_compat_map[key])
982
983 value = default if value is None else value
984
985 str_fmt = f'{fmt[:-1]}s'
986 if fmt[-1] == 'l':
987 value, fmt = ', '.join(variadic(value)), str_fmt
988 elif fmt[-1] == 'j':
989 value, fmt = json.dumps(value), str_fmt
990 elif fmt[-1] == 'q':
991 value, fmt = compat_shlex_quote(str(value)), str_fmt
992 elif fmt[-1] == 'c':
993 value = str(value)
994 if value is None:
995 value, fmt = default, 's'
996 else:
997 value = value[0]
998 elif fmt[-1] not in 'rs': # numeric
999 value = float_or_none(value)
1000 if value is None:
1001 value, fmt = default, 's'
1002
1003 if sanitize:
1004 if fmt[-1] == 'r':
1005 # If value is an object, sanitize might convert it to a string
1006 # So we convert it to repr first
1007 value, fmt = repr(value), str_fmt
1008 if fmt[-1] in 'csr':
1009 value = sanitize(mobj['fields'].split('.')[-1], value)
1010
1011 key = '%s\0%s' % (key.replace('%', '%\0'), original_fmt)
1012 TMPL_DICT[key] = value
1013 return f'{prefix}%({key}){fmt}'
1014
1015 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1016
1017 def _prepare_filename(self, info_dict, tmpl_type='default'):
1018 try:
1019 sanitize = lambda k, v: sanitize_filename(
1020 compat_str(v),
1021 restricted=self.params.get('restrictfilenames'),
1022 is_id=(k == 'id' or k.endswith('_id')))
1023 outtmpl = self.outtmpl_dict.get(tmpl_type, self.outtmpl_dict['default'])
1024 outtmpl, template_dict = self.prepare_outtmpl(outtmpl, info_dict, sanitize)
1025 outtmpl = self.escape_outtmpl(self._outtmpl_expandpath(outtmpl))
1026 filename = outtmpl % template_dict
1027
1028 force_ext = OUTTMPL_TYPES.get(tmpl_type)
1029 if force_ext is not None:
1030 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1031
1032 # https://github.com/blackjack4494/youtube-dlc/issues/85
1033 trim_file_name = self.params.get('trim_file_name', False)
1034 if trim_file_name:
1035 fn_groups = filename.rsplit('.')
1036 ext = fn_groups[-1]
1037 sub_ext = ''
1038 if len(fn_groups) > 2:
1039 sub_ext = fn_groups[-2]
1040 filename = '.'.join(filter(None, [fn_groups[0][:trim_file_name], sub_ext, ext]))
1041
1042 return filename
1043 except ValueError as err:
1044 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1045 return None
1046
1047 def prepare_filename(self, info_dict, dir_type='', warn=False):
1048 """Generate the output filename."""
1049
1050 filename = self._prepare_filename(info_dict, dir_type or 'default')
1051
1052 if warn:
1053 if not self.params.get('paths'):
1054 pass
1055 elif filename == '-':
1056 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1057 elif os.path.isabs(filename):
1058 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1059 self.__prepare_filename_warned = True
1060 if filename == '-' or not filename:
1061 return filename
1062
1063 return self.get_output_path(dir_type, filename)
1064
1065 def _match_entry(self, info_dict, incomplete=False, silent=False):
1066 """ Returns None if the file should be downloaded """
1067
1068 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1069
1070 def check_filter():
1071 if 'title' in info_dict:
1072 # This can happen when we're just evaluating the playlist
1073 title = info_dict['title']
1074 matchtitle = self.params.get('matchtitle', False)
1075 if matchtitle:
1076 if not re.search(matchtitle, title, re.IGNORECASE):
1077 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1078 rejecttitle = self.params.get('rejecttitle', False)
1079 if rejecttitle:
1080 if re.search(rejecttitle, title, re.IGNORECASE):
1081 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1082 date = info_dict.get('upload_date')
1083 if date is not None:
1084 dateRange = self.params.get('daterange', DateRange())
1085 if date not in dateRange:
1086 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
1087 view_count = info_dict.get('view_count')
1088 if view_count is not None:
1089 min_views = self.params.get('min_views')
1090 if min_views is not None and view_count < min_views:
1091 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1092 max_views = self.params.get('max_views')
1093 if max_views is not None and view_count > max_views:
1094 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1095 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1096 return 'Skipping "%s" because it is age restricted' % video_title
1097
1098 if not incomplete:
1099 match_filter = self.params.get('match_filter')
1100 if match_filter is not None:
1101 ret = match_filter(info_dict)
1102 if ret is not None:
1103 return ret
1104 return None
1105
1106 if self.in_download_archive(info_dict):
1107 reason = '%s has already been recorded in the archive' % video_title
1108 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1109 else:
1110 reason = check_filter()
1111 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1112 if reason is not None:
1113 if not silent:
1114 self.to_screen('[download] ' + reason)
1115 if self.params.get(break_opt, False):
1116 raise break_err()
1117 return reason
1118
1119 @staticmethod
1120 def add_extra_info(info_dict, extra_info):
1121 '''Set the keys from extra_info in info dict if they are missing'''
1122 for key, value in extra_info.items():
1123 info_dict.setdefault(key, value)
1124
1125 def extract_info(self, url, download=True, ie_key=None, extra_info={},
1126 process=True, force_generic_extractor=False):
1127 """
1128 Return a list with a dictionary for each video extracted.
1129
1130 Arguments:
1131 url -- URL to extract
1132
1133 Keyword arguments:
1134 download -- whether to download videos during extraction
1135 ie_key -- extractor key hint
1136 extra_info -- dictionary containing the extra values to add to each result
1137 process -- whether to resolve all unresolved references (URLs, playlist items),
1138 must be True for download to work.
1139 force_generic_extractor -- force using the generic extractor
1140 """
1141
1142 if not ie_key and force_generic_extractor:
1143 ie_key = 'Generic'
1144
1145 if ie_key:
1146 ies = [self.get_info_extractor(ie_key)]
1147 else:
1148 ies = self._ies
1149
1150 for ie in ies:
1151 if not ie.suitable(url):
1152 continue
1153
1154 ie_key = ie.ie_key()
1155 ie = self.get_info_extractor(ie_key)
1156 if not ie.working():
1157 self.report_warning('The program functionality for this site has been marked as broken, '
1158 'and will probably not work.')
1159
1160 try:
1161 temp_id = str_or_none(
1162 ie.extract_id(url) if callable(getattr(ie, 'extract_id', None))
1163 else ie._match_id(url))
1164 except (AssertionError, IndexError, AttributeError):
1165 temp_id = None
1166 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1167 self.to_screen("[%s] %s: has already been recorded in archive" % (
1168 ie_key, temp_id))
1169 break
1170 return self.__extract_info(url, ie, download, extra_info, process)
1171 else:
1172 self.report_error('no suitable InfoExtractor for URL %s' % url)
1173
1174 def __handle_extraction_exceptions(func, handle_all_errors=True):
1175 def wrapper(self, *args, **kwargs):
1176 try:
1177 return func(self, *args, **kwargs)
1178 except GeoRestrictedError as e:
1179 msg = e.msg
1180 if e.countries:
1181 msg += '\nThis video is available in %s.' % ', '.join(
1182 map(ISO3166Utils.short2full, e.countries))
1183 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1184 self.report_error(msg)
1185 except ExtractorError as e: # An error we somewhat expected
1186 self.report_error(compat_str(e), e.format_traceback())
1187 except ThrottledDownload:
1188 self.to_stderr('\r')
1189 self.report_warning('The download speed is below throttle limit. Re-extracting data')
1190 return wrapper(self, *args, **kwargs)
1191 except (MaxDownloadsReached, ExistingVideoReached, RejectedVideoReached):
1192 raise
1193 except Exception as e:
1194 if handle_all_errors and self.params.get('ignoreerrors', False):
1195 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
1196 else:
1197 raise
1198 return wrapper
1199
1200 @__handle_extraction_exceptions
1201 def __extract_info(self, url, ie, download, extra_info, process):
1202 ie_result = ie.extract(url)
1203 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1204 return
1205 if isinstance(ie_result, list):
1206 # Backwards compatibility: old IE result format
1207 ie_result = {
1208 '_type': 'compat_list',
1209 'entries': ie_result,
1210 }
1211 if extra_info.get('original_url'):
1212 ie_result.setdefault('original_url', extra_info['original_url'])
1213 self.add_default_extra_info(ie_result, ie, url)
1214 if process:
1215 return self.process_ie_result(ie_result, download, extra_info)
1216 else:
1217 return ie_result
1218
1219 def add_default_extra_info(self, ie_result, ie, url):
1220 if url is not None:
1221 self.add_extra_info(ie_result, {
1222 'webpage_url': url,
1223 'original_url': url,
1224 'webpage_url_basename': url_basename(url),
1225 })
1226 if ie is not None:
1227 self.add_extra_info(ie_result, {
1228 'extractor': ie.IE_NAME,
1229 'extractor_key': ie.ie_key(),
1230 })
1231
1232 def process_ie_result(self, ie_result, download=True, extra_info={}):
1233 """
1234 Take the result of the ie(may be modified) and resolve all unresolved
1235 references (URLs, playlist items).
1236
1237 It will also download the videos if 'download'.
1238 Returns the resolved ie_result.
1239 """
1240 result_type = ie_result.get('_type', 'video')
1241
1242 if result_type in ('url', 'url_transparent'):
1243 ie_result['url'] = sanitize_url(ie_result['url'])
1244 if ie_result.get('original_url'):
1245 extra_info.setdefault('original_url', ie_result['original_url'])
1246
1247 extract_flat = self.params.get('extract_flat', False)
1248 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1249 or extract_flat is True):
1250 info_copy = ie_result.copy()
1251 self.add_extra_info(info_copy, extra_info)
1252 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1253 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1254 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1255 return ie_result
1256
1257 if result_type == 'video':
1258 self.add_extra_info(ie_result, extra_info)
1259 ie_result = self.process_video_result(ie_result, download=download)
1260 additional_urls = (ie_result or {}).get('additional_urls')
1261 if additional_urls:
1262 # TODO: Improve MetadataFromFieldPP to allow setting a list
1263 if isinstance(additional_urls, compat_str):
1264 additional_urls = [additional_urls]
1265 self.to_screen(
1266 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1267 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1268 ie_result['additional_entries'] = [
1269 self.extract_info(
1270 url, download, extra_info,
1271 force_generic_extractor=self.params.get('force_generic_extractor'))
1272 for url in additional_urls
1273 ]
1274 return ie_result
1275 elif result_type == 'url':
1276 # We have to add extra_info to the results because it may be
1277 # contained in a playlist
1278 return self.extract_info(
1279 ie_result['url'], download,
1280 ie_key=ie_result.get('ie_key'),
1281 extra_info=extra_info)
1282 elif result_type == 'url_transparent':
1283 # Use the information from the embedding page
1284 info = self.extract_info(
1285 ie_result['url'], ie_key=ie_result.get('ie_key'),
1286 extra_info=extra_info, download=False, process=False)
1287
1288 # extract_info may return None when ignoreerrors is enabled and
1289 # extraction failed with an error, don't crash and return early
1290 # in this case
1291 if not info:
1292 return info
1293
1294 force_properties = dict(
1295 (k, v) for k, v in ie_result.items() if v is not None)
1296 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1297 if f in force_properties:
1298 del force_properties[f]
1299 new_result = info.copy()
1300 new_result.update(force_properties)
1301
1302 # Extracted info may not be a video result (i.e.
