]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
11add73ade0af662e3d42f3c7a51e5e0e828cbf8
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 #!/usr/bin/env python3
2 # coding: utf-8
3
4 from __future__ import absolute_import, unicode_literals
5
6 import collections
7 import contextlib
8 import copy
9 import datetime
10 import errno
11 import fileinput
12 import io
13 import itertools
14 import json
15 import locale
16 import operator
17 import os
18 import platform
19 import re
20 import shutil
21 import subprocess
22 import sys
23 import tempfile
24 import time
25 import tokenize
26 import traceback
27 import random
28
29 from string import ascii_letters
30 from zipimport import zipimporter
31
32 from .compat import (
33 compat_basestring,
34 compat_get_terminal_size,
35 compat_kwargs,
36 compat_numeric_types,
37 compat_os_name,
38 compat_shlex_quote,
39 compat_str,
40 compat_tokenize_tokenize,
41 compat_urllib_error,
42 compat_urllib_request,
43 compat_urllib_request_DataHandler,
44 )
45 from .cookies import load_cookies
46 from .utils import (
47 age_restricted,
48 args_to_str,
49 ContentTooShortError,
50 date_from_str,
51 DateRange,
52 DEFAULT_OUTTMPL,
53 determine_ext,
54 determine_protocol,
55 DOT_DESKTOP_LINK_TEMPLATE,
56 DOT_URL_LINK_TEMPLATE,
57 DOT_WEBLOC_LINK_TEMPLATE,
58 DownloadError,
59 encode_compat_str,
60 encodeFilename,
61 EntryNotInPlaylist,
62 error_to_compat_str,
63 ExistingVideoReached,
64 expand_path,
65 ExtractorError,
66 float_or_none,
67 format_bytes,
68 format_field,
69 STR_FORMAT_RE_TMPL,
70 STR_FORMAT_TYPES,
71 formatSeconds,
72 GeoRestrictedError,
73 HEADRequest,
74 int_or_none,
75 iri_to_uri,
76 ISO3166Utils,
77 LazyList,
78 locked_file,
79 make_dir,
80 make_HTTPS_handler,
81 MaxDownloadsReached,
82 network_exceptions,
83 orderedSet,
84 OUTTMPL_TYPES,
85 PagedList,
86 parse_filesize,
87 PerRequestProxyHandler,
88 platform_name,
89 PostProcessingError,
90 preferredencoding,
91 prepend_extension,
92 process_communicate_or_kill,
93 register_socks_protocols,
94 RejectedVideoReached,
95 render_table,
96 replace_extension,
97 SameFileError,
98 sanitize_filename,
99 sanitize_path,
100 sanitize_url,
101 sanitized_Request,
102 std_headers,
103 str_or_none,
104 strftime_or_none,
105 subtitles_filename,
106 ThrottledDownload,
107 to_high_limit_path,
108 traverse_obj,
109 try_get,
110 UnavailableVideoError,
111 url_basename,
112 variadic,
113 version_tuple,
114 write_json_file,
115 write_string,
116 YoutubeDLCookieProcessor,
117 YoutubeDLHandler,
118 YoutubeDLRedirectHandler,
119 )
120 from .cache import Cache
121 from .extractor import (
122 gen_extractor_classes,
123 get_info_extractor,
124 _LAZY_LOADER,
125 _PLUGIN_CLASSES
126 )
127 from .extractor.openload import PhantomJSwrapper
128 from .downloader import (
129 FFmpegFD,
130 get_suitable_downloader,
131 shorten_protocol_name
132 )
133 from .downloader.rtmp import rtmpdump_version
134 from .postprocessor import (
135 get_postprocessor,
136 FFmpegFixupDurationPP,
137 FFmpegFixupM3u8PP,
138 FFmpegFixupM4aPP,
139 FFmpegFixupStretchedPP,
140 FFmpegFixupTimestampPP,
141 FFmpegMergerPP,
142 FFmpegPostProcessor,
143 MoveFilesAfterDownloadPP,
144 )
145 from .version import __version__
146
147 if compat_os_name == 'nt':
148 import ctypes
149
150
151 class YoutubeDL(object):
152 """YoutubeDL class.
153
154 YoutubeDL objects are the ones responsible of downloading the
155 actual video file and writing it to disk if the user has requested
156 it, among some other tasks. In most cases there should be one per
157 program. As, given a video URL, the downloader doesn't know how to
158 extract all the needed information, task that InfoExtractors do, it
159 has to pass the URL to one of them.
160
161 For this, YoutubeDL objects have a method that allows
162 InfoExtractors to be registered in a given order. When it is passed
163 a URL, the YoutubeDL object handles it to the first InfoExtractor it
164 finds that reports being able to handle it. The InfoExtractor extracts
165 all the information about the video or videos the URL refers to, and
166 YoutubeDL process the extracted information, possibly using a File
167 Downloader to download the video.
168
169 YoutubeDL objects accept a lot of parameters. In order not to saturate
170 the object constructor with arguments, it receives a dictionary of
171 options instead. These options are available through the params
172 attribute for the InfoExtractors to use. The YoutubeDL also
173 registers itself as the downloader in charge for the InfoExtractors
174 that are added to it, so this is a "mutual registration".
175
176 Available options:
177
178 username: Username for authentication purposes.
179 password: Password for authentication purposes.
180 videopassword: Password for accessing a video.
181 ap_mso: Adobe Pass multiple-system operator identifier.
182 ap_username: Multiple-system operator account username.
183 ap_password: Multiple-system operator account password.
184 usenetrc: Use netrc for authentication instead.
185 verbose: Print additional info to stdout.
186 quiet: Do not print messages to stdout.
187 no_warnings: Do not print out anything for warnings.
188 forceprint: A list of templates to force print
189 forceurl: Force printing final URL. (Deprecated)
190 forcetitle: Force printing title. (Deprecated)
191 forceid: Force printing ID. (Deprecated)
192 forcethumbnail: Force printing thumbnail URL. (Deprecated)
193 forcedescription: Force printing description. (Deprecated)
194 forcefilename: Force printing final filename. (Deprecated)
195 forceduration: Force printing duration. (Deprecated)
196 forcejson: Force printing info_dict as JSON.
197 dump_single_json: Force printing the info_dict of the whole playlist
198 (or video) as a single JSON line.
199 force_write_download_archive: Force writing download archive regardless
200 of 'skip_download' or 'simulate'.
201 simulate: Do not download the video files.
202 format: Video format code. see "FORMAT SELECTION" for more details.
203 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
204 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
205 extracting metadata even if the video is not actually
206 available for download (experimental)
207 format_sort: How to sort the video formats. see "Sorting Formats"
208 for more details.
209 format_sort_force: Force the given format_sort. see "Sorting Formats"
210 for more details.
211 allow_multiple_video_streams: Allow multiple video streams to be merged
212 into a single file
213 allow_multiple_audio_streams: Allow multiple audio streams to be merged
214 into a single file
215 check_formats Whether to test if the formats are downloadable.
216 Can be True (check all), False (check none)
217 or None (check only if requested by extractor)
218 paths: Dictionary of output paths. The allowed keys are 'home'
219 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
220 outtmpl: Dictionary of templates for output names. Allowed keys
221 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
222 A string a also accepted for backward compatibility
223 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
224 restrictfilenames: Do not allow "&" and spaces in file names
225 trim_file_name: Limit length of filename (extension excluded)
226 windowsfilenames: Force the filenames to be windows compatible
227 ignoreerrors: Do not stop on download errors
228 (Default True when running yt-dlp,
229 but False when directly accessing YoutubeDL class)
230 skip_playlist_after_errors: Number of allowed failures until the rest of
231 the playlist is skipped
232 force_generic_extractor: Force downloader to use the generic extractor
233 overwrites: Overwrite all video and metadata files if True,
234 overwrite only non-video files if None
235 and don't overwrite any file if False
236 playliststart: Playlist item to start at.
237 playlistend: Playlist item to end at.
238 playlist_items: Specific indices of playlist to download.
239 playlistreverse: Download playlist items in reverse order.
240 playlistrandom: Download playlist items in random order.
241 matchtitle: Download only matching titles.
242 rejecttitle: Reject downloads for matching titles.
243 logger: Log messages to a logging.Logger instance.
244 logtostderr: Log messages to stderr instead of stdout.
245 writedescription: Write the video description to a .description file
246 writeinfojson: Write the video description to a .info.json file
247 clean_infojson: Remove private fields from the infojson
248 writecomments: Extract video comments. This will not be written to disk
249 unless writeinfojson is also given
250 writeannotations: Write the video annotations to a .annotations.xml file
251 writethumbnail: Write the thumbnail image to a file
252 allow_playlist_files: Whether to write playlists' description, infojson etc
253 also to disk when using the 'write*' options
254 write_all_thumbnails: Write all thumbnail formats to files
255 writelink: Write an internet shortcut file, depending on the
256 current platform (.url/.webloc/.desktop)
257 writeurllink: Write a Windows internet shortcut file (.url)
258 writewebloclink: Write a macOS internet shortcut file (.webloc)
259 writedesktoplink: Write a Linux internet shortcut file (.desktop)
260 writesubtitles: Write the video subtitles to a file
261 writeautomaticsub: Write the automatically generated subtitles to a file
262 allsubtitles: Deprecated - Use subtitleslangs = ['all']
263 Downloads all the subtitles of the video
264 (requires writesubtitles or writeautomaticsub)
265 listsubtitles: Lists all available subtitles for the video
266 subtitlesformat: The format code for subtitles
267 subtitleslangs: List of languages of the subtitles to download (can be regex).
