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