1303 # info.get('_type', 'video') != video) but rather an url or
1304 # url_transparent. In such cases outer metadata (from ie_result)
1305 # should be propagated to inner one (info). For this to happen
1306 # _type of info should be overridden with url_transparent. This
1307 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1308 if new_result.get('_type') == 'url':
1309 new_result['_type'] = 'url_transparent'
1310
1311 return self.process_ie_result(
1312 new_result, download=download, extra_info=extra_info)
1313 elif result_type in ('playlist', 'multi_video'):
1314 # Protect from infinite recursion due to recursively nested playlists
1315 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1316 webpage_url = ie_result['webpage_url']
1317 if webpage_url in self._playlist_urls:
1318 self.to_screen(
1319 '[download] Skipping already downloaded playlist: %s'
1320 % ie_result.get('title') or ie_result.get('id'))
1321 return
1322
1323 self._playlist_level += 1
1324 self._playlist_urls.add(webpage_url)
1325 self._sanitize_thumbnails(ie_result)
1326 try:
1327 return self.__process_playlist(ie_result, download)
1328 finally:
1329 self._playlist_level -= 1
1330 if not self._playlist_level:
1331 self._playlist_urls.clear()
1332 elif result_type == 'compat_list':
1333 self.report_warning(
1334 'Extractor %s returned a compat_list result. '
1335 'It needs to be updated.' % ie_result.get('extractor'))
1336
1337 def _fixup(r):
1338 self.add_extra_info(
1339 r,
1340 {
1341 'extractor': ie_result['extractor'],
1342 'webpage_url': ie_result['webpage_url'],
1343 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1344 'extractor_key': ie_result['extractor_key'],
1345 }
1346 )
1347 return r
1348 ie_result['entries'] = [
1349 self.process_ie_result(_fixup(r), download, extra_info)
1350 for r in ie_result['entries']
1351 ]
1352 return ie_result
1353 else:
1354 raise Exception('Invalid result type: %s' % result_type)
1355
1356 def _ensure_dir_exists(self, path):
1357 return make_dir(path, self.report_error)
1358
1359 def __process_playlist(self, ie_result, download):
1360 # We process each entry in the playlist
1361 playlist = ie_result.get('title') or ie_result.get('id')
1362 self.to_screen('[download] Downloading playlist: %s' % playlist)
1363
1364 if 'entries' not in ie_result:
1365 raise EntryNotInPlaylist()
1366 incomplete_entries = bool(ie_result.get('requested_entries'))
1367 if incomplete_entries:
1368 def fill_missing_entries(entries, indexes):
1369 ret = [None] * max(*indexes)
1370 for i, entry in zip(indexes, entries):
1371 ret[i - 1] = entry
1372 return ret
1373 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1374
1375 playlist_results = []
1376
1377 playliststart = self.params.get('playliststart', 1)
1378 playlistend = self.params.get('playlistend')
1379 # For backwards compatibility, interpret -1 as whole list
1380 if playlistend == -1:
1381 playlistend = None
1382
1383 playlistitems_str = self.params.get('playlist_items')
1384 playlistitems = None
1385 if playlistitems_str is not None:
1386 def iter_playlistitems(format):
1387 for string_segment in format.split(','):
1388 if '-' in string_segment:
1389 start, end = string_segment.split('-')
1390 for item in range(int(start), int(end) + 1):
1391 yield int(item)
1392 else:
1393 yield int(string_segment)
1394 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1395
1396 ie_entries = ie_result['entries']
1397 msg = (
1398 'Downloading %d videos' if not isinstance(ie_entries, list)
1399 else 'Collected %d videos; downloading %%d of them' % len(ie_entries))
1400 if not isinstance(ie_entries, (list, PagedList)):
1401 ie_entries = LazyList(ie_entries)
1402
1403 def get_entry(i):
1404 return YoutubeDL.__handle_extraction_exceptions(
1405 lambda self, i: ie_entries[i - 1],
1406 False
1407 )(self, i)
1408
1409 entries = []
1410 for i in playlistitems or itertools.count(playliststart):
1411 if playlistitems is None and playlistend is not None and playlistend < i:
1412 break
1413 entry = None
1414 try:
1415 entry = get_entry(i)
1416 if entry is None:
1417 raise EntryNotInPlaylist()
1418 except (IndexError, EntryNotInPlaylist):
1419 if incomplete_entries:
1420 raise EntryNotInPlaylist()
1421 elif not playlistitems:
1422 break
1423 entries.append(entry)
1424 try:
1425 if entry is not None:
1426 self._match_entry(entry, incomplete=True, silent=True)
1427 except (ExistingVideoReached, RejectedVideoReached):
1428 break
1429 ie_result['entries'] = entries
1430
1431 # Save playlist_index before re-ordering
1432 entries = [
1433 ((playlistitems[i - 1] if playlistitems else i), entry)
1434 for i, entry in enumerate(entries, 1)
1435 if entry is not None]
1436 n_entries = len(entries)
1437
1438 if not playlistitems and (playliststart or playlistend):
1439 playlistitems = list(range(playliststart, playliststart + n_entries))
1440 ie_result['requested_entries'] = playlistitems
1441
1442 if self.params.get('allow_playlist_files', True):
1443 ie_copy = {
1444 'playlist': playlist,
1445 'playlist_id': ie_result.get('id'),
1446 'playlist_title': ie_result.get('title'),
1447 'playlist_uploader': ie_result.get('uploader'),
1448 'playlist_uploader_id': ie_result.get('uploader_id'),
1449 'playlist_index': 0,
1450 }
1451 ie_copy.update(dict(ie_result))
1452
1453 if self.params.get('writeinfojson', False):
1454 infofn = self.prepare_filename(ie_copy, 'pl_infojson')
1455 if not self._ensure_dir_exists(encodeFilename(infofn)):
1456 return
1457 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
1458 self.to_screen('[info] Playlist metadata is already present')
1459 else:
1460 self.to_screen('[info] Writing playlist metadata as JSON to: ' + infofn)
1461 try:
1462 write_json_file(self.filter_requested_info(ie_result, self.params.get('clean_infojson', True)), infofn)
1463 except (OSError, IOError):
1464 self.report_error('Cannot write playlist metadata to JSON file ' + infofn)
1465
1466 # TODO: This should be passed to ThumbnailsConvertor if necessary
1467 self._write_thumbnails(ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1468
1469 if self.params.get('writedescription', False):
1470 descfn = self.prepare_filename(ie_copy, 'pl_description')
1471 if not self._ensure_dir_exists(encodeFilename(descfn)):
1472 return
1473 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
1474 self.to_screen('[info] Playlist description is already present')
1475 elif ie_result.get('description') is None:
1476 self.report_warning('There\'s no playlist description to write.')
1477 else:
1478 try:
1479 self.to_screen('[info] Writing playlist description to: ' + descfn)
1480 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1481 descfile.write(ie_result['description'])
1482 except (OSError, IOError):
1483 self.report_error('Cannot write playlist description file ' + descfn)
1484 return
1485
1486 if self.params.get('playlistreverse', False):
1487 entries = entries[::-1]
1488 if self.params.get('playlistrandom', False):
1489 random.shuffle(entries)
1490
1491 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1492
1493 self.to_screen('[%s] playlist %s: %s' % (ie_result['extractor'], playlist, msg % n_entries))
1494 failures = 0
1495 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1496 for i, entry_tuple in enumerate(entries, 1):
1497 playlist_index, entry = entry_tuple
1498 if 'playlist_index' in self.params.get('compat_options', []):
1499 playlist_index = playlistitems[i - 1] if playlistitems else i
1500 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1501 # This __x_forwarded_for_ip thing is a bit ugly but requires
1502 # minimal changes
1503 if x_forwarded_for:
1504 entry['__x_forwarded_for_ip'] = x_forwarded_for
1505 extra = {
1506 'n_entries': n_entries,
1507 '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1508 'playlist_index': playlist_index,
1509 'playlist_autonumber': i,
1510 'playlist': playlist,
1511 'playlist_id': ie_result.get('id'),
1512 'playlist_title': ie_result.get('title'),
1513 'playlist_uploader': ie_result.get('uploader'),
1514 'playlist_uploader_id': ie_result.get('uploader_id'),
1515 'extractor': ie_result['extractor'],
1516 'webpage_url': ie_result['webpage_url'],
1517 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1518 'extractor_key': ie_result['extractor_key'],
1519 }
1520
1521 if self._match_entry(entry, incomplete=True) is not None:
1522 continue
1523
1524 entry_result = self.__process_iterable_entry(entry, download, extra)
1525 if not entry_result:
1526 failures += 1
1527 if failures >= max_failures:
1528 self.report_error(
1529 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1530 break
1531 # TODO: skip failed (empty) entries?