268 The list may contain "all" to refer to all the available
269 subtitles. The language can be prefixed with a "-" to
270 exclude it from the requested languages. Eg: ['all', '-live_chat']
271 keepvideo: Keep the video file after post-processing
272 daterange: A DateRange object, download only if the upload_date is in the range.
273 skip_download: Skip the actual download of the video file
274 cachedir: Location of the cache files in the filesystem.
275 False to disable filesystem cache.
276 noplaylist: Download single video instead of a playlist if in doubt.
277 age_limit: An integer representing the user's age in years.
278 Unsuitable videos for the given age are skipped.
279 min_views: An integer representing the minimum view count the video
280 must have in order to not be skipped.
281 Videos without view count information are always
282 downloaded. None for no limit.
283 max_views: An integer representing the maximum view count.
284 Videos that are more popular than that are not
285 downloaded.
286 Videos without view count information are always
287 downloaded. None for no limit.
288 download_archive: File name of a file where all downloads are recorded.
289 Videos already present in the file are not downloaded
290 again.
291 break_on_existing: Stop the download process after attempting to download a
292 file that is in the archive.
293 break_on_reject: Stop the download process when encountering a video that
294 has been filtered out.
295 cookiefile: File name where cookies should be read from and dumped to
296 cookiesfrombrowser: A tuple containing the name of the browser and the profile
297 name/path from where cookies are loaded.
298 Eg: ('chrome', ) or (vivaldi, 'default')
299 nocheckcertificate:Do not verify SSL certificates
300 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
301 At the moment, this is only supported by YouTube.
302 proxy: URL of the proxy server to use
303 geo_verification_proxy: URL of the proxy to use for IP address verification
304 on geo-restricted sites.
305 socket_timeout: Time to wait for unresponsive hosts, in seconds
306 bidi_workaround: Work around buggy terminals without bidirectional text
307 support, using fridibi
308 debug_printtraffic:Print out sent and received HTTP traffic
309 include_ads: Download ads as well
310 default_search: Prepend this string if an input url is not valid.
311 'auto' for elaborate guessing
312 encoding: Use this encoding instead of the system-specified.
313 extract_flat: Do not resolve URLs, return the immediate result.
314 Pass in 'in_playlist' to only show this behavior for
315 playlist items.
316 postprocessors: A list of dictionaries, each with an entry
317 * key: The name of the postprocessor. See
318 yt_dlp/postprocessor/__init__.py for a list.
319 * when: When to run the postprocessor. Can be one of
320 pre_process|before_dl|post_process|after_move.
321 Assumed to be 'post_process' if not given
322 post_hooks: A list of functions that get called as the final step
323 for each video file, after all postprocessors have been
324 called. The filename will be passed as the only argument.
325 progress_hooks: A list of functions that get called on download
326 progress, with a dictionary with the entries
327 * status: One of "downloading", "error", or "finished".
328 Check this first and ignore unknown values.
329 * info_dict: The extracted info_dict
330
331 If status is one of "downloading", or "finished", the
332 following properties may also be present:
333 * filename: The final filename (always present)
334 * tmpfilename: The filename we're currently writing to
335 * downloaded_bytes: Bytes on disk
336 * total_bytes: Size of the whole file, None if unknown
337 * total_bytes_estimate: Guess of the eventual file size,
338 None if unavailable.
339 * elapsed: The number of seconds since download started.
340 * eta: The estimated time in seconds, None if unknown
341 * speed: The download speed in bytes/second, None if
342 unknown
343 * fragment_index: The counter of the currently
344 downloaded video fragment.
345 * fragment_count: The number of fragments (= individual
346 files that will be merged)
347
348 Progress hooks are guaranteed to be called at least once
349 (with status "finished") if the download is successful.
350 merge_output_format: Extension to use when merging formats.
351 final_ext: Expected final extension; used to detect when the file was
352 already downloaded and converted. "merge_output_format" is
353 replaced by this extension when given
354 fixup: Automatically correct known faults of the file.
355 One of:
356 - "never": do nothing
357 - "warn": only emit a warning
358 - "detect_or_warn": check whether we can do anything
359 about it, warn otherwise (default)
360 source_address: Client-side IP address to bind to.
361 call_home: Boolean, true iff we are allowed to contact the
362 yt-dlp servers for debugging. (BROKEN)
363 sleep_interval_requests: Number of seconds to sleep between requests
364 during extraction
365 sleep_interval: Number of seconds to sleep before each download when
366 used alone or a lower bound of a range for randomized
367 sleep before each download (minimum possible number
368 of seconds to sleep) when used along with
369 max_sleep_interval.
370 max_sleep_interval:Upper bound of a range for randomized sleep before each
371 download (maximum possible number of seconds to sleep).
372 Must only be used along with sleep_interval.
373 Actual sleep time will be a random float from range
374 [sleep_interval; max_sleep_interval].
375 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
376 listformats: Print an overview of available video formats and exit.
377 list_thumbnails: Print a table of all thumbnails and exit.
378 match_filter: A function that gets called with the info_dict of
379 every video.
380 If it returns a message, the video is ignored.
381 If it returns None, the video is downloaded.
382 match_filter_func in utils.py is one example for this.
383 no_color: Do not emit color codes in output.
384 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
385 HTTP header
386 geo_bypass_country:
387 Two-letter ISO 3166-2 country code that will be used for
388 explicit geographic restriction bypassing via faking
389 X-Forwarded-For HTTP header
390 geo_bypass_ip_block:
391 IP range in CIDR notation that will be used similarly to
392 geo_bypass_country
393
394 The following options determine which downloader is picked:
395 external_downloader: A dictionary of protocol keys and the executable of the
396 external downloader to use for it. The allowed protocols
397 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
398 Set the value to 'native' to use the native downloader
399 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
400 or {'m3u8': 'ffmpeg'} instead.
401 Use the native HLS downloader instead of ffmpeg/avconv
402 if True, otherwise use ffmpeg/avconv if False, otherwise
403 use downloader suggested by extractor if None.
404 compat_opts: Compatibility options. See "Differences in default behavior".
405 The following options do not work when used through the API:
406 filename, abort-on-error, multistreams, no-live-chat,
407 no-clean-infojson, no-playlist-metafiles.
408 Refer __init__.py for their implementation
409
410 The following parameters are not used by YoutubeDL itself, they are used by
411 the downloader (see yt_dlp/downloader/common.py):
412 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
413 max_filesize, test, noresizebuffer, retries, continuedl, noprogress, consoletitle,
414 xattr_set_filesize, external_downloader_args, hls_use_mpegts, http_chunk_size.
415
416 The following options are used by the post processors:
417 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
418 otherwise prefer ffmpeg. (avconv support is deprecated)
419 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
420 to the binary or its containing directory.
421 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
422 and a list of additional command-line arguments for the
423 postprocessor/executable. The dict can also have "PP+EXE" keys
424 which are used when the given exe is used by the given PP.
425 Use 'default' as the name for arguments to passed to all PP
426
427 The following options are used by the extractors:
428 extractor_retries: Number of times to retry for known errors
429 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
430 hls_split_discontinuity: Split HLS playlists to different formats at
431 discontinuities such as ad breaks (default: False)
432 extractor_args: A dictionary of arguments to be passed to the extractors.
433 See "EXTRACTOR ARGUMENTS" for details.
434 Eg: {'youtube': {'skip': ['dash', 'hls']}}
435 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
436 If True (default), DASH manifests and related
437 data will be downloaded and processed by extractor.
438 You can reduce network I/O by disabling it if you don't
439 care about DASH. (only for youtube)
440 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
441 If True (default), HLS manifests and related
442 data will be downloaded and processed by extractor.
443 You can reduce network I/O by disabling it if you don't
444 care about HLS. (only for youtube)
445 """
446
447 _NUMERIC_FIELDS = set((
448 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
449 'timestamp', 'upload_year', 'upload_month', 'upload_day',
450 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
451 'average_rating', 'comment_count', 'age_limit',
452 'start_time', 'end_time',
453 'chapter_number', 'season_number', 'episode_number',
454 'track_number', 'disc_number', 'release_year',
455 'playlist_index',
456 ))
457
458 params = None
459 _ies = []
460 _pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
461 _printed_messages = set()
462 _first_webpage_request = True
463 _download_retcode = None
464 _num_downloads = None
465 _playlist_level = 0
466 _playlist_urls = set()
467 _screen_file = None
468
469 def __init__(self, params=None, auto_init=True):
470 """Create a FileDownloader object with the given options."""
