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