1532 playlist_results.append(entry_result)
1533 ie_result['entries'] = playlist_results
1534 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1535 return ie_result
1536
1537 @__handle_extraction_exceptions
1538 def __process_iterable_entry(self, entry, download, extra_info):
1539 return self.process_ie_result(
1540 entry, download=download, extra_info=extra_info)
1541
1542 def _build_format_filter(self, filter_spec):
1543 " Returns a function to filter the formats according to the filter_spec "
1544
1545 OPERATORS = {
1546 '<': operator.lt,
1547 '<=': operator.le,
1548 '>': operator.gt,
1549 '>=': operator.ge,
1550 '=': operator.eq,
1551 '!=': operator.ne,
1552 }
1553 operator_rex = re.compile(r'''(?x)\s*
1554 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1555 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1556 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1557 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1558 m = operator_rex.fullmatch(filter_spec)
1559 if m:
1560 try:
1561 comparison_value = int(m.group('value'))
1562 except ValueError:
1563 comparison_value = parse_filesize(m.group('value'))
1564 if comparison_value is None:
1565 comparison_value = parse_filesize(m.group('value') + 'B')
1566 if comparison_value is None:
1567 raise ValueError(
1568 'Invalid value %r in format specification %r' % (
1569 m.group('value'), filter_spec))
1570 op = OPERATORS[m.group('op')]
1571
1572 if not m:
1573 STR_OPERATORS = {
1574 '=': operator.eq,
1575 '^=': lambda attr, value: attr.startswith(value),
1576 '$=': lambda attr, value: attr.endswith(value),
1577 '*=': lambda attr, value: value in attr,
1578 }
1579 str_operator_rex = re.compile(r'''(?x)\s*
1580 (?P<key>[a-zA-Z0-9._-]+)\s*
1581 (?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1582 (?P<value>[a-zA-Z0-9._-]+)\s*
1583 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1584 m = str_operator_rex.fullmatch(filter_spec)
1585 if m:
1586 comparison_value = m.group('value')
1587 str_op = STR_OPERATORS[m.group('op')]
1588 if m.group('negation'):
1589 op = lambda attr, value: not str_op(attr, value)
1590 else:
1591 op = str_op
1592
1593 if not m:
1594 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1595
1596 def _filter(f):
1597 actual_value = f.get(m.group('key'))
1598 if actual_value is None:
1599 return m.group('none_inclusive')
1600 return op(actual_value, comparison_value)
1601 return _filter
1602
1603 def _default_format_spec(self, info_dict, download=True):
1604
1605 def can_merge():
1606 merger = FFmpegMergerPP(self)
1607 return merger.available and merger.can_merge()
1608
1609 prefer_best = (
1610 not self.params.get('simulate', False)
1611 and download
1612 and (
1613 not can_merge()
1614 or info_dict.get('is_live', False)
1615 or self.outtmpl_dict['default'] == '-'))
1616 compat = (
1617 prefer_best
1618 or self.params.get('allow_multiple_audio_streams', False)
1619 or 'format-spec' in self.params.get('compat_opts', []))
1620
1621 return (
1622 'best/bestvideo+bestaudio' if prefer_best
1623 else 'bestvideo*+bestaudio/best' if not compat
1624 else 'bestvideo+bestaudio/best')
1625
1626 def build_format_selector(self, format_spec):
1627 def syntax_error(note, start):
1628 message = (
1629 'Invalid format specification: '
1630 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1631 return SyntaxError(message)
1632
1633 PICKFIRST = 'PICKFIRST'
1634 MERGE = 'MERGE'
1635 SINGLE = 'SINGLE'
1636 GROUP = 'GROUP'
1637 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1638
1639 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1640 'video': self.params.get('allow_multiple_video_streams', False)}
1641
1642 check_formats = self.params.get('check_formats')
1643
1644 def _parse_filter(tokens):
1645 filter_parts = []
1646 for type, string, start, _, _ in tokens:
1647 if type == tokenize.OP and string == ']':
1648 return ''.join(filter_parts)
1649 else:
1650 filter_parts.append(string)
1651
1652 def _remove_unused_ops(tokens):
1653 # Remove operators that we don't use and join them with the surrounding strings
1654 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1655 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1656 last_string, last_start, last_end, last_line = None, None, None, None
1657 for type, string, start, end, line in tokens:
1658 if type == tokenize.OP and string == '[':
1659 if last_string:
1660 yield tokenize.NAME, last_string, last_start, last_end, last_line
1661 last_string = None
1662 yield type, string, start, end, line
1663 # everything inside brackets will be handled by _parse_filter
1664 for type, string, start, end, line in tokens:
1665 yield type, string, start, end, line
1666 if type == tokenize.OP and string == ']':
1667 break
1668 elif type == tokenize.OP and string in ALLOWED_OPS:
1669 if last_string:
1670 yield tokenize.NAME, last_string, last_start, last_end, last_line
1671 last_string = None
1672 yield type, string, start, end, line
1673 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1674 if not last_string:
1675 last_string = string
1676 last_start = start
1677 last_end = end
1678 else:
1679 last_string += string
1680 if last_string:
1681 yield tokenize.NAME, last_string, last_start, last_end, last_line
1682
1683 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1684 selectors = []
1685 current_selector = None
1686 for type, string, start, _, _ in tokens:
1687 # ENCODING is only defined in python 3.x
1688 if type == getattr(tokenize, 'ENCODING', None):
1689 continue
1690 elif type in [tokenize.NAME, tokenize.NUMBER]:
1691 current_selector = FormatSelector(SINGLE, string, [])
1692 elif type == tokenize.OP:
1693 if string == ')':
1694 if not inside_group:
1695 # ')' will be handled by the parentheses group
1696 tokens.restore_last_token()
1697 break
1698 elif inside_merge and string in ['/', ',']:
1699 tokens.restore_last_token()
1700 break
1701 elif inside_choice and string == ',':
1702 tokens.restore_last_token()
1703 break
1704 elif string == ',':
1705 if not current_selector:
1706 raise syntax_error('"," must follow a format selector', start)
1707 selectors.append(current_selector)
1708 current_selector = None
1709 elif string == '/':
1710 if not current_selector:
1711 raise syntax_error('"/" must follow a format selector', start)
1712 first_choice = current_selector
1713 second_choice = _parse_format_selection(tokens, inside_choice=True)
1714 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1715 elif string == '[':
1716 if not current_selector:
1717 current_selector = FormatSelector(SINGLE, 'best', [])
1718 format_filter = _parse_filter(tokens)
1719 current_selector.filters.append(format_filter)
1720 elif string == '(':
1721 if current_selector:
1722 raise syntax_error('Unexpected "("', start)
1723 group = _parse_format_selection(tokens, inside_group=True)
1724 current_selector = FormatSelector(GROUP, group, [])
1725 elif string == '+':
1726 if not current_selector:
1727 raise syntax_error('Unexpected "+"', start)
1728 selector_1 = current_selector
1729 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1730 if not selector_2:
1731 raise syntax_error('Expected a selector', start)
1732 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1733 else:
1734 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1735 elif type == tokenize.ENDMARKER:
1736 break
1737 if current_selector:
1738 selectors.append(current_selector)
1739 return selectors
1740
1741 def _merge(formats_pair):
1742 format_1, format_2 = formats_pair
1743
1744 formats_info = []
1745 formats_info.extend(format_1.get('requested_formats', (format_1,)))
1746 formats_info.extend(format_2.get('requested_formats', (format_2,)))
1747
1748 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
1749 get_no_more = {'video': False, 'audio': False}
1750 for (i, fmt_info) in enumerate(formats_info):
1751 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
1752 formats_info.pop(i)
1753 continue
1754 for aud_vid in ['audio', 'video']:
1755 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
1756 if get_no_more[aud_vid]:
1757 formats_info.pop(i)
1758 get_no_more[aud_vid] = True
1759
1760 if len(formats_info) == 1:
1761 return formats_info[0]
1762
1763 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
1764 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
1765
1766 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
1767 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
1768
1769 output_ext = self.params.get('merge_output_format')
1770 if not output_ext:
1771 if the_only_video:
1772 output_ext = the_only_video['ext']
1773 elif the_only_audio and not video_fmts:
1774 output_ext = the_only_audio['ext']
1775 else:
1776 output_ext = 'mkv'
1777
1778 new_dict = {
1779 'requested_formats': formats_info,
1780 'format': '+'.join(fmt_info.get('format') for fmt_info in formats_info),
1781 'format_id': '+'.join(fmt_info.get('format_id') for fmt_info in formats_info),
1782 'ext': output_ext,
1783 }
1784
1785 if the_only_video:
1786 new_dict.update({
1787 'width': the_only_video.get('width'),
1788 'height': the_only_video.get('height'),
1789 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
1790 'fps': the_only_video.get('fps'),
1791 'vcodec': the_only_video.get('vcodec'),
1792 'vbr': the_only_video.get('vbr'),
1793 'stretched_ratio': the_only_video.get('stretched_ratio'),
1794 })
1795
1796 if the_only_audio:
1797 new_dict.update({
1798 'acodec': the_only_audio.get('acodec'),
1799 'abr': the_only_audio.get('abr'),
1800 })
1801
1802 return new_dict
1803
1804 def _check_formats(formats):
1805 if not check_formats:
1806 yield from formats
1807 return
1808 for f in formats:
1809 self.to_screen('[info] Testing format %s' % f['format_id'])
1810 temp_file = tempfile.NamedTemporaryFile(
1811 suffix='.tmp', delete=False,
1812 dir=self.get_output_path('temp') or None)
1813 temp_file.close()
1814 try:
1815 success, _ = self.dl(temp_file.name, f, test=True)
1816 except (DownloadError, IOError, OSError, ValueError) + network_exceptions:
1817 success = False
1818 finally:
1819 if os.path.exists(temp_file.name):
1820 try:
1821 os.remove(temp_file.name)
1822 except OSError:
1823 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1824 if success:
1825 yield f
1826 else:
1827 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1828
1829 def _build_selector_function(selector):
1830 if isinstance(selector, list): # ,
1831 fs = [_build_selector_function(s) for s in selector]
1832
1833 def selector_function(ctx):
1834 for f in fs:
1835 yield from f(ctx)
1836 return selector_function
1837
1838 elif selector.type == GROUP: # ()
1839 selector_function = _build_selector_function(selector.selector)
1840
1841 elif selector.type == PICKFIRST: # /
1842 fs = [_build_selector_function(s) for s in selector.selector]
1843
1844 def selector_function(ctx):
1845 for f in fs:
1846 picked_formats = list(f(ctx))
1847 if picked_formats:
1848 return picked_formats
1849 return []
1850
1851 elif selector.type == MERGE: # +
1852 selector_1, selector_2 = map(_build_selector_function, selector.selector)
1853
1854 def selector_function(ctx):
1855 for pair in itertools.product(
1856 selector_1(copy.deepcopy(ctx)), selector_2(copy.deepcopy(ctx))):
1857 yield _merge(pair)
1858
1859 elif selector.type == SINGLE: # atom
1860 format_spec = selector.selector or 'best'
1861
1862 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
1863 if format_spec == 'all':
1864 def selector_function(ctx):
1865 yield from _check_formats(ctx['formats'])
1866 elif format_spec == 'mergeall':
1867 def selector_function(ctx):
1868 formats = list(_check_formats(ctx['formats']))
1869 if not formats:
1870 return
1871 merged_format = formats[-1]
1872 for f in formats[-2::-1]:
1873 merged_format = _merge((merged_format, f))
1874 yield merged_format
1875
1876 else:
1877 format_fallback, format_reverse, format_idx = False, True, 1
1878 mobj = re.match(
1879 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
1880 format_spec)
1881 if mobj is not None:
1882 format_idx = int_or_none(mobj.group('n'), default=1)
1883 format_reverse = mobj.group('bw')[0] == 'b'
1884 format_type = (mobj.group('type') or [None])[0]
1885 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
1886 format_modified = mobj.group('mod') is not None
1887
1888 format_fallback = not format_type and not format_modified # for b, w
1889 _filter_f = (
1890 (lambda f: f.get('%scodec' % format_type) != 'none')
1891 if format_type and format_modified # bv*, ba*, wv*, wa*