471 if params is None:
472 params = {}
473 self._ies = []
474 self._ies_instances = {}
475 self._pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
476 self._printed_messages = set()
477 self._first_webpage_request = True
478 self._post_hooks = []
479 self._progress_hooks = []
480 self._download_retcode = 0
481 self._num_downloads = 0
482 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
483 self._err_file = sys.stderr
484 self.params = {
485 # Default parameters
486 'nocheckcertificate': False,
487 }
488 self.params.update(params)
489 self.cache = Cache(self)
490
491 if sys.version_info < (3, 6):
492 self.report_warning(
493 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
494
495 def check_deprecated(param, option, suggestion):
496 if self.params.get(param) is not None:
497 self.report_warning('%s is deprecated. Use %s instead' % (option, suggestion))
498 return True
499 return False
500
501 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
502 if self.params.get('geo_verification_proxy') is None:
503 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
504
505 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
506 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
507 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
508
509 for msg in self.params.get('warnings', []):
510 self.report_warning(msg)
511
512 if self.params.get('final_ext'):
513 if self.params.get('merge_output_format'):
514 self.report_warning('--merge-output-format will be ignored since --remux-video or --recode-video is given')
515 self.params['merge_output_format'] = self.params['final_ext']
516
517 if 'overwrites' in self.params and self.params['overwrites'] is None:
518 del self.params['overwrites']
519
520 if params.get('bidi_workaround', False):
521 try:
522 import pty
523 master, slave = pty.openpty()
524 width = compat_get_terminal_size().columns
525 if width is None:
526 width_args = []
527 else:
528 width_args = ['-w', str(width)]
529 sp_kwargs = dict(
530 stdin=subprocess.PIPE,
531 stdout=slave,
532 stderr=self._err_file)
533 try:
534 self._output_process = subprocess.Popen(
535 ['bidiv'] + width_args, **sp_kwargs
536 )
537 except OSError:
538 self._output_process = subprocess.Popen(
539 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
540 self._output_channel = os.fdopen(master, 'rb')
541 except OSError as ose:
542 if ose.errno == errno.ENOENT:
543 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.')
544 else:
545 raise
546
547 if (sys.platform != 'win32'
548 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
549 and not params.get('restrictfilenames', False)):
550 # Unicode filesystem API will throw errors (#1474, #13027)
551 self.report_warning(
552 'Assuming --restrict-filenames since file system encoding '
553 'cannot encode all characters. '
554 'Set the LC_ALL environment variable to fix this.')
555 self.params['restrictfilenames'] = True
556
557 self.outtmpl_dict = self.parse_outtmpl()
558
559 # Creating format selector here allows us to catch syntax errors before the extraction
560 self.format_selector = (
561 None if self.params.get('format') is None
562 else self.build_format_selector(self.params['format']))
563
564 self._setup_opener()
565
566 """Preload the archive, if any is specified"""
567 def preload_download_archive(fn):
568 if fn is None:
569 return False
570 self.write_debug('Loading archive file %r\n' % fn)
571 try:
572 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
573 for line in archive_file:
574 self.archive.add(line.strip())
575 except IOError as ioe:
576 if ioe.errno != errno.ENOENT:
577 raise
578 return False
579 return True
580
581 self.archive = set()
582 preload_download_archive(self.params.get('download_archive'))
583
584 if auto_init:
585 self.print_debug_header()
586 self.add_default_info_extractors()
587
588 for pp_def_raw in self.params.get('postprocessors', []):
589 pp_def = dict(pp_def_raw)
590 when = pp_def.pop('when', 'post_process')
591 pp_class = get_postprocessor(pp_def.pop('key'))
592 pp = pp_class(self, **compat_kwargs(pp_def))
593 self.add_post_processor(pp, when=when)
594
595 for ph in self.params.get('post_hooks', []):
596 self.add_post_hook(ph)
597
598 for ph in self.params.get('progress_hooks', []):
599 self.add_progress_hook(ph)
600
601 register_socks_protocols()
602
603 def warn_if_short_id(self, argv):
604 # short YouTube ID starting with dash?
605 idxs = [
606 i for i, a in enumerate(argv)
607 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
608 if idxs:
609 correct_argv = (
610 ['yt-dlp']
611 + [a for i, a in enumerate(argv) if i not in idxs]
612 + ['--'] + [argv[i] for i in idxs]
613 )
614 self.report_warning(
615 'Long argument string detected. '
616 'Use -- to separate parameters and URLs, like this:\n%s\n' %
617 args_to_str(correct_argv))
618
619 def add_info_extractor(self, ie):
620 """Add an InfoExtractor object to the end of the list."""
621 self._ies.append(ie)
622 if not isinstance(ie, type):
623 self._ies_instances[ie.ie_key()] = ie
624 ie.set_downloader(self)
625
626 def get_info_extractor(self, ie_key):
627 """
628 Get an instance of an IE with name ie_key, it will try to get one from
629 the _ies list, if there's no instance it will create a new one and add
630 it to the extractor list.
631 """
632 ie = self._ies_instances.get(ie_key)
633 if ie is None:
634 ie = get_info_extractor(ie_key)()
635 self.add_info_extractor(ie)
636 return ie
637
638 def add_default_info_extractors(self):
639 """
640 Add the InfoExtractors returned by gen_extractors to the end of the list
641 """
642 for ie in gen_extractor_classes():
643 self.add_info_extractor(ie)
644
645 def add_post_processor(self, pp, when='post_process'):
646 """Add a PostProcessor object to the end of the chain."""
647 self._pps[when].append(pp)
648 pp.set_downloader(self)
649
650 def add_post_hook(self, ph):
651 """Add the post hook"""
652 self._post_hooks.append(ph)
653
654 def add_progress_hook(self, ph):
655 """Add the progress hook (currently only for the file downloader)"""
656 self._progress_hooks.append(ph)
657
658 def _bidi_workaround(self, message):
659 if not hasattr(self, '_output_channel'):
660 return message
661
662 assert hasattr(self, '_output_process')
663 assert isinstance(message, compat_str)
664 line_count = message.count('\n') + 1
665 self._output_process.stdin.write((message + '\n').encode('utf-8'))
666 self._output_process.stdin.flush()
667 res = ''.join(self._output_channel.readline().decode('utf-8')
668 for _ in range(line_count))
669 return res[:-len('\n')]
670
671 def _write_string(self, message, out=None, only_once=False):
672 if only_once:
673 if message in self._printed_messages:
674 return
675 self._printed_messages.add(message)
676 write_string(message, out=out, encoding=self.params.get('encoding'))
677
678 def to_stdout(self, message, skip_eol=False, quiet=False):
679 """Print message to stdout"""
680 if self.params.get('logger'):
681 self.params['logger'].debug(message)
682 elif not quiet or self.params.get('verbose'):
683 self._write_string(
684 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
685 self._err_file if quiet else self._screen_file)
686
687 def to_stderr(self, message, only_once=False):
688 """Print message to stderr"""
689 assert isinstance(message, compat_str)
690 if self.params.get('logger'):
691 self.params['logger'].error(message)
692 else:
693 self._write_string('%s\n' % self._bidi_workaround(message), self._err_file, only_once=only_once)
694
695 def to_console_title(self, message):
696 if not self.params.get('consoletitle', False):
697 return
698 if compat_os_name == 'nt':
699 if ctypes.windll.kernel32.GetConsoleWindow():
700 # c_wchar_p() might not be necessary if `message` is
701 # already of type unicode()
702 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
703 elif 'TERM' in os.environ:
704 self._write_string('\033]0;%s\007' % message, self._screen_file)
705
706 def save_console_title(self):
707 if not self.params.get('consoletitle', False):
708 return
709 if self.params.get('simulate', False):
710 return
711 if compat_os_name != 'nt' and 'TERM' in os.environ:
712 # Save the title on stack
713 self._write_string('\033[22;0t', self._screen_file)
714
715 def restore_console_title(self):
716 if not self.params.get('consoletitle', False):
717 return
718 if self.params.get('simulate', False):
719 return
720 if compat_os_name != 'nt' and 'TERM' in os.environ:
721 # Restore the title from stack
722 self._write_string('\033[23;0t', self._screen_file)
723
724 def __enter__(self):
725 self.save_console_title()
726 return self
727
728 def __exit__(self, *args):
729 self.restore_console_title()
730
731 if self.params.get('cookiefile') is not None:
732 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
733
734 def trouble(self, message=None, tb=None):
735 """Determine action to take when a download problem appears.
736
737 Depending on if the downloader has been configured to ignore
738 download errors or not, this method may throw an exception or
739 not when errors are found, after printing the message.
740
741 tb, if given, is additional traceback information.
742 """
743 if message is not None:
744 self.to_stderr(message)
745 if self.params.get('verbose'):
746 if tb is None:
747 if sys.exc_info()[0]: # if .trouble has been called from an except block
748 tb = ''
749 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
750 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
751 tb += encode_compat_str(traceback.format_exc())
752 else:
753 tb_data = traceback.format_list(traceback.extract_stack())
754 tb = ''.join(tb_data)
755 if tb:
756 self.to_stderr(tb)
757 if not self.params.get('ignoreerrors', False):
758 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
759 exc_info = sys.exc_info()[1].exc_info
760 else:
761 exc_info = sys.exc_info()
762 raise DownloadError(message, exc_info)
763 self._download_retcode = 1
764
765 def to_screen(self, message, skip_eol=False):
766 """Print message to stdout if not in quiet mode"""
767 self.to_stdout(
768 message, skip_eol, quiet=self.params.get('quiet', False))
769
770 def report_warning(self, message, only_once=False):
771 '''
772 Print the message to stderr, it will be prefixed with 'WARNING:'
773 If stderr is a tty file the 'WARNING:' will be colored
774 '''
775 if self.params.get('logger') is not None:
776 self.params['logger'].warning(message)
777 else:
778 if self.params.get('no_warnings'):
779 return
780 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
781 _msg_header = '\033[0;33mWARNING:\033[0m'
782 else:
783 _msg_header = 'WARNING:'
784 warning_message = '%s %s' % (_msg_header, message)
785 self.to_stderr(warning_message, only_once)
786
787 def report_error(self, message, tb=None):
788 '''
789 Do the same as trouble, but prefixes the message with 'ERROR:', colored
790 in red if stderr is a tty file.