1892 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
1893 if format_type # bv, ba, wv, wa
1894 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
1895 if not format_modified # b, w
1896 else lambda f: True) # b*, w*
1897 filter_f = lambda f: _filter_f(f) and (
1898 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
1899 else:
1900 filter_f = ((lambda f: f.get('ext') == format_spec)
1901 if format_spec in ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] # extension
1902 else (lambda f: f.get('format_id') == format_spec)) # id
1903
1904 def selector_function(ctx):
1905 formats = list(ctx['formats'])
1906 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
1907 if format_fallback and ctx['incomplete_formats'] and not matches:
1908 # for extractors with incomplete formats (audio only (soundcloud)
1909 # or video only (imgur)) best/worst will fallback to
1910 # best/worst {video,audio}-only format
1911 matches = formats
1912 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
1913 try:
1914 yield matches[format_idx - 1]
1915 except IndexError:
1916 return
1917
1918 filters = [self._build_format_filter(f) for f in selector.filters]
1919
1920 def final_selector(ctx):
1921 ctx_copy = copy.deepcopy(ctx)
1922 for _filter in filters:
1923 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1924 return selector_function(ctx_copy)
1925 return final_selector
1926
1927 stream = io.BytesIO(format_spec.encode('utf-8'))
1928 try:
1929 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1930 except tokenize.TokenError:
1931 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1932
1933 class TokenIterator(object):
1934 def __init__(self, tokens):
1935 self.tokens = tokens
1936 self.counter = 0
1937
1938 def __iter__(self):
1939 return self
1940
1941 def __next__(self):
1942 if self.counter >= len(self.tokens):
1943 raise StopIteration()
1944 value = self.tokens[self.counter]
1945 self.counter += 1
1946 return value
1947
1948 next = __next__
1949
1950 def restore_last_token(self):
1951 self.counter -= 1
1952
1953 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1954 return _build_selector_function(parsed_selector)
1955
1956 def _calc_headers(self, info_dict):
1957 res = std_headers.copy()
1958
1959 add_headers = info_dict.get('http_headers')
1960 if add_headers:
1961 res.update(add_headers)
1962
1963 cookies = self._calc_cookies(info_dict)
1964 if cookies:
1965 res['Cookie'] = cookies
1966
1967 if 'X-Forwarded-For' not in res:
1968 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1969 if x_forwarded_for_ip:
1970 res['X-Forwarded-For'] = x_forwarded_for_ip
1971
1972 return res
1973
1974 def _calc_cookies(self, info_dict):
1975 pr = sanitized_Request(info_dict['url'])
1976 self.cookiejar.add_cookie_header(pr)
1977 return pr.get_header('Cookie')
1978
1979 def _sanitize_thumbnails(self, info_dict):
1980 thumbnails = info_dict.get('thumbnails')
1981 if thumbnails is None:
1982 thumbnail = info_dict.get('thumbnail')
1983 if thumbnail:
1984 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1985 if thumbnails:
1986 thumbnails.sort(key=lambda t: (
1987 t.get('preference') if t.get('preference') is not None else -1,
1988 t.get('width') if t.get('width') is not None else -1,
1989 t.get('height') if t.get('height') is not None else -1,
1990 t.get('id') if t.get('id') is not None else '',
1991 t.get('url')))
1992
1993 def thumbnail_tester():
1994 if self.params.get('check_formats'):
1995 test_all = True
1996 to_screen = lambda msg: self.to_screen(f'[info] {msg}')
1997 else:
1998 test_all = False
1999 to_screen = self.write_debug
2000
2001 def test_thumbnail(t):
2002 if not test_all and not t.get('_test_url'):
2003 return True
2004 to_screen('Testing thumbnail %s' % t['id'])
2005 try:
2006 self.urlopen(HEADRequest(t['url']))
2007 except network_exceptions as err:
2008 to_screen('Unable to connect to thumbnail %s URL "%s" - %s. Skipping...' % (
2009 t['id'], t['url'], error_to_compat_str(err)))
2010 return False
2011 return True
2012
2013 return test_thumbnail
2014
2015 for i, t in enumerate(thumbnails):
2016 if t.get('id') is None:
2017 t['id'] = '%d' % i
2018 if t.get('width') and t.get('height'):
2019 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2020 t['url'] = sanitize_url(t['url'])
2021
2022 if self.params.get('check_formats') is not False:
2023 info_dict['thumbnails'] = LazyList(filter(thumbnail_tester(), thumbnails[::-1])).reverse()
2024 else:
2025 info_dict['thumbnails'] = thumbnails
2026
2027 def process_video_result(self, info_dict, download=True):
2028 assert info_dict.get('_type', 'video') == 'video'
2029
2030 if 'id' not in info_dict:
2031 raise ExtractorError('Missing "id" field in extractor result')
2032 if 'title' not in info_dict:
2033 raise ExtractorError('Missing "title" field in extractor result')
2034
2035 def report_force_conversion(field, field_not, conversion):
2036 self.report_warning(
2037 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2038 % (field, field_not, conversion))
2039
2040 def sanitize_string_field(info, string_field):
2041 field = info.get(string_field)
2042 if field is None or isinstance(field, compat_str):
2043 return
2044 report_force_conversion(string_field, 'a string', 'string')
2045 info[string_field] = compat_str(field)
2046
2047 def sanitize_numeric_fields(info):
2048 for numeric_field in self._NUMERIC_FIELDS:
2049 field = info.get(numeric_field)
2050 if field is None or isinstance(field, compat_numeric_types):
2051 continue
2052 report_force_conversion(numeric_field, 'numeric', 'int')
2053 info[numeric_field] = int_or_none(field)
2054
2055 sanitize_string_field(info_dict, 'id')
2056 sanitize_numeric_fields(info_dict)
2057
2058 if 'playlist' not in info_dict:
2059 # It isn't part of a playlist
2060 info_dict['playlist'] = None
2061 info_dict['playlist_index'] = None
2062
2063 self._sanitize_thumbnails(info_dict)
2064
2065 thumbnail = info_dict.get('thumbnail')
2066 thumbnails = info_dict.get('thumbnails')
2067 if thumbnail:
2068 info_dict['thumbnail'] = sanitize_url(thumbnail)
2069 elif thumbnails:
2070 info_dict['thumbnail'] = thumbnails[-1]['url']
2071
2072 if info_dict.get('display_id') is None and 'id' in info_dict:
2073 info_dict['display_id'] = info_dict['id']
2074
2075 for ts_key, date_key in (
2076 ('timestamp', 'upload_date'),
2077 ('release_timestamp', 'release_date'),
2078 ):
2079 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2080 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2081 # see http://bugs.python.org/issue1646728)
2082 try:
2083 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2084 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2085 except (ValueError, OverflowError, OSError):
2086 pass
2087
2088 live_keys = ('is_live', 'was_live')
2089 live_status = info_dict.get('live_status')
2090 if live_status is None:
2091 for key in live_keys:
2092 if info_dict.get(key) is False:
2093 continue
2094 if info_dict.get(key):
2095 live_status = key
2096 break
2097 if all(info_dict.get(key) is False for key in live_keys):
2098 live_status = 'not_live'
2099 if live_status:
2100 info_dict['live_status'] = live_status
2101 for key in live_keys:
2102 if info_dict.get(key) is None:
2103 info_dict[key] = (live_status == key)
2104
2105 # Auto generate title fields corresponding to the *_number fields when missing
2106 # in order to always have clean titles. This is very common for TV series.
2107 for field in ('chapter', 'season', 'episode'):
2108 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2109 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2110
2111 for cc_kind in ('subtitles', 'automatic_captions'):
2112 cc = info_dict.get(cc_kind)
2113 if cc:
2114 for _, subtitle in cc.items():
2115 for subtitle_format in subtitle:
2116 if subtitle_format.get('url'):
2117 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2118 if subtitle_format.get('ext') is None:
2119 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2120
2121 automatic_captions = info_dict.get('automatic_captions')
2122 subtitles = info_dict.get('subtitles')
2123
2124 info_dict['requested_subtitles'] = self.process_subtitles(
2125 info_dict['id'], subtitles, automatic_captions)
2126
2127 # We now pick which formats have to be downloaded
2128 if info_dict.get('formats') is None:
2129 # There's only one format available
2130 formats = [info_dict]
2131 else:
2132 formats = info_dict['formats']
2133
2134 if not formats:
2135 if not self.params.get('ignore_no_formats_error'):
2136 raise ExtractorError('No video formats found!')
2137 else:
2138 self.report_warning('No video formats found!')
2139
2140 def is_wellformed(f):
2141 url = f.get('url')
2142 if not url:
2143 self.report_warning(
2144 '"url" field is missing or empty - skipping format, '
2145 'there is an error in extractor')
2146 return False
2147 if isinstance(url, bytes):
2148 sanitize_string_field(f, 'url')
2149 return True
2150
2151 # Filter out malformed formats for better extraction robustness
2152 formats = list(filter(is_wellformed, formats))
2153
2154 formats_dict = {}
2155
2156 # We check that all the formats have the format and format_id fields
2157 for i, format in enumerate(formats):
2158 sanitize_string_field(format, 'format_id')
2159 sanitize_numeric_fields(format)
2160 format['url'] = sanitize_url(format['url'])
2161 if not format.get('format_id'):
2162 format['format_id'] = compat_str(i)
2163 else:
2164 # Sanitize format_id from characters used in format selector expression
2165 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2166 format_id = format['format_id']
2167 if format_id not in formats_dict:
2168 formats_dict[format_id] = []
2169 formats_dict[format_id].append(format)
2170
2171 # Make sure all formats have unique format_id
2172 for format_id, ambiguous_formats in formats_dict.items():
2173 if len(ambiguous_formats) > 1:
2174 for i, format in enumerate(ambiguous_formats):
2175 format['format_id'] = '%s-%d' % (format_id, i)
2176
2177 for i, format in enumerate(formats):
2178 if format.get('format') is None:
2179 format['format'] = '{id} - {res}{note}'.format(
2180 id=format['format_id'],
2181 res=self.format_resolution(format),
2182 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
2183 )
2184 # Automatically determine file extension if missing
2185 if format.get('ext') is None:
2186 format['ext'] = determine_ext(format['url']).lower()
2187 # Automatically determine protocol if missing (useful for format
2188 # selection purposes)
2189 if format.get('protocol') is None:
2190 format['protocol'] = determine_protocol(format)
2191 # Add HTTP headers, so that external programs can use them from the
2192 # json output
2193 full_format_info = info_dict.copy()
2194 full_format_info.update(format)
2195 format['http_headers'] = self._calc_headers(full_format_info)
2196 # Remove private housekeeping stuff
2197 if '__x_forwarded_for_ip' in info_dict:
2198 del info_dict['__x_forwarded_for_ip']
2199
2200 # TODO Central sorting goes here
2201
2202 if formats and formats[0] is not info_dict:
2203 # only set the 'formats' fields if the original info_dict list them
2204 # otherwise we end up with a circular reference, the first (and unique)
2205 # element in the 'formats' field in info_dict is info_dict itself,
2206 # which can't be exported to json
2207 info_dict['formats'] = formats
2208
2209 info_dict, _ = self.pre_process(info_dict)
2210
2211 list_only = self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles')
2212 if list_only:
2213 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2214 if self.params.get('list_thumbnails'):
2215 self.list_thumbnails(info_dict)
2216 if self.params.get('listformats'):
2217 if not info_dict.get('formats'):
2218 raise ExtractorError('No video formats found', expected=True)
2219 self.list_formats(info_dict)
2220 if self.params.get('listsubtitles'):
2221 if 'automatic_captions' in info_dict:
2222 self.list_subtitles(
2223 info_dict['id'], automatic_captions, 'automatic captions')
2224 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2225 return
2226
2227 format_selector = self.format_selector
2228 if format_selector is None:
2229 req_format = self._default_format_spec(info_dict, download=download)
2230 self.write_debug('Default format spec: %s' % req_format)
2231 format_selector = self.build_format_selector(req_format)
2232
2233 # While in format selection we may need to have an access to the original
2234 # format set in order to calculate some metrics or do some processing.
2235 # For now we need to be able to guess whether original formats provided
2236 # by extractor are incomplete or not (i.e. whether extractor provides only
2237 # video-only or audio-only formats) for proper formats selection for
2238 # extractors with such incomplete formats (see
2239 # https://github.com/ytdl-org/youtube-dl/pull/5556).
2240 # Since formats may be filtered during format selection and may not match
2241 # the original formats the results may be incorrect. Thus original formats
2242 # or pre-calculated metrics should be passed to format selection routines
2243 # as well.
2244 # We will pass a context object containing all necessary additional data
2245 # instead of just formats.
2246 # This fixes incorrect format selection issue (see
2247 # https://github.com/ytdl-org/youtube-dl/issues/10083).
2248 incomplete_formats = (
2249 # All formats are video-only or
2250 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2251 # all formats are audio-only
2252 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
2253
2254 ctx = {
2255 'formats': formats,
2256 'incomplete_formats': incomplete_formats,
2257 }
2258
2259 formats_to_download = list(format_selector(ctx))
2260 if not formats_to_download:
2261 if not self.params.get('ignore_no_formats_error'):
2262 raise ExtractorError('Requested format is not available', expected=True)
2263 else:
2264 self.report_warning('Requested format is not available')
2265 # Process what we can, even without any available formats.
2266 self.process_info(dict(info_dict))
2267 elif download:
2268 self.to_screen(
2269 '[info] %s: Downloading %d format(s): %s' % (
2270 info_dict['id'], len(formats_to_download),
2271 ", ".join([f['format_id'] for f in formats_to_download])))
2272 for fmt in formats_to_download:
2273 new_info = dict(info_dict)
2274 # Save a reference to the original info_dict so that it can be modified in process_info if needed
2275 new_info['__original_infodict'] = info_dict
2276 new_info.update(fmt)
2277 self.process_info(new_info)
2278 # We update the info dict with the best quality format (backwards compatibility)
2279 if formats_to_download:
2280 info_dict.update(formats_to_download[-1])
2281 return info_dict
2282
2283 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2284 """Select the requested subtitles and their format"""
2285 available_subs = {}
2286 if normal_subtitles and self.params.get('writesubtitles'):
2287 available_subs.update(normal_subtitles)
2288 if automatic_captions and self.params.get('writeautomaticsub'):
2289 for lang, cap_info in automatic_captions.items():
2290 if lang not in available_subs:
2291 available_subs[lang] = cap_info
2292
2293 if (not self.params.get('writesubtitles') and not
2294 self.params.get('writeautomaticsub') or not
2295 available_subs):
2296 return None
2297
2298 all_sub_langs = available_subs.keys()
2299 if self.params.get('allsubtitles', False):
2300 requested_langs = all_sub_langs
2301 elif self.params.get('subtitleslangs', False):
2302 requested_langs = set()
2303 for lang in self.params.get('subtitleslangs'):
2304 if lang == 'all':
2305 requested_langs.update(all_sub_langs)
2306 continue
2307 discard = lang[0] == '-'
2308 if discard:
2309 lang = lang[1:]
2310 current_langs = filter(re.compile(lang + '$').match, all_sub_langs)
2311 if discard:
2312 for lang in current_langs:
2313 requested_langs.discard(lang)
2314 else:
2315 requested_langs.update(current_langs)
2316 elif 'en' in available_subs:
2317 requested_langs = ['en']
2318 else:
2319 requested_langs = [list(all_sub_langs)[0]]
2320 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2321
2322 formats_query = self.params.get('subtitlesformat', 'best')
2323 formats_preference = formats_query.split('/') if formats_query else []
2324 subs = {}
2325 for lang in requested_langs:
2326 formats = available_subs.get(lang)
2327 if formats is None:
2328 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
2329 continue
2330 for ext in formats_preference:
2331 if ext == 'best':
2332 f = formats[-1]
2333 break
2334 matches = list(filter(lambda f: f['ext'] == ext, formats))
2335 if matches:
2336 f = matches[-1]
2337 break
2338 else:
2339 f = formats[-1]
2340 self.report_warning(
2341 'No subtitle format found matching "%s" for language %s, '
2342 'using %s' % (formats_query, lang, f['ext']))
2343 subs[lang] = f
2344 return subs
2345
2346 def __forced_printings(self, info_dict, filename, incomplete):
2347 def print_mandatory(field, actual_field=None):
2348 if actual_field is None:
2349 actual_field = field
2350 if (self.params.get('force%s' % field, False)
2351 and (not incomplete or info_dict.get(actual_field) is not None)):
2352 self.to_stdout(info_dict[actual_field])
2353
2354 def print_optional(field):
2355 if (self.params.get('force%s' % field, False)
2356 and info_dict.get(field) is not None):
2357 self.to_stdout(info_dict[field])
2358
2359 info_dict = info_dict.copy()
2360 if filename is not None:
2361 info_dict['filename'] = filename
2362 if info_dict.get('requested_formats') is not None:
2363 # For RTMP URLs, also include the playpath
2364 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2365 elif 'url' in info_dict:
2366 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2367
2368 for tmpl in self.params.get('forceprint', []):
2369 if re.match(r'\w+$', tmpl):
2370 tmpl = '%({})s'.format(tmpl)
2371 tmpl, info_copy = self.prepare_outtmpl(tmpl, info_dict)
2372 self.to_stdout(self.escape_outtmpl(tmpl) % info_copy)
2373
2374 print_mandatory('title')
2375 print_mandatory('id')
2376 print_mandatory('url', 'urls')
2377 print_optional('thumbnail')
2378 print_optional('description')
2379 print_optional('filename')
2380 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
2381 self.to_stdout(formatSeconds(info_dict['duration']))
2382 print_mandatory('format')
2383
2384 if self.params.get('forcejson', False):
2385 self.post_extract(info_dict)
2386 self.to_stdout(json.dumps(info_dict, default=repr))
2387
2388 def dl(self, name, info, subtitle=False, test=False):
2389
2390 if test:
2391 verbose = self.params.get('verbose')
2392 params = {
2393 'test': True,
2394 'quiet': not verbose,
2395 'verbose': verbose,
2396 'noprogress': not verbose,
2397 'nopart': True,
2398 'skip_unavailable_fragments': False,
2399 'keep_fragments': False,
2400 'overwrites': True,
2401 '_no_ytdl_file': True,
2402 }
2403 else:
2404 params = self.params
2405 fd = get_suitable_downloader(info, params)(self, params)
2406 if not test:
2407 for ph in self._progress_hooks:
2408 fd.add_progress_hook(ph)
2409 urls = '", "'.join([f['url'] for f in info.get('requested_formats', [])] or [info['url']])
2410 self.write_debug('Invoking downloader on "%s"' % urls)
2411 new_info = dict(info)
2412 if new_info.get('http_headers') is None:
2413 new_info['http_headers'] = self._calc_headers(new_info)
2414 return fd.download(name, new_info, subtitle)
2415
2416 def process_info(self, info_dict):
2417 """Process a single resolved IE result."""
2418
2419 assert info_dict.get('_type', 'video') == 'video'
2420
2421 info_dict.setdefault('__postprocessors', [])
2422
2423 max_downloads = self.params.get('max_downloads')
2424 if max_downloads is not None:
2425 if self._num_downloads >= int(max_downloads):
2426 raise MaxDownloadsReached()
2427
2428 # TODO: backward compatibility, to be removed
2429 info_dict['fulltitle'] = info_dict['title']
2430
2431 if 'format' not in info_dict and 'ext' in info_dict:
2432 info_dict['format'] = info_dict['ext']
2433
2434 if self._match_entry(info_dict) is not None:
2435 return
2436
2437 self.post_extract(info_dict)
2438 self._num_downloads += 1
2439
2440 # info_dict['_filename'] needs to be set for backward compatibility
2441 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2442 temp_filename = self.prepare_filename(info_dict, 'temp')
2443 files_to_move = {}
2444
2445 # Forced printings
2446 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2447
2448 if self.params.get('simulate', False):
2449 if self.params.get('force_write_download_archive', False):
2450 self.record_download_archive(info_dict)
2451
2452 # Do nothing else if in simulate mode
2453 return
2454
2455 if full_filename is None:
2456 return
2457
2458 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2459 return
2460 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2461 return
2462
2463 if self.params.get('writedescription', False):
2464 descfn = self.prepare_filename(info_dict, 'description')
2465 if not self._ensure_dir_exists(encodeFilename(descfn)):
2466 return
2467 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
2468 self.to_screen('[info] Video description is already present')
2469 elif info_dict.get('description') is None:
2470 self.report_warning('There\'s no description to write.')
2471 else:
2472 try:
2473 self.to_screen('[info] Writing video description to: ' + descfn)
2474 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
2475 descfile.write(info_dict['description'])
2476 except (OSError, IOError):
2477 self.report_error('Cannot write description file ' + descfn)
2478 return
2479
2480 if self.params.get('writeannotations', False):
2481 annofn = self.prepare_filename(info_dict, 'annotation')
2482 if not self._ensure_dir_exists(encodeFilename(annofn)):
2483 return
2484 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2485 self.to_screen('[info] Video annotations are already present')
2486 elif not info_dict.get('annotations'):
2487 self.report_warning('There are no annotations to write.')
2488 else:
2489 try:
2490 self.to_screen('[info] Writing video annotations to: ' + annofn)
2491 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2492 annofile.write(info_dict['annotations'])
2493 except (KeyError, TypeError):
2494 self.report_warning('There are no annotations to write.')