791 '''
792 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
793 _msg_header = '\033[0;31mERROR:\033[0m'
794 else:
795 _msg_header = 'ERROR:'
796 error_message = '%s %s' % (_msg_header, message)
797 self.trouble(error_message, tb)
798
799 def write_debug(self, message, only_once=False):
800 '''Log debug message or Print message to stderr'''
801 if not self.params.get('verbose', False):
802 return
803 message = '[debug] %s' % message
804 if self.params.get('logger'):
805 self.params['logger'].debug(message)
806 else:
807 self.to_stderr(message, only_once)
808
809 def report_file_already_downloaded(self, file_name):
810 """Report file has already been fully downloaded."""
811 try:
812 self.to_screen('[download] %s has already been downloaded' % file_name)
813 except UnicodeEncodeError:
814 self.to_screen('[download] The file has already been downloaded')
815
816 def report_file_delete(self, file_name):
817 """Report that existing file will be deleted."""
818 try:
819 self.to_screen('Deleting existing file %s' % file_name)
820 except UnicodeEncodeError:
821 self.to_screen('Deleting existing file')
822
823 def parse_outtmpl(self):
824 outtmpl_dict = self.params.get('outtmpl', {})
825 if not isinstance(outtmpl_dict, dict):
826 outtmpl_dict = {'default': outtmpl_dict}
827 outtmpl_dict.update({
828 k: v for k, v in DEFAULT_OUTTMPL.items()
829 if not outtmpl_dict.get(k)})
830 for key, val in outtmpl_dict.items():
831 if isinstance(val, bytes):
832 self.report_warning(
833 'Parameter outtmpl is bytes, but should be a unicode string. '
834 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
835 return outtmpl_dict
836
837 def get_output_path(self, dir_type='', filename=None):
838 paths = self.params.get('paths', {})
839 assert isinstance(paths, dict)
840 path = os.path.join(
841 expand_path(paths.get('home', '').strip()),
842 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
843 filename or '')
844
845 # Temporary fix for #4787
846 # 'Treat' all problem characters by passing filename through preferredencoding
847 # to workaround encoding issues with subprocess on python2 @ Windows
848 if sys.version_info < (3, 0) and sys.platform == 'win32':
849 path = encodeFilename(path, True).decode(preferredencoding())
850 return sanitize_path(path, force=self.params.get('windowsfilenames'))
851
852 @staticmethod
853 def _outtmpl_expandpath(outtmpl):
854 # expand_path translates '%%' into '%' and '$$' into '$'
855 # correspondingly that is not what we want since we need to keep
856 # '%%' intact for template dict substitution step. Working around
857 # with boundary-alike separator hack.
858 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
859 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
860
861 # outtmpl should be expand_path'ed before template dict substitution
862 # because meta fields may contain env variables we don't want to
863 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
864 # title "Hello $PATH", we don't want `$PATH` to be expanded.
865 return expand_path(outtmpl).replace(sep, '')
866
867 @staticmethod
868 def escape_outtmpl(outtmpl):
869 ''' Escape any remaining strings like %s, %abc% etc. '''
870 return re.sub(
871 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
872 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
873 outtmpl)
874
875 @classmethod
876 def validate_outtmpl(cls, outtmpl):
877 ''' @return None or Exception object '''
878 outtmpl = re.sub(
879 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljq]'),
880 lambda mobj: f'{mobj.group(0)[:-1]}s',
881 cls._outtmpl_expandpath(outtmpl))
882 try:
883 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
884 return None
885 except ValueError as err:
886 return err
887
888 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=None):
889 """ Make the template and info_dict suitable for substitution : ydl.outtmpl_escape(outtmpl) % info_dict """
890 info_dict = dict(info_dict)
891 na = self.params.get('outtmpl_na_placeholder', 'NA')
892
893 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
894 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
895 if info_dict.get('duration', None) is not None
896 else None)
897 info_dict['epoch'] = int(time.time())
898 info_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
899 if info_dict.get('resolution') is None:
900 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
901
902 # For fields playlist_index and autonumber convert all occurrences
903 # of %(field)s to %(field)0Nd for backward compatibility
904 field_size_compat_map = {
905 'playlist_index': len(str(info_dict.get('_last_playlist_index') or '')),
906 'autonumber': self.params.get('autonumber_size') or 5,
907 }
908
909 TMPL_DICT = {}
910 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljq]'))
911 MATH_FUNCTIONS = {
912 '+': float.__add__,
913 '-': float.__sub__,
914 }
915 # Field is of the form key1.key2...
916 # where keys (except first) can be string, int or slice
917 FIELD_RE = r'\w+(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
918 MATH_FIELD_RE = r'''{field}|{num}'''.format(field=FIELD_RE, num=r'-?\d+(?:.\d+)?')
919 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
920 INTERNAL_FORMAT_RE = re.compile(r'''(?x)
921 (?P<negate>-)?
922 (?P<fields>{field})
923 (?P<maths>(?:{math_op}{math_field})*)
924 (?:>(?P<strf_format>.+?))?
925 (?:\|(?P<default>.*?))?
926 $'''.format(field=FIELD_RE, math_op=MATH_OPERATORS_RE, math_field=MATH_FIELD_RE))
927
928 get_key = lambda k: traverse_obj(
929 info_dict, k.split('.'), is_user_input=True, traverse_string=True)
930
931 def get_value(mdict):
932 # Object traversal
933 value = get_key(mdict['fields'])
934 # Negative
935 if mdict['negate']:
936 value = float_or_none(value)
937 if value is not None:
938 value *= -1
939 # Do maths
940 offset_key = mdict['maths']
941 if offset_key:
942 value = float_or_none(value)
943 operator = None
944 while offset_key:
945 item = re.match(
946 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
947 offset_key).group(0)
948 offset_key = offset_key[len(item):]
949 if operator is None:
950 operator = MATH_FUNCTIONS[item]
951 continue
952 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
953 offset = float_or_none(item)
954 if offset is None:
955 offset = float_or_none(get_key(item))
956 try:
957 value = operator(value, multiplier * offset)
958 except (TypeError, ZeroDivisionError):
959 return None
960 operator = None
961 # Datetime formatting
962 if mdict['strf_format']:
963 value = strftime_or_none(value, mdict['strf_format'])
964
965 return value
966
967 def create_key(outer_mobj):
968 if not outer_mobj.group('has_key'):
969 return f'%{outer_mobj.group(0)}'
970
971 prefix = outer_mobj.group('prefix')
972 key = outer_mobj.group('key')
973 original_fmt = fmt = outer_mobj.group('format')
974 mobj = re.match(INTERNAL_FORMAT_RE, key)
975 if mobj is None:
976 value, default, mobj = None, na, {'fields': ''}
977 else:
978 mobj = mobj.groupdict()
979 default = mobj['default'] if mobj['default'] is not None else na
980 value = get_value(mobj)
981
982 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
983 fmt = '0{:d}d'.format(field_size_compat_map[key])
984
985 value = default if value is None else value
986
987 str_fmt = f'{fmt[:-1]}s'
988 if fmt[-1] == 'l':
989 value, fmt = ', '.join(variadic(value)), str_fmt
990 elif fmt[-1] == 'j':
991 value, fmt = json.dumps(value), str_fmt
992 elif fmt[-1] == 'q':
993 value, fmt = compat_shlex_quote(str(value)), str_fmt
994 elif fmt[-1] == 'c':
995 value = str(value)
996 if value is None:
997 value, fmt = default, 's'
998 else:
999 value = value[0]
1000 elif fmt[-1] not in 'rs': # numeric
1001 value = float_or_none(value)
1002 if value is None:
1003 value, fmt = default, 's'
1004
1005 if sanitize:
1006 if fmt[-1] == 'r':
1007 # If value is an object, sanitize might convert it to a string
1008 # So we convert it to repr first
1009 value, fmt = repr(value), str_fmt
1010 if fmt[-1] in 'csr':
1011 value = sanitize(mobj['fields'].split('.')[-1], value)
1012
1013 key = '%s\0%s' % (key.replace('%', '%\0'), original_fmt)
1014 TMPL_DICT[key] = value
1015 return f'{prefix}%({key}){fmt}'
1016
1017 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1018
1019 def _prepare_filename(self, info_dict, tmpl_type='default'):
1020 try:
1021 sanitize = lambda k, v: sanitize_filename(
1022 compat_str(v),
1023 restricted=self.params.get('restrictfilenames'),
1024 is_id=(k == 'id' or k.endswith('_id')))
1025 outtmpl = self.outtmpl_dict.get(tmpl_type, self.outtmpl_dict['default'])
1026 outtmpl, template_dict = self.prepare_outtmpl(outtmpl, info_dict, sanitize)
1027 outtmpl = self.escape_outtmpl(self._outtmpl_expandpath(outtmpl))
1028 filename = outtmpl % template_dict
1029
1030 force_ext = OUTTMPL_TYPES.get(tmpl_type)
1031 if force_ext is not None:
1032 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1033
1034 # https://github.com/blackjack4494/youtube-dlc/issues/85
1035 trim_file_name = self.params.get('trim_file_name', False)
1036 if trim_file_name:
1037 fn_groups = filename.rsplit('.')