2495 except (OSError, IOError):
2496 self.report_error('Cannot write annotations file: ' + annofn)
2497 return
2498
2499 subtitles_are_requested = any([self.params.get('writesubtitles', False),
2500 self.params.get('writeautomaticsub')])
2501
2502 if subtitles_are_requested and info_dict.get('requested_subtitles'):
2503 # subtitles download errors are already managed as troubles in relevant IE
2504 # that way it will silently go on when used with unsupporting IE
2505 subtitles = info_dict['requested_subtitles']
2506 # ie = self.get_info_extractor(info_dict['extractor_key'])
2507 for sub_lang, sub_info in subtitles.items():
2508 sub_format = sub_info['ext']
2509 sub_filename = subtitles_filename(temp_filename, sub_lang, sub_format, info_dict.get('ext'))
2510 sub_filename_final = subtitles_filename(
2511 self.prepare_filename(info_dict, 'subtitle'), sub_lang, sub_format, info_dict.get('ext'))
2512 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(sub_filename)):
2513 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
2514 sub_info['filepath'] = sub_filename
2515 files_to_move[sub_filename] = sub_filename_final
2516 else:
2517 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
2518 if sub_info.get('data') is not None:
2519 try:
2520 # Use newline='' to prevent conversion of newline characters
2521 # See https://github.com/ytdl-org/youtube-dl/issues/10268
2522 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
2523 subfile.write(sub_info['data'])
2524 sub_info['filepath'] = sub_filename
2525 files_to_move[sub_filename] = sub_filename_final
2526 except (OSError, IOError):
2527 self.report_error('Cannot write subtitles file ' + sub_filename)
2528 return
2529 else:
2530 try:
2531 self.dl(sub_filename, sub_info.copy(), subtitle=True)
2532 sub_info['filepath'] = sub_filename
2533 files_to_move[sub_filename] = sub_filename_final
2534 except (ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
2535 self.report_warning('Unable to download subtitle for "%s": %s' %
2536 (sub_lang, error_to_compat_str(err)))
2537 continue
2538
2539 if self.params.get('writeinfojson', False):
2540 infofn = self.prepare_filename(info_dict, 'infojson')
2541 if not self._ensure_dir_exists(encodeFilename(infofn)):
2542 return
2543 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
2544 self.to_screen('[info] Video metadata is already present')
2545 else:
2546 self.to_screen('[info] Writing video metadata as JSON to: ' + infofn)
2547 try:
2548 write_json_file(self.filter_requested_info(info_dict, self.params.get('clean_infojson', True)), infofn)
2549 except (OSError, IOError):
2550 self.report_error('Cannot write video metadata to JSON file ' + infofn)
2551 return
2552 info_dict['__infojson_filename'] = infofn
2553
2554 for thumb_ext in self._write_thumbnails(info_dict, temp_filename):
2555 thumb_filename_temp = replace_extension(temp_filename, thumb_ext, info_dict.get('ext'))
2556 thumb_filename = replace_extension(
2557 self.prepare_filename(info_dict, 'thumbnail'), thumb_ext, info_dict.get('ext'))
2558 files_to_move[thumb_filename_temp] = thumb_filename
2559
2560 # Write internet shortcut files
2561 url_link = webloc_link = desktop_link = False
2562 if self.params.get('writelink', False):
2563 if sys.platform == "darwin": # macOS.
2564 webloc_link = True
2565 elif sys.platform.startswith("linux"):
2566 desktop_link = True
2567 else: # if sys.platform in ['win32', 'cygwin']:
2568 url_link = True
2569 if self.params.get('writeurllink', False):
2570 url_link = True
2571 if self.params.get('writewebloclink', False):
2572 webloc_link = True
2573 if self.params.get('writedesktoplink', False):
2574 desktop_link = True
2575
2576 if url_link or webloc_link or desktop_link:
2577 if 'webpage_url' not in info_dict:
2578 self.report_error('Cannot write internet shortcut file because the "webpage_url" field is missing in the media information')
2579 return
2580 ascii_url = iri_to_uri(info_dict['webpage_url'])
2581
2582 def _write_link_file(extension, template, newline, embed_filename):
2583 linkfn = replace_extension(full_filename, extension, info_dict.get('ext'))
2584 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2585 self.to_screen('[info] Internet shortcut is already present')
2586 else:
2587 try:
2588 self.to_screen('[info] Writing internet shortcut to: ' + linkfn)
2589 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', newline=newline) as linkfile:
2590 template_vars = {'url': ascii_url}
2591 if embed_filename:
2592 template_vars['filename'] = linkfn[:-(len(extension) + 1)]
2593 linkfile.write(template % template_vars)
2594 except (OSError, IOError):
2595 self.report_error('Cannot write internet shortcut ' + linkfn)
2596 return False
2597 return True
2598
2599 if url_link:
2600 if not _write_link_file('url', DOT_URL_LINK_TEMPLATE, '\r\n', embed_filename=False):
2601 return
2602 if webloc_link:
2603 if not _write_link_file('webloc', DOT_WEBLOC_LINK_TEMPLATE, '\n', embed_filename=False):
2604 return
2605 if desktop_link:
2606 if not _write_link_file('desktop', DOT_DESKTOP_LINK_TEMPLATE, '\n', embed_filename=True):
2607 return
2608
2609 try:
2610 info_dict, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2611 except PostProcessingError as err:
2612 self.report_error('Preprocessing: %s' % str(err))
2613 return
2614
2615 must_record_download_archive = False
2616 if self.params.get('skip_download', False):
2617 info_dict['filepath'] = temp_filename
2618 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2619 info_dict['__files_to_move'] = files_to_move
2620 info_dict = self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict)
2621 else:
2622 # Download
2623 try:
2624
2625 def existing_file(*filepaths):
2626 ext = info_dict.get('ext')
2627 final_ext = self.params.get('final_ext', ext)
2628 existing_files = []
2629 for file in orderedSet(filepaths):
2630 if final_ext != ext:
2631 converted = replace_extension(file, final_ext, ext)
2632 if os.path.exists(encodeFilename(converted)):
2633 existing_files.append(converted)
2634 if os.path.exists(encodeFilename(file)):
2635 existing_files.append(file)
2636
2637 if not existing_files or self.params.get('overwrites', False):
2638 for file in orderedSet(existing_files):
2639 self.report_file_delete(file)
2640 os.remove(encodeFilename(file))
2641 return None
2642
2643 self.report_file_already_downloaded(existing_files[0])
2644 info_dict['ext'] = os.path.splitext(existing_files[0])[1][1:]
2645 return existing_files[0]
2646
2647 success = True
2648 if info_dict.get('requested_formats') is not None:
2649
2650 def compatible_formats(formats):
2651 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2652 video_formats = [format for format in formats if format.get('vcodec') != 'none']
2653 audio_formats = [format for format in formats if format.get('acodec') != 'none']
2654 if len(video_formats) > 2 or len(audio_formats) > 2:
2655 return False
2656
2657 # Check extension
2658 exts = set(format.get('ext') for format in formats)
2659 COMPATIBLE_EXTS = (
2660 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
2661 set(('webm',)),
2662 )
2663 for ext_sets in COMPATIBLE_EXTS:
2664 if ext_sets.issuperset(exts):
2665 return True
2666 # TODO: Check acodec/vcodec
2667 return False
2668
2669 requested_formats = info_dict['requested_formats']
2670 old_ext = info_dict['ext']
2671 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
2672 info_dict['ext'] = 'mkv'
2673 self.report_warning(
2674 'Requested formats are incompatible for merge and will be merged into mkv.')
2675
2676 def correct_ext(filename):
2677 filename_real_ext = os.path.splitext(filename)[1][1:]
2678 filename_wo_ext = (
2679 os.path.splitext(filename)[0]
2680 if filename_real_ext == old_ext
2681 else filename)
2682 return '%s.%s' % (filename_wo_ext, info_dict['ext'])
2683
2684 # Ensure filename always has a correct extension for successful merge
2685 full_filename = correct_ext(full_filename)
2686 temp_filename = correct_ext(temp_filename)
2687 dl_filename = existing_file(full_filename, temp_filename)
2688 info_dict['__real_download'] = False
2689
2690 _protocols = set(determine_protocol(f) for f in requested_formats)
2691 if len(_protocols) == 1:
2692 info_dict['protocol'] = _protocols.pop()
2693 directly_mergable = (
2694 'no-direct-merge' not in self.params.get('compat_opts', [])
2695 and info_dict.get('protocol') is not None # All requested formats have same protocol
2696 and not self.params.get('allow_unplayable_formats')
2697 and get_suitable_downloader(info_dict, self.params).__name__ == 'FFmpegFD')
2698 if directly_mergable:
2699 info_dict['url'] = requested_formats[0]['url']
2700 # Treat it as a single download
2701 dl_filename = existing_file(full_filename, temp_filename)
2702 if dl_filename is None:
2703 success, real_download = self.dl(temp_filename, info_dict)
2704 info_dict['__real_download'] = real_download
2705 else:
2706 downloaded = []
2707 merger = FFmpegMergerPP(self)
2708 if self.params.get('allow_unplayable_formats'):
2709 self.report_warning(
2710 'You have requested merging of multiple formats '
2711 'while also allowing unplayable formats to be downloaded. '
2712 'The formats won\'t be merged to prevent data corruption.')
2713 elif not merger.available:
2714 self.report_warning(
2715 'You have requested merging of multiple formats but ffmpeg is not installed. '
2716 'The formats won\'t be merged.')