1038 ext = fn_groups[-1]
1039 sub_ext = ''
1040 if len(fn_groups) > 2:
1041 sub_ext = fn_groups[-2]
1042 filename = '.'.join(filter(None, [fn_groups[0][:trim_file_name], sub_ext, ext]))
1043
1044 return filename
1045 except ValueError as err:
1046 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1047 return None
1048
1049 def prepare_filename(self, info_dict, dir_type='', warn=False):
1050 """Generate the output filename."""
1051
1052 filename = self._prepare_filename(info_dict, dir_type or 'default')
1053
1054 if warn:
1055 if not self.params.get('paths'):
1056 pass
1057 elif filename == '-':
1058 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1059 elif os.path.isabs(filename):
1060 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1061 self.__prepare_filename_warned = True
1062 if filename == '-' or not filename:
1063 return filename
1064
1065 return self.get_output_path(dir_type, filename)
1066
1067 def _match_entry(self, info_dict, incomplete=False, silent=False):
1068 """ Returns None if the file should be downloaded """
1069
1070 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1071
1072 def check_filter():
1073 if 'title' in info_dict:
1074 # This can happen when we're just evaluating the playlist
1075 title = info_dict['title']
1076 matchtitle = self.params.get('matchtitle', False)
1077 if matchtitle:
1078 if not re.search(matchtitle, title, re.IGNORECASE):
1079 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1080 rejecttitle = self.params.get('rejecttitle', False)
1081 if rejecttitle:
1082 if re.search(rejecttitle, title, re.IGNORECASE):
1083 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1084 date = info_dict.get('upload_date')
1085 if date is not None:
1086 dateRange = self.params.get('daterange', DateRange())
1087 if date not in dateRange:
1088 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
1089 view_count = info_dict.get('view_count')
1090 if view_count is not None:
1091 min_views = self.params.get('min_views')
1092 if min_views is not None and view_count < min_views:
1093 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1094 max_views = self.params.get('max_views')
1095 if max_views is not None and view_count > max_views:
1096 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1097 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1098 return 'Skipping "%s" because it is age restricted' % video_title
1099
1100 if not incomplete:
1101 match_filter = self.params.get('match_filter')
1102 if match_filter is not None:
1103 ret = match_filter(info_dict)
1104 if ret is not None:
1105 return ret
1106 return None
1107
1108 if self.in_download_archive(info_dict):
1109 reason = '%s has already been recorded in the archive' % video_title
1110 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1111 else:
1112 reason = check_filter()
1113 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1114 if reason is not None:
1115 if not silent:
1116 self.to_screen('[download] ' + reason)
1117 if self.params.get(break_opt, False):
1118 raise break_err()
1119 return reason
1120
1121 @staticmethod
1122 def add_extra_info(info_dict, extra_info):
1123 '''Set the keys from extra_info in info dict if they are missing'''
1124 for key, value in extra_info.items():
1125 info_dict.setdefault(key, value)
1126
1127 def extract_info(self, url, download=True, ie_key=None, extra_info={},
1128 process=True, force_generic_extractor=False):
1129 """
1130 Return a list with a dictionary for each video extracted.
1131
1132 Arguments:
1133 url -- URL to extract
1134
1135 Keyword arguments:
1136 download -- whether to download videos during extraction
1137 ie_key -- extractor key hint
1138 extra_info -- dictionary containing the extra values to add to each result
1139 process -- whether to resolve all unresolved references (URLs, playlist items),
1140 must be True for download to work.
1141 force_generic_extractor -- force using the generic extractor
1142 """
1143
1144 if not ie_key and force_generic_extractor:
1145 ie_key = 'Generic'
1146
1147 if ie_key:
1148 ies = [self.get_info_extractor(ie_key)]
1149 else:
1150 ies = self._ies
1151
1152 for ie in ies:
1153 if not ie.suitable(url):
1154 continue
1155
1156 ie_key = ie.ie_key()
1157 ie = self.get_info_extractor(ie_key)
1158 if not ie.working():
1159 self.report_warning('The program functionality for this site has been marked as broken, '
1160 'and will probably not work.')
1161
1162 try:
1163 temp_id = str_or_none(
1164 ie.extract_id(url) if callable(getattr(ie, 'extract_id', None))
1165 else ie._match_id(url))
1166 except (AssertionError, IndexError, AttributeError):
1167 temp_id = None
1168 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1169 self.to_screen("[%s] %s: has already been recorded in archive" % (
1170 ie_key, temp_id))
1171 break
1172 return self.__extract_info(url, ie, download, extra_info, process)
1173 else:
1174 self.report_error('no suitable InfoExtractor for URL %s' % url)
1175
1176 def __handle_extraction_exceptions(func, handle_all_errors=True):
1177 def wrapper(self, *args, **kwargs):
1178 try:
1179 return func(self, *args, **kwargs)
1180 except GeoRestrictedError as e:
1181 msg = e.msg
1182 if e.countries:
1183 msg += '\nThis video is available in %s.' % ', '.join(
1184 map(ISO3166Utils.short2full, e.countries))
1185 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1186 self.report_error(msg)
1187 except ExtractorError as e: # An error we somewhat expected
1188 self.report_error(compat_str(e), e.format_traceback())
1189 except ThrottledDownload:
1190 self.to_stderr('\r')
1191 self.report_warning('The download speed is below throttle limit. Re-extracting data')
1192 return wrapper(self, *args, **kwargs)
1193 except (MaxDownloadsReached, ExistingVideoReached, RejectedVideoReached):
1194 raise
1195 except Exception as e:
1196 if handle_all_errors and self.params.get('ignoreerrors', False):
1197 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
1198 else:
1199 raise
1200 return wrapper
1201
1202 @__handle_extraction_exceptions
1203 def __extract_info(self, url, ie, download, extra_info, process):
1204 ie_result = ie.extract(url)
1205 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1206 return
1207 if isinstance(ie_result, list):
1208 # Backwards compatibility: old IE result format
1209 ie_result = {
1210 '_type': 'compat_list',
1211 'entries': ie_result,
1212 }
1213 if extra_info.get('original_url'):
1214 ie_result.setdefault('original_url', extra_info['original_url'])
1215 self.add_default_extra_info(ie_result, ie, url)
1216 if process:
1217 return self.process_ie_result(ie_result, download, extra_info)
1218 else:
1219 return ie_result
1220
1221 def add_default_extra_info(self, ie_result, ie, url):
1222 if url is not None:
1223 self.add_extra_info(ie_result, {
1224 'webpage_url': url,
1225 'original_url': url,
1226 'webpage_url_basename': url_basename(url),
1227 })
1228 if ie is not None:
1229 self.add_extra_info(ie_result, {
1230 'extractor': ie.IE_NAME,
1231 'extractor_key': ie.ie_key(),
1232 })
1233
1234 def process_ie_result(self, ie_result, download=True, extra_info={}):
1235 """
1236 Take the result of the ie(may be modified) and resolve all unresolved
1237 references (URLs, playlist items).
1238
1239 It will also download the videos if 'download'.
1240 Returns the resolved ie_result.
1241 """
1242 result_type = ie_result.get('_type', 'video')
1243
1244 if result_type in ('url', 'url_transparent'):
1245 ie_result['url'] = sanitize_url(ie_result['url'])
1246 if ie_result.get('original_url'):
1247 extra_info.setdefault('original_url', ie_result['original_url'])
1248
1249 extract_flat = self.params.get('extract_flat', False)
1250 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1251 or extract_flat is True):
1252 info_copy = ie_result.copy()
1253 self.add_extra_info(info_copy, extra_info)
1254 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1255 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1256 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1257 return ie_result
1258
1259 if result_type == 'video':
1260 self.add_extra_info(ie_result, extra_info)
1261 ie_result = self.process_video_result(ie_result, download=download)
1262 additional_urls = (ie_result or {}).get('additional_urls')
1263 if additional_urls:
1264 # TODO: Improve MetadataFromFieldPP to allow setting a list
1265 if isinstance(additional_urls, compat_str):
1266 additional_urls = [additional_urls]
1267 self.to_screen(
1268 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1269 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1270 ie_result['additional_entries'] = [
1271 self.extract_info(
1272 url, download, extra_info,
1273 force_generic_extractor=self.params.get('force_generic_extractor'))
1274 for url in additional_urls
1275 ]
1276 return ie_result
1277 elif result_type == 'url':
1278 # We have to add extra_info to the results because it may be
1279 # contained in a playlist
1280 return self.extract_info(
1281 ie_result['url'], download,
1282 ie_key=ie_result.get('ie_key'),
1283 extra_info=extra_info)
1284 elif result_type == 'url_transparent':
1285 # Use the information from the embedding page
1286 info = self.extract_info(
1287 ie_result['url'], ie_key=ie_result.get('ie_key'),
1288 extra_info=extra_info, download=False, process=False)
1289
1290 # extract_info may return None when ignoreerrors is enabled and
1291 # extraction failed with an error, don't crash and return early
1292 # in this case
1293 if not info:
1294 return info
1295
1296 force_properties = dict(
1297 (k, v) for k, v in ie_result.items() if v is not None)
1298 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1299 if f in force_properties:
1300 del force_properties[f]
1301 new_result = info.copy()
1302 new_result.update(force_properties)
1303
1304 # Extracted info may not be a video result (i.e.
1305 # info.get('_type', 'video') != video) but rather an url or
1306 # url_transparent. In such cases outer metadata (from ie_result)
1307 # should be propagated to inner one (info). For this to happen
1308 # _type of info should be overridden with url_transparent. This
1309 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1310 if new_result.get('_type') == 'url':
1311 new_result['_type'] = 'url_transparent'
1312
1313 return self.process_ie_result(
1314 new_result, download=download, extra_info=extra_info)
1315 elif result_type in ('playlist', 'multi_video'):
1316 # Protect from infinite recursion due to recursively nested playlists
1317 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1318 webpage_url = ie_result['webpage_url']
1319 if webpage_url in self._playlist_urls:
1320 self.to_screen(
1321 '[download] Skipping already downloaded playlist: %s'
1322 % ie_result.get('title') or ie_result.get('id'))
1323 return
1324
1325 self._playlist_level += 1
1326 self._playlist_urls.add(webpage_url)
1327 self._sanitize_thumbnails(ie_result)
1328 try:
1329 return self.__process_playlist(ie_result, download)
1330 finally:
1331 self._playlist_level -= 1
1332 if not self._playlist_level:
1333 self._playlist_urls.clear()
1334 elif result_type == 'compat_list':
1335 self.report_warning(
1336 'Extractor %s returned a compat_list result. '
1337 'It needs to be updated.' % ie_result.get('extractor'))
1338
1339 def _fixup(r):
1340 self.add_extra_info(
1341 r,
1342 {
1343 'extractor': ie_result['extractor'],
1344 'webpage_url': ie_result['webpage_url'],
1345 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1346 'extractor_key': ie_result['extractor_key'],
1347 }
1348 )
1349 return r
1350 ie_result['entries'] = [
1351 self.process_ie_result(_fixup(r), download, extra_info)
1352 for r in ie_result['entries']
1353 ]
1354 return ie_result
1355 else:
1356 raise Exception('Invalid result type: %s' % result_type)
1357
1358 def _ensure_dir_exists(self, path):
1359 return make_dir(path, self.report_error)
1360
1361 def __process_playlist(self, ie_result, download):
1362 # We process each entry in the playlist
1363 playlist = ie_result.get('title') or ie_result.get('id')
1364 self.to_screen('[download] Downloading playlist: %s' % playlist)
1365
1366 if 'entries' not in ie_result:
1367 raise EntryNotInPlaylist()
1368 incomplete_entries = bool(ie_result.get('requested_entries'))
1369 if incomplete_entries:
1370 def fill_missing_entries(entries, indexes):
1371 ret = [None] * max(*indexes)
1372 for i, entry in zip(indexes, entries):
1373 ret[i - 1] = entry
1374 return ret
1375 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1376
1377 playlist_results = []
1378
1379 playliststart = self.params.get('playliststart', 1)
1380 playlistend = self.params.get('playlistend')
1381 # For backwards compatibility, interpret -1 as whole list
1382 if playlistend == -1:
1383 playlistend = None
1384
1385 playlistitems_str = self.params.get('playlist_items')
1386 playlistitems = None
1387 if playlistitems_str is not None:
1388 def iter_playlistitems(format):
1389 for string_segment in format.split(','):
1390 if '-' in string_segment:
1391 start, end = string_segment.split('-')
1392 for item in range(int(start), int(end) + 1):
1393 yield int(item)
1394 else:
1395 yield int(string_segment)
1396 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1397
1398 ie_entries = ie_result['entries']
1399 msg = (
1400 'Downloading %d videos' if not isinstance(ie_entries, list)
1401 else 'Collected %d videos; downloading %%d of them' % len(ie_entries))
1402 if not isinstance(ie_entries, (list, PagedList)):
1403 ie_entries = LazyList(ie_entries)
1404
1405 def get_entry(i):
1406 return YoutubeDL.__handle_extraction_exceptions(
1407 lambda self, i: ie_entries[i - 1],
1408 False
1409 )(self, i)
1410
1411 entries = []
1412 for i in playlistitems or itertools.count(playliststart):
1413 if playlistitems is None and playlistend is not None and playlistend < i:
1414 break
1415 entry = None
1416 try:
1417 entry = get_entry(i)
1418 if entry is None:
1419 raise EntryNotInPlaylist()
1420 except (IndexError, EntryNotInPlaylist):
1421 if incomplete_entries:
1422 raise EntryNotInPlaylist()
1423 elif not playlistitems:
1424 break
1425 entries.append(entry)
1426 try:
1427 if entry is not None:
1428 self._match_entry(entry, incomplete=True, silent=True)
1429 except (ExistingVideoReached, RejectedVideoReached):
1430 break
1431 ie_result['entries'] = entries
1432
1433 # Save playlist_index before re-ordering
1434 entries = [
1435 ((playlistitems[i - 1] if playlistitems else i), entry)
1436 for i, entry in enumerate(entries, 1)
1437 if entry is not None]
1438 n_entries = len(entries)
1439
1440 if not playlistitems and (playliststart or playlistend):
1441 playlistitems = list(range(playliststart, playliststart + n_entries))
1442 ie_result['requested_entries'] = playlistitems
1443
1444 if self.params.get('allow_playlist_files', True):
1445 ie_copy = {
1446 'playlist': playlist,
1447 'playlist_id': ie_result.get('id'),
1448 'playlist_title': ie_result.get('title'),
1449 'playlist_uploader': ie_result.get('uploader'),
1450 'playlist_uploader_id': ie_result.get('uploader_id'),
1451 'playlist_index': 0,
1452 }
1453 ie_copy.update(dict(ie_result))
1454
1455 if self.params.get('writeinfojson', False):
1456 infofn = self.prepare_filename(ie_copy, 'pl_infojson')
1457 if not self._ensure_dir_exists(encodeFilename(infofn)):
1458 return
1459 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
1460 self.to_screen('[info] Playlist metadata is already present')
1461 else:
1462 self.to_screen('[info] Writing playlist metadata as JSON to: ' + infofn)
1463 try:
1464 write_json_file(self.filter_requested_info(ie_result, self.params.get('clean_infojson', True)), infofn)
1465 except (OSError, IOError):
1466 self.report_error('Cannot write playlist metadata to JSON file ' + infofn)
1467
1468 # TODO: This should be passed to ThumbnailsConvertor if necessary
1469 self._write_thumbnails(ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1470
1471 if self.params.get('writedescription', False):
1472 descfn = self.prepare_filename(ie_copy, 'pl_description')
1473 if not self._ensure_dir_exists(encodeFilename(descfn)):
1474 return
1475 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
1476 self.to_screen('[info] Playlist description is already present')
1477 elif ie_result.get('description') is None:
1478 self.report_warning('There\'s no playlist description to write.')
1479 else:
1480 try:
1481 self.to_screen('[info] Writing playlist description to: ' + descfn)
1482 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1483 descfile.write(ie_result['description'])
1484 except (OSError, IOError):
1485 self.report_error('Cannot write playlist description file ' + descfn)
1486 return
1487
1488 if self.params.get('playlistreverse', False):
1489 entries = entries[::-1]
1490 if self.params.get('playlistrandom', False):
1491 random.shuffle(entries)
1492
1493 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1494
1495 self.to_screen('[%s] playlist %s: %s' % (ie_result['extractor'], playlist, msg % n_entries))
1496 failures = 0
1497 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1498 for i, entry_tuple in enumerate(entries, 1):
1499 playlist_index, entry = entry_tuple
1500 if 'playlist_index' in self.params.get('compat_options', []):
1501 playlist_index = playlistitems[i - 1] if playlistitems else i
1502 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1503 # This __x_forwarded_for_ip thing is a bit ugly but requires
1504 # minimal changes
1505 if x_forwarded_for:
1506 entry['__x_forwarded_for_ip'] = x_forwarded_for
1507 extra = {
1508 'n_entries': n_entries,
1509 '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1510 'playlist_index': playlist_index,
1511 'playlist_autonumber': i,
1512 'playlist': playlist,
1513 'playlist_id': ie_result.get('id'),
1514 'playlist_title': ie_result.get('title'),
1515 'playlist_uploader': ie_result.get('uploader'),
1516 'playlist_uploader_id': ie_result.get('uploader_id'),
1517 'extractor': ie_result['extractor'],
1518 'webpage_url': ie_result['webpage_url'],
1519 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1520 'extractor_key': ie_result['extractor_key'],
1521 }
1522
1523 if self._match_entry(entry, incomplete=True) is not None:
1524 continue
1525
1526 entry_result = self.__process_iterable_entry(entry, download, extra)
1527 if not entry_result:
1528 failures += 1
1529 if failures >= max_failures:
1530 self.report_error(
1531 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1532 break
1533 # TODO: skip failed (empty) entries?