2717
2718 if dl_filename is None:
2719 for f in requested_formats:
2720 new_info = dict(info_dict)
2721 del new_info['requested_formats']
2722 new_info.update(f)
2723 fname = prepend_extension(
2724 self.prepare_filename(new_info, 'temp'),
2725 'f%s' % f['format_id'], new_info['ext'])
2726 if not self._ensure_dir_exists(fname):
2727 return
2728 downloaded.append(fname)
2729 partial_success, real_download = self.dl(fname, new_info)
2730 info_dict['__real_download'] = info_dict['__real_download'] or real_download
2731 success = success and partial_success
2732 if merger.available and not self.params.get('allow_unplayable_formats'):
2733 info_dict['__postprocessors'].append(merger)
2734 info_dict['__files_to_merge'] = downloaded
2735 # Even if there were no downloads, it is being merged only now
2736 info_dict['__real_download'] = True
2737 else:
2738 for file in downloaded:
2739 files_to_move[file] = None
2740 else:
2741 # Just a single file
2742 dl_filename = existing_file(full_filename, temp_filename)
2743 if dl_filename is None:
2744 success, real_download = self.dl(temp_filename, info_dict)
2745 info_dict['__real_download'] = real_download
2746
2747 dl_filename = dl_filename or temp_filename
2748 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2749
2750 except network_exceptions as err:
2751 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
2752 return
2753 except (OSError, IOError) as err:
2754 raise UnavailableVideoError(err)
2755 except (ContentTooShortError, ) as err:
2756 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
2757 return
2758
2759 if success and full_filename != '-':
2760
2761 def fixup():
2762 do_fixup = True
2763 fixup_policy = self.params.get('fixup')
2764 vid = info_dict['id']
2765
2766 if fixup_policy in ('ignore', 'never'):
2767 return
2768 elif fixup_policy == 'warn':
2769 do_fixup = False
2770 elif fixup_policy != 'force':
2771 assert fixup_policy in ('detect_or_warn', None)
2772 if not info_dict.get('__real_download'):
2773 do_fixup = False
2774
2775 def ffmpeg_fixup(cndn, msg, cls):
2776 if not cndn:
2777 return
2778 if not do_fixup:
2779 self.report_warning(f'{vid}: {msg}')
2780 return
2781 pp = cls(self)
2782 if pp.available:
2783 info_dict['__postprocessors'].append(pp)
2784 else:
2785 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
2786
2787 stretched_ratio = info_dict.get('stretched_ratio')
2788 ffmpeg_fixup(
2789 stretched_ratio not in (1, None),
2790 f'Non-uniform pixel ratio {stretched_ratio}',
2791 FFmpegFixupStretchedPP)
2792
2793 ffmpeg_fixup(
2794 (info_dict.get('requested_formats') is None
2795 and info_dict.get('container') == 'm4a_dash'
2796 and info_dict.get('ext') == 'm4a'),
2797 'writing DASH m4a. Only some players support this container',
2798 FFmpegFixupM4aPP)
2799
2800 downloader = (get_suitable_downloader(info_dict, self.params).__name__
2801 if 'protocol' in info_dict else None)
2802 ffmpeg_fixup(downloader == 'HlsFD', 'malformed AAC bitstream detected', FFmpegFixupM3u8PP)
2803 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'malformed timestamps detected', FFmpegFixupTimestampPP)
2804 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'malformed duration detected', FFmpegFixupDurationPP)
2805
2806 fixup()
2807 try:
2808 info_dict = self.post_process(dl_filename, info_dict, files_to_move)
2809 except PostProcessingError as err:
2810 self.report_error('Postprocessing: %s' % str(err))
2811 return
2812 try:
2813 for ph in self._post_hooks:
2814 ph(info_dict['filepath'])
2815 except Exception as err:
2816 self.report_error('post hooks: %s' % str(err))
2817 return
2818 must_record_download_archive = True
2819
2820 if must_record_download_archive or self.params.get('force_write_download_archive', False):
2821 self.record_download_archive(info_dict)
2822 max_downloads = self.params.get('max_downloads')
2823 if max_downloads is not None and self._num_downloads >= int(max_downloads):
2824 raise MaxDownloadsReached()
2825
2826 def download(self, url_list):
2827 """Download a given list of URLs."""
2828 outtmpl = self.outtmpl_dict['default']
2829 if (len(url_list) > 1
2830 and outtmpl != '-'
2831 and '%' not in outtmpl
2832 and self.params.get('max_downloads') != 1):
2833 raise SameFileError(outtmpl)
2834
2835 for url in url_list:
2836 try:
2837 # It also downloads the videos
2838 res = self.extract_info(
2839 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2840 except UnavailableVideoError:
2841 self.report_error('unable to download video')
2842 except MaxDownloadsReached:
2843 self.to_screen('[info] Maximum number of downloaded files reached')
2844 raise
2845 except ExistingVideoReached:
2846 self.to_screen('[info] Encountered a file that is already in the archive, stopping due to --break-on-existing')
2847 raise
2848 except RejectedVideoReached:
2849 self.to_screen('[info] Encountered a file that did not match filter, stopping due to --break-on-reject')
2850 raise
2851 else:
2852 if self.params.get('dump_single_json', False):
2853 self.post_extract(res)
2854 self.to_stdout(json.dumps(res, default=repr))
2855
2856 return self._download_retcode
2857
2858 def download_with_info_file(self, info_filename):
2859 with contextlib.closing(fileinput.FileInput(
2860 [info_filename], mode='r',
2861 openhook=fileinput.hook_encoded('utf-8'))) as f:
2862 # FileInput doesn't have a read method, we can't call json.load
2863 info = self.filter_requested_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
2864 try:
2865 self.process_ie_result(info, download=True)
2866 except (DownloadError, EntryNotInPlaylist, ThrottledDownload):
2867 webpage_url = info.get('webpage_url')
2868 if webpage_url is not None:
2869 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2870 return self.download([webpage_url])
2871 else:
2872 raise
2873 return self._download_retcode
2874
2875 @staticmethod
2876 def filter_requested_info(info_dict, actually_filter=True):
2877 remove_keys = ['__original_infodict'] # Always remove this since this may contain a copy of the entire dict
2878 keep_keys = ['_type'], # Always keep this to facilitate load-info-json
2879 if actually_filter:
2880 remove_keys += ('requested_formats', 'requested_subtitles', 'requested_entries', 'filepath', 'entries', 'original_url')
2881 empty_values = (None, {}, [], set(), tuple())
2882 reject = lambda k, v: k not in keep_keys and (
2883 k.startswith('_') or k in remove_keys or v in empty_values)
2884 else:
2885 info_dict['epoch'] = int(time.time())
2886 reject = lambda k, v: k in remove_keys
2887 filter_fn = lambda obj: (
2888 list(map(filter_fn, obj)) if isinstance(obj, (LazyList, list, tuple, set))
2889 else obj if not isinstance(obj, dict)
2890 else dict((k, filter_fn(v)) for k, v in obj.items() if not reject(k, v)))
2891 return filter_fn(info_dict)
2892
2893 def run_pp(self, pp, infodict):
2894 files_to_delete = []
2895 if '__files_to_move' not in infodict:
2896 infodict['__files_to_move'] = {}
2897 files_to_delete, infodict = pp.run(infodict)
2898 if not files_to_delete:
2899 return infodict
2900
2901 if self.params.get('keepvideo', False):
2902 for f in files_to_delete:
2903 infodict['__files_to_move'].setdefault(f, '')
2904 else:
2905 for old_filename in set(files_to_delete):
2906 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2907 try:
2908 os.remove(encodeFilename(old_filename))
2909 except (IOError, OSError):
2910 self.report_warning('Unable to remove downloaded original file')
2911 if old_filename in infodict['__files_to_move']:
2912 del infodict['__files_to_move'][old_filename]
2913 return infodict
2914
2915 @staticmethod
2916 def post_extract(info_dict):
2917 def actual_post_extract(info_dict):
2918 if info_dict.get('_type') in ('playlist', 'multi_video'):
2919 for video_dict in info_dict.get('entries', {}):
2920 actual_post_extract(video_dict or {})
2921 return
2922
2923 post_extractor = info_dict.get('__post_extractor') or (lambda: {})
2924 extra = post_extractor().items()
2925 info_dict.update(extra)
2926 info_dict.pop('__post_extractor', None)
2927
2928 original_infodict = info_dict.get('__original_infodict') or {}
2929 original_infodict.update(extra)
2930 original_infodict.pop('__post_extractor', None)
2931
2932 actual_post_extract(info_dict or {})
2933
2934 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
2935 info = dict(ie_info)
2936 info['__files_to_move'] = files_to_move or {}
2937 for pp in self._pps[key]:
2938 info = self.run_pp(pp, info)
2939 return info, info.pop('__files_to_move', None)
2940
2941 def post_process(self, filename, ie_info, files_to_move=None):
2942 """Run all the postprocessors on the given file."""
2943 info = dict(ie_info)
2944 info['filepath'] = filename
2945 info['__files_to_move'] = files_to_move or {}
2946
2947 for pp in ie_info.get('__postprocessors', []) + self._pps['post_process']:
2948 info = self.run_pp(pp, info)
2949 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
2950 del info['__files_to_move']
2951 for pp in self._pps['after_move']:
2952 info = self.run_pp(pp, info)
2953 return info
2954
2955 def _make_archive_id(self, info_dict):
2956 video_id = info_dict.get('id')
2957 if not video_id:
2958 return
2959 # Future-proof against any change in case
2960 # and backwards compatibility with prior versions
2961 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
2962 if extractor is None:
2963 url = str_or_none(info_dict.get('url'))
2964 if not url:
2965 return
2966 # Try to find matching extractor for the URL and take its ie_key
2967 for ie in self._ies:
2968 if ie.suitable(url):
2969 extractor = ie.ie_key()
2970 break
2971 else:
2972 return
2973 return '%s %s' % (extractor.lower(), video_id)
2974
2975 def in_download_archive(self, info_dict):
2976 fn = self.params.get('download_archive')
2977 if fn is None:
2978 return False
2979
2980 vid_id = self._make_archive_id(info_dict)
2981 if not vid_id:
2982 return False # Incomplete video information
2983
2984 return vid_id in self.archive
2985
2986 def record_download_archive(self, info_dict):
2987 fn = self.params.get('download_archive')
2988 if fn is None:
2989 return
2990 vid_id = self._make_archive_id(info_dict)
2991 assert vid_id
2992 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2993 archive_file.write(vid_id + '\n')
2994 self.archive.add(vid_id)
2995
2996 @staticmethod
2997 def format_resolution(format, default='unknown'):
2998 if format.get('vcodec') == 'none':
2999 if format.get('acodec') == 'none':
3000 return 'images'
3001 return 'audio only'
3002 if format.get('resolution') is not None:
3003 return format['resolution']
3004 if format.get('width') and format.get('height'):
3005 res = '%dx%d' % (format['width'], format['height'])
3006 elif format.get('height'):
3007 res = '%sp' % format['height']
3008 elif format.get('width'):
3009 res = '%dx?' % format['width']
3010 else:
3011 res = default
3012 return res
3013
3014 def _format_note(self, fdict):
3015 res = ''
3016 if fdict.get('ext') in ['f4f', 'f4m']:
3017 res += '(unsupported) '
3018 if fdict.get('language'):
3019 if res:
3020 res += ' '
3021 res += '[%s] ' % fdict['language']
3022 if fdict.get('format_note') is not None:
3023 res += fdict['format_note'] + ' '
3024 if fdict.get('tbr') is not None:
3025 res += '%4dk ' % fdict['tbr']
3026 if fdict.get('container') is not None:
3027 if res:
3028 res += ', '
3029 res += '%s container' % fdict['container']
3030 if (fdict.get('vcodec') is not None
3031 and fdict.get('vcodec') != 'none'):
3032 if res:
3033 res += ', '
3034 res += fdict['vcodec']
3035 if fdict.get('vbr') is not None:
3036 res += '@'
3037 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3038 res += 'video@'
3039 if fdict.get('vbr') is not None:
3040 res += '%4dk' % fdict['vbr']
3041 if fdict.get('fps') is not None:
3042 if res:
3043 res += ', '
3044 res += '%sfps' % fdict['fps']
3045 if fdict.get('acodec') is not None:
3046 if res:
3047 res += ', '
3048 if fdict['acodec'] == 'none':
3049 res += 'video only'
3050 else:
3051 res += '%-5s' % fdict['acodec']
3052 elif fdict.get('abr') is not None:
3053 if res:
3054 res += ', '
3055 res += 'audio'
3056 if fdict.get('abr') is not None:
3057 res += '@%3dk' % fdict['abr']
3058 if fdict.get('asr') is not None:
3059 res += ' (%5dHz)' % fdict['asr']
3060 if fdict.get('filesize') is not None:
3061 if res:
3062 res += ', '
3063 res += format_bytes(fdict['filesize'])
3064 elif fdict.get('filesize_approx') is not None:
3065 if res:
3066 res += ', '
3067 res += '~' + format_bytes(fdict['filesize_approx'])
3068 return res
3069
3070 def list_formats(self, info_dict):
3071 formats = info_dict.get('formats', [info_dict])
3072 new_format = (
3073 'list-formats' not in self.params.get('compat_opts', [])
3074 and self.params.get('listformats_table', True) is not False)
3075 if new_format:
3076 table = [
3077 [
3078 format_field(f, 'format_id'),
3079 format_field(f, 'ext'),
3080 self.format_resolution(f),
3081 format_field(f, 'fps', '%d'),
3082 '|',
3083 format_field(f, 'filesize', ' %s', func=format_bytes) + format_field(f, 'filesize_approx', '~%s', func=format_bytes),
3084 format_field(f, 'tbr', '%4dk'),
3085 shorten_protocol_name(f.get('protocol', '').replace("native", "n")),
3086 '|',
3087 format_field(f, 'vcodec', default='unknown').replace('none', ''),
3088 format_field(f, 'vbr', '%4dk'),
3089 format_field(f, 'acodec', default='unknown').replace('none', ''),
3090 format_field(f, 'abr', '%3dk'),
3091 format_field(f, 'asr', '%5dHz'),
3092 ', '.join(filter(None, (
3093 'UNSUPPORTED' if f.get('ext') in ('f4f', 'f4m') else '',
3094 format_field(f, 'language', '[%s]'),
3095 format_field(f, 'format_note'),
3096 format_field(f, 'container', ignore=(None, f.get('ext'))),
3097 ))),
3098 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3099 header_line = ['ID', 'EXT', 'RESOLUTION', 'FPS', '|', ' FILESIZE', ' TBR', 'PROTO',
3100 '|', 'VCODEC', ' VBR', 'ACODEC', ' ABR', ' ASR', 'MORE INFO']
3101 else:
3102 table = [
3103 [
3104 format_field(f, 'format_id'),
3105 format_field(f, 'ext'),
3106 self.format_resolution(f),
3107 self._format_note(f)]
3108 for f in formats
3109 if f.get('preference') is None or f['preference'] >= -1000]
3110 header_line = ['format code', 'extension', 'resolution', 'note']
3111
3112 self.to_screen(
3113 '[info] Available formats for %s:' % info_dict['id'])
3114 self.to_stdout(render_table(
3115 header_line, table, delim=new_format, extraGap=(0 if new_format else 1), hideEmpty=new_format))
3116
3117 def list_thumbnails(self, info_dict):
3118 thumbnails = list(info_dict.get('thumbnails'))
3119 if not thumbnails:
3120 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
3121 return
3122
3123 self.to_screen(
3124 '[info] Thumbnails for %s:' % info_dict['id'])
3125 self.to_stdout(render_table(
3126 ['ID', 'width', 'height', 'URL'],
3127 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
3128
3129 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3130 if not subtitles:
3131 self.to_screen('%s has no %s' % (video_id, name))
3132 return
3133 self.to_screen(
3134 'Available %s for %s:' % (name, video_id))
3135
3136 def _row(lang, formats):
3137 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3138 if len(set(names)) == 1:
3139 names = [] if names[0] == 'unknown' else names[:1]
3140 return [lang, ', '.join(names), ', '.join(exts)]
3141
3142 self.to_stdout(render_table(
3143 ['Language', 'Name', 'Formats'],
3144 [_row(lang, formats) for lang, formats in subtitles.items()],
3145 hideEmpty=True))
3146
3147 def urlopen(self, req):
3148 """ Start an HTTP download """
3149 if isinstance(req, compat_basestring):
3150 req = sanitized_Request(req)
3151 return self._opener.open(req, timeout=self._socket_timeout)
3152
3153 def print_debug_header(self):
3154 if not self.params.get('verbose'):
3155 return
3156
3157 if type('') is not compat_str:
3158 # Python 2.6 on SLES11 SP1 (https://github.com/ytdl-org/youtube-dl/issues/3326)
3159 self.report_warning(
3160 'Your Python is broken! Update to a newer and supported version')
3161
3162 stdout_encoding = getattr(
3163 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
3164 encoding_str = (
3165 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
3166 locale.getpreferredencoding(),
3167 sys.getfilesystemencoding(),
3168 stdout_encoding,
3169 self.get_encoding()))
3170 write_string(encoding_str, encoding=None)
3171
3172 source = (
3173 '(exe)' if hasattr(sys, 'frozen')
3174 else '(zip)' if isinstance(globals().get('__loader__'), zipimporter)
3175 else '(source)' if os.path.basename(sys.argv[0]) == '__main__.py'
3176 else '')
3177 self._write_string('[debug] yt-dlp version %s %s\n' % (__version__, source))
3178 if _LAZY_LOADER:
3179 self._write_string('[debug] Lazy loading extractors enabled\n')
3180 if _PLUGIN_CLASSES:
3181 self._write_string(
3182 '[debug] Plugin Extractors: %s\n' % [ie.ie_key() for ie in _PLUGIN_CLASSES])
3183 if self.params.get('compat_opts'):
3184 self._write_string(
3185 '[debug] Compatibility options: %s\n' % ', '.join(self.params.get('compat_opts')))
3186 try:
3187 sp = subprocess.Popen(
3188 ['git', 'rev-parse', '--short', 'HEAD'],
3189 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3190 cwd=os.path.dirname(os.path.abspath(__file__)))
3191 out, err = process_communicate_or_kill(sp)
3192 out = out.decode().strip()
3193 if re.match('[0-9a-f]+', out):
3194 self._write_string('[debug] Git HEAD: %s\n' % out)
3195 except Exception:
3196 try:
3197 sys.exc_clear()
3198 except Exception:
3199 pass
3200
3201 def python_implementation():
3202 impl_name = platform.python_implementation()
3203 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3204 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3205 return impl_name
3206
3207 self._write_string('[debug] Python version %s (%s %s) - %s\n' % (
3208 platform.python_version(),
3209 python_implementation(),
3210 platform.architecture()[0],
3211 platform_name()))
3212
3213 exe_versions = FFmpegPostProcessor.get_versions(self)
3214 exe_versions['rtmpdump'] = rtmpdump_version()
3215 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3216 exe_str = ', '.join(
3217 '%s %s' % (exe, v)
3218 for exe, v in sorted(exe_versions.items())
3219 if v
3220 )
3221 if not exe_str:
3222 exe_str = 'none'
3223 self._write_string('[debug] exe versions: %s\n' % exe_str)
3224
3225 proxy_map = {}
3226 for handler in self._opener.handlers:
3227 if hasattr(handler, 'proxies'):
3228 proxy_map.update(handler.proxies)
3229 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
3230
3231 if self.params.get('call_home', False):
3232 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3233 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
3234 return
3235 latest_version = self.urlopen(
3236 'https://yt-dl.org/latest/version').read().decode('utf-8')
3237 if version_tuple(latest_version) > version_tuple(__version__):
3238 self.report_warning(
3239 'You are using an outdated version (newest version: %s)! '
3240 'See https://yt-dl.org/update if you need help updating.' %
3241 latest_version)
3242
3243 def _setup_opener(self):
3244 timeout_val = self.params.get('socket_timeout')
3245 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
3246
3247 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3248 opts_cookiefile = self.params.get('cookiefile')
3249 opts_proxy = self.params.get('proxy')
3250
3251 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3252
3253 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3254 if opts_proxy is not None:
3255 if opts_proxy == '':
3256 proxies = {}
3257 else:
3258 proxies = {'http': opts_proxy, 'https': opts_proxy}
3259 else:
3260 proxies = compat_urllib_request.getproxies()
3261 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3262 if 'http' in proxies and 'https' not in proxies:
3263 proxies['https'] = proxies['http']
3264 proxy_handler = PerRequestProxyHandler(proxies)
3265
3266 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3267 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3268 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3269 redirect_handler = YoutubeDLRedirectHandler()
3270 data_handler = compat_urllib_request_DataHandler()
3271
3272 # When passing our own FileHandler instance, build_opener won't add the
3273 # default FileHandler and allows us to disable the file protocol, which
3274 # can be used for malicious purposes (see
3275 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3276 file_handler = compat_urllib_request.FileHandler()
3277
3278 def file_open(*args, **kwargs):
3279 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3280 file_handler.file_open = file_open
3281
3282 opener = compat_urllib_request.build_opener(
3283 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3284
3285 # Delete the default user-agent header, which would otherwise apply in
3286 # cases where our custom HTTP handler doesn't come into play
3287 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3288 opener.addheaders = []
3289 self._opener = opener
3290
3291 def encode(self, s):
3292 if isinstance(s, bytes):
3293 return s # Already encoded
3294
3295 try:
3296 return s.encode(self.get_encoding())
3297 except UnicodeEncodeError as err:
3298 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3299 raise
3300
3301 def get_encoding(self):
3302 encoding = self.params.get('encoding')
3303 if encoding is None:
3304 encoding = preferredencoding()
3305 return encoding
3306
3307 def _write_thumbnails(self, info_dict, filename): # return the extensions
3308 write_all = self.params.get('write_all_thumbnails', False)
3309 thumbnails = []
3310 if write_all or self.params.get('writethumbnail', False):
3311 thumbnails = info_dict.get('thumbnails') or []
3312 multiple = write_all and len(thumbnails) > 1
3313
3314 ret = []
3315 for t in thumbnails[::-1]:
3316 thumb_ext = determine_ext(t['url'], 'jpg')
3317 suffix = '%s.' % t['id'] if multiple else ''
3318 thumb_display_id = '%s ' % t['id'] if multiple else ''
3319 thumb_filename = replace_extension(filename, suffix + thumb_ext, info_dict.get('ext'))
3320
3321 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(thumb_filename)):
3322 ret.append(suffix + thumb_ext)
3323 t['filepath'] = thumb_filename
3324 self.to_screen('[%s] %s: Thumbnail %sis already present' %
3325 (info_dict['extractor'], info_dict['id'], thumb_display_id))
3326 else:
3327 self.to_screen('[%s] %s: Downloading thumbnail %s ...' %
3328 (info_dict['extractor'], info_dict['id'], thumb_display_id))
3329 try:
3330 uf = self.urlopen(t['url'])
3331 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3332 shutil.copyfileobj(uf, thumbf)
3333 ret.append(suffix + thumb_ext)
3334 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
3335 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
3336 t['filepath'] = thumb_filename
3337 except network_exceptions as err:
3338 self.report_warning('Unable to download thumbnail "%s": %s' %
3339 (t['url'], error_to_compat_str(err)))
3340 if ret and not write_all:
3341 break
3342 return ret