1534 playlist_results.append(entry_result)
1535 ie_result['entries'] = playlist_results
1536 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1537 return ie_result
1538
1539 @__handle_extraction_exceptions
1540 def __process_iterable_entry(self, entry, download, extra_info):
1541 return self.process_ie_result(
1542 entry, download=download, extra_info=extra_info)
1543
1544 def _build_format_filter(self, filter_spec):
1545 " Returns a function to filter the formats according to the filter_spec "
1546
1547 OPERATORS = {
1548 '<': operator.lt,
1549 '<=': operator.le,
1550 '>': operator.gt,
1551 '>=': operator.ge,
1552 '=': operator.eq,
1553 '!=': operator.ne,
1554 }
1555 operator_rex = re.compile(r'''(?x)\s*
1556 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1557 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1558 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1559 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1560 m = operator_rex.fullmatch(filter_spec)
1561 if m:
1562 try:
1563 comparison_value = int(m.group('value'))
1564 except ValueError:
1565 comparison_value = parse_filesize(m.group('value'))
1566 if comparison_value is None:
1567 comparison_value = parse_filesize(m.group('value') + 'B')
1568 if comparison_value is None:
1569 raise ValueError(
1570 'Invalid value %r in format specification %r' % (
1571 m.group('value'), filter_spec))
1572 op = OPERATORS[m.group('op')]
1573
1574 if not m:
1575 STR_OPERATORS = {
1576 '=': operator.eq,
1577 '^=': lambda attr, value: attr.startswith(value),
1578 '$=': lambda attr, value: attr.endswith(value),
1579 '*=': lambda attr, value: value in attr,
1580 }
1581 str_operator_rex = re.compile(r'''(?x)\s*
1582 (?P<key>[a-zA-Z0-9._-]+)\s*
1583 (?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1584 (?P<value>[a-zA-Z0-9._-]+)\s*
1585 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1586 m = str_operator_rex.fullmatch(filter_spec)
1587 if m:
1588 comparison_value = m.group('value')
1589 str_op = STR_OPERATORS[m.group('op')]
1590 if m.group('negation'):
1591 op = lambda attr, value: not str_op(attr, value)
1592 else:
1593 op = str_op
1594
1595 if not m:
1596 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1597
1598 def _filter(f):
1599 actual_value = f.get(m.group('key'))
1600 if actual_value is None:
1601 return m.group('none_inclusive')
1602 return op(actual_value, comparison_value)
1603 return _filter
1604
1605 def _default_format_spec(self, info_dict, download=True):
1606
1607 def can_merge():
1608 merger = FFmpegMergerPP(self)
1609 return merger.available and merger.can_merge()
1610
1611 prefer_best = (
1612 not self.params.get('simulate', False)
1613 and download
1614 and (
1615 not can_merge()
1616 or info_dict.get('is_live', False)
1617 or self.outtmpl_dict['default'] == '-'))
1618 compat = (
1619 prefer_best
1620 or self.params.get('allow_multiple_audio_streams', False)
1621 or 'format-spec' in self.params.get('compat_opts', []))
1622
1623 return (
1624 'best/bestvideo+bestaudio' if prefer_best
1625 else 'bestvideo*+bestaudio/best' if not compat
1626 else 'bestvideo+bestaudio/best')
1627
1628 def build_format_selector(self, format_spec):
1629 def syntax_error(note, start):
1630 message = (
1631 'Invalid format specification: '
1632 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1633 return SyntaxError(message)
1634
1635 PICKFIRST = 'PICKFIRST'
1636 MERGE = 'MERGE'
1637 SINGLE = 'SINGLE'
1638 GROUP = 'GROUP'
1639 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1640
1641 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1642 'video': self.params.get('allow_multiple_video_streams', False)}
1643
1644 check_formats = self.params.get('check_formats')
1645
1646 def _parse_filter(tokens):
1647 filter_parts = []
1648 for type, string, start, _, _ in tokens:
1649 if type == tokenize.OP and string == ']':
1650 return ''.join(filter_parts)
1651 else:
1652 filter_parts.append(string)
1653
1654 def _remove_unused_ops(tokens):
1655 # Remove operators that we don't use and join them with the surrounding strings
1656 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1657 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1658 last_string, last_start, last_end, last_line = None, None, None, None
1659 for type, string, start, end, line in tokens:
1660 if type == tokenize.OP and string == '[':
1661 if last_string:
1662 yield tokenize.NAME, last_string, last_start, last_end, last_line
1663 last_string = None
1664 yield type, string, start, end, line
1665 # everything inside brackets will be handled by _parse_filter
1666 for type, string, start, end, line in tokens:
1667 yield type, string, start, end, line
1668 if type == tokenize.OP and string == ']':
1669 break
1670 elif type == tokenize.OP and string in ALLOWED_OPS:
1671 if last_string:
1672 yield tokenize.NAME, last_string, last_start, last_end, last_line
1673 last_string = None
1674 yield type, string, start, end, line
1675 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1676 if not last_string:
1677 last_string = string
1678 last_start = start
1679 last_end = end
1680 else:
1681 last_string += string
1682 if last_string:
1683 yield tokenize.NAME, last_string, last_start, last_end, last_line
1684
1685 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1686 selectors = []
1687 current_selector = None
1688 for type, string, start, _, _ in tokens:
1689 # ENCODING is only defined in python 3.x
1690 if type == getattr(tokenize, 'ENCODING', None):
1691 continue
1692 elif type in [tokenize.NAME, tokenize.NUMBER]:
1693 current_selector = FormatSelector(SINGLE, string, [])
1694 elif type == tokenize.OP:
1695 if string == ')':
1696 if not inside_group:
1697 # ')' will be handled by the parentheses group
1698 tokens.restore_last_token()
1699 break
1700 elif inside_merge and string in ['/', ',']:
1701 tokens.restore_last_token()
1702 break
1703 elif inside_choice and string == ',':
1704 tokens.restore_last_token()
1705 break
1706 elif string == ',':
1707 if not current_selector:
1708 raise syntax_error('"," must follow a format selector', start)
1709 selectors.append(current_selector)
1710 current_selector = None
1711 elif string == '/':
1712 if not current_selector:
1713 raise syntax_error('"/" must follow a format selector', start)
1714 first_choice = current_selector
1715 second_choice = _parse_format_selection(tokens, inside_choice=True)
1716 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1717 elif string == '[':
1718 if not current_selector:
1719 current_selector = FormatSelector(SINGLE, 'best', [])
1720 format_filter = _parse_filter(tokens)
1721 current_selector.filters.append(format_filter)
1722 elif string == '(':
1723 if current_selector:
1724 raise syntax_error('Unexpected "("', start)
1725 group = _parse_format_selection(tokens, inside_group=True)
1726 current_selector = FormatSelector(GROUP, group, [])
1727 elif string == '+':
1728 if not current_selector:
1729 raise syntax_error('Unexpected "+"', start)
1730 selector_1 = current_selector
1731 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1732 if not selector_2:
1733 raise syntax_error('Expected a selector', start)
1734 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1735 else:
1736 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1737 elif type == tokenize.ENDMARKER:
1738 break
1739 if current_selector:
1740 selectors.append(current_selector)
1741 return selectors
1742
1743 def _merge(formats_pair):
1744 format_1, format_2 = formats_pair
1745
1746 formats_info = []
1747 formats_info.extend(format_1.get('requested_formats', (format_1,)))
1748 formats_info.extend(format_2.get('requested_formats', (format_2,)))
1749
1750 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
1751 get_no_more = {'video': False, 'audio': False}
1752 for (i, fmt_info) in enumerate(formats_info):
1753 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
1754 formats_info.pop(i)
1755 continue
1756 for aud_vid in ['audio', 'video']:
1757 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
1758 if get_no_more[aud_vid]:
1759 formats_info.pop(i)
1760 break
1761 get_no_more[aud_vid] = True
1762
1763 if len(formats_info) == 1:
1764 return formats_info[0]
1765
1766 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
1767 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
1768
1769 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
1770 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
1771
1772 output_ext = self.params.get('merge_output_format')
1773 if not output_ext:
1774 if the_only_video:
1775 output_ext = the_only_video['ext']
1776 elif the_only_audio and not video_fmts:
1777 output_ext = the_only_audio['ext']
1778 else:
1779 output_ext = 'mkv'
1780
1781 new_dict = {
1782 'requested_formats': formats_info,
1783 'format': '+'.join(fmt_info.get('format') for fmt_info in formats_info),
1784 'format_id': '+'.join(fmt_info.get('format_id') for fmt_info in formats_info),
1785 'ext': output_ext,
1786 }
1787
1788 if the_only_video:
1789 new_dict.update({
1790 'width': the_only_video.get('width'),
1791 'height': the_only_video.get('height'),
1792 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
1793 'fps': the_only_video.get('fps'),
1794 'vcodec': the_only_video.get('vcodec'),
1795 'vbr': the_only_video.get('vbr'),
1796 'stretched_ratio': the_only_video.get('stretched_ratio'),
1797 })
1798
1799 if the_only_audio:
1800 new_dict.update({
1801 'acodec': the_only_audio.get('acodec'),
1802 'abr': the_only_audio.get('abr'),
1803 })
1804
1805 return new_dict
1806
1807 def _check_formats(formats):
1808 if not check_formats:
1809 yield from formats
1810 return
1811 for f in formats:
1812 self.to_screen('[info] Testing format %s' % f['format_id'])
1813 temp_file = tempfile.NamedTemporaryFile(
1814 suffix='.tmp', delete=False,
1815 dir=self.get_output_path('temp') or None)
1816 temp_file.close()
1817 try:
1818 success, _ = self.dl(temp_file.name, f, test=True)
1819 except (DownloadError, IOError, OSError, ValueError) + network_exceptions:
1820 success = False
1821 finally:
1822 if os.path.exists(temp_file.name):
1823 try:
1824 os.remove(temp_file.name)
1825 except OSError:
1826 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1827 if success:
1828 yield f
1829 else:
1830 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1831
1832 def _build_selector_function(selector):
1833 if isinstance(selector, list): # ,
1834 fs = [_build_selector_function(s) for s in selector]
1835
1836 def selector_function(ctx):
1837 for f in fs:
1838 yield from f(ctx)
1839 return selector_function
1840
1841 elif selector.type == GROUP: # ()
1842 selector_function = _build_selector_function(selector.selector)
1843
1844 elif selector.type == PICKFIRST: # /
1845 fs = [_build_selector_function(s) for s in selector.selector]
1846
1847 def selector_function(ctx):
1848 for f in fs:
1849 picked_formats = list(f(ctx))
1850 if picked_formats:
1851 return picked_formats
1852 return []
1853
1854 elif selector.type == MERGE: # +
1855 selector_1, selector_2 = map(_build_selector_function, selector.selector)
1856
1857 def selector_function(ctx):
1858 for pair in itertools.product(
1859 selector_1(copy.deepcopy(ctx)), selector_2(copy.deepcopy(ctx))):
1860 yield _merge(pair)
1861
1862 elif selector.type == SINGLE: # atom
1863 format_spec = selector.selector or 'best'
1864
1865 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
1866 if format_spec == 'all':
1867 def selector_function(ctx):
1868 yield from _check_formats(ctx['formats'])
1869 elif format_spec == 'mergeall':
1870 def selector_function(ctx):
1871 formats = list(_check_formats(ctx['formats']))
1872 if not formats:
1873 return
1874 merged_format = formats[-1]
1875 for f in formats[-2::-1]:
1876 merged_format = _merge((merged_format, f))
1877 yield merged_format
1878
1879 else:
1880 format_fallback, format_reverse, format_idx = False, True, 1
1881 mobj = re.match(
1882 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
1883 format_spec)
1884 if mobj is not None:
1885 format_idx = int_or_none(mobj.group('n'), default=1)
1886 format_reverse = mobj.group('bw')[0] == 'b'
1887 format_type = (mobj.group('type') or [None])[0]
1888 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
1889 format_modified = mobj.group('mod') is not None
1890
1891 format_fallback = not format_type and not format_modified # for b, w
1892 _filter_f = (
1893 (lambda f: f.get('%scodec' % format_type) != 'none')
1894 if format_type and format_modified # bv*, ba*, wv*, wa*
1895 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
1896 if format_type # bv, ba, wv, wa
1897 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
1898 if not format_modified # b, w
1899 else lambda f: True) # b*, w*
1900 filter_f = lambda f: _filter_f(f) and (
1901 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
1902 else:
1903 filter_f = ((lambda f: f.get('ext') == format_spec)
1904 if format_spec in ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] # extension
1905 else (lambda f: f.get('format_id') == format_spec)) # id
1906
1907 def selector_function(ctx):
1908 formats = list(ctx['formats'])
1909 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
1910 if format_fallback and ctx['incomplete_formats'] and not matches:
1911 # for extractors with incomplete formats (audio only (soundcloud)
1912 # or video only (imgur)) best/worst will fallback to
1913 # best/worst {video,audio}-only format
1914 matches = formats
1915 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
1916 try:
1917 yield matches[format_idx - 1]
1918 except IndexError:
1919 return
1920
1921 filters = [self._build_format_filter(f) for f in selector.filters]
1922
1923 def final_selector(ctx):
1924 ctx_copy = copy.deepcopy(ctx)
1925 for _filter in filters:
1926 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1927 return selector_function(ctx_copy)
1928 return final_selector
1929
1930 stream = io.BytesIO(format_spec.encode('utf-8'))
1931 try:
1932 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1933 except tokenize.TokenError:
1934 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1935
1936 class TokenIterator(object):
1937 def __init__(self, tokens):
1938 self.tokens = tokens
1939 self.counter = 0
1940
1941 def __iter__(self):
1942 return self
1943
1944 def __next__(self):
1945 if self.counter >= len(self.tokens):
1946 raise StopIteration()
1947 value = self.tokens[self.counter]
1948 self.counter += 1
1949 return value
1950
1951 next = __next__
1952
1953 def restore_last_token(self):
1954 self.counter -= 1
1955
1956 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1957 return _build_selector_function(parsed_selector)
1958
1959 def _calc_headers(self, info_dict):
1960 res = std_headers.copy()
1961
1962 add_headers = info_dict.get('http_headers')
1963 if add_headers:
1964 res.update(add_headers)
1965
1966 cookies = self._calc_cookies(info_dict)
1967 if cookies:
1968 res['Cookie'] = cookies
1969
1970 if 'X-Forwarded-For' not in res:
1971 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1972 if x_forwarded_for_ip:
1973 res['X-Forwarded-For'] = x_forwarded_for_ip
1974
1975 return res
1976
1977 def _calc_cookies(self, info_dict):
1978 pr = sanitized_Request(info_dict['url'])
1979 self.cookiejar.add_cookie_header(pr)
1980 return pr.get_header('Cookie')
1981
1982 def _sanitize_thumbnails(self, info_dict):
1983 thumbnails = info_dict.get('thumbnails')
1984 if thumbnails is None:
1985 thumbnail = info_dict.get('thumbnail')
1986 if thumbnail:
1987 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1988 if thumbnails:
1989 thumbnails.sort(key=lambda t: (
1990 t.get('preference') if t.get('preference') is not None else -1,
1991 t.get('width') if t.get('width') is not None else -1,
1992 t.get('height') if t.get('height') is not None else -1,
1993 t.get('id') if t.get('id') is not None else '',
1994 t.get('url')))
1995
1996 def thumbnail_tester():
1997 if self.params.get('check_formats'):
1998 test_all = True
1999 to_screen = lambda msg: self.to_screen(f'[info] {msg}')
2000 else:
2001 test_all = False
2002 to_screen = self.write_debug
2003
2004 def test_thumbnail(t):
2005 if not test_all and not t.get('_test_url'):
2006 return True
2007 to_screen('Testing thumbnail %s' % t['id'])
2008 try:
2009 self.urlopen(HEADRequest(t['url']))
2010 except network_exceptions as err:
2011 to_screen('Unable to connect to thumbnail %s URL "%s" - %s. Skipping...' % (
2012 t['id'], t['url'], error_to_compat_str(err)))
2013 return False
2014 return True
2015
2016 return test_thumbnail
2017
2018 for i, t in enumerate(thumbnails):
2019 if t.get('id') is None:
2020 t['id'] = '%d' % i
2021 if t.get('width') and t.get('height'):
2022 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2023 t['url'] = sanitize_url(t['url'])
2024
2025 if self.params.get('check_formats') is not False:
2026 info_dict['thumbnails'] = LazyList(filter(thumbnail_tester(), thumbnails[::-1])).reverse()
2027 else:
2028 info_dict['thumbnails'] = thumbnails
2029
2030 def process_video_result(self, info_dict, download=True):
2031 assert info_dict.get('_type', 'video') == 'video'
2032
2033 if 'id' not in info_dict:
2034 raise ExtractorError('Missing "id" field in extractor result')
2035 if 'title' not in info_dict:
2036 raise ExtractorError('Missing "title" field in extractor result')
2037
2038 def report_force_conversion(field, field_not, conversion):
2039 self.report_warning(
2040 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2041 % (field, field_not, conversion))
2042
2043 def sanitize_string_field(info, string_field):
2044 field = info.get(string_field)
2045 if field is None or isinstance(field, compat_str):
2046 return
2047 report_force_conversion(string_field, 'a string', 'string')
2048 info[string_field] = compat_str(field)
2049
2050 def sanitize_numeric_fields(info):
2051 for numeric_field in self._NUMERIC_FIELDS:
2052 field = info.get(numeric_field)
2053 if field is None or isinstance(field, compat_numeric_types):
2054 continue
2055 report_force_conversion(numeric_field, 'numeric', 'int')
2056 info[numeric_field] = int_or_none(field)
2057
2058 sanitize_string_field(info_dict, 'id')
2059 sanitize_numeric_fields(info_dict)
2060
2061 if 'playlist' not in info_dict:
2062 # It isn't part of a playlist
2063 info_dict['playlist'] = None
2064 info_dict['playlist_index'] = None
2065
2066 self._sanitize_thumbnails(info_dict)
2067
2068 thumbnail = info_dict.get('thumbnail')
2069 thumbnails = info_dict.get('thumbnails')
2070 if thumbnail:
2071 info_dict['thumbnail'] = sanitize_url(thumbnail)
2072 elif thumbnails:
2073 info_dict['thumbnail'] = thumbnails[-1]['url']
2074
2075 if info_dict.get('display_id') is None and 'id' in info_dict:
2076 info_dict['display_id'] = info_dict['id']
2077
2078 for ts_key, date_key in (
2079 ('timestamp', 'upload_date'),
2080 ('release_timestamp', 'release_date'),
2081 ):
2082 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2083 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2084 # see http://bugs.python.org/issue1646728)
2085 try:
2086 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2087 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2088 except (ValueError, OverflowError, OSError):
2089 pass
2090
2091 live_keys = ('is_live', 'was_live')
2092 live_status = info_dict.get('live_status')
2093 if live_status is None:
2094 for key in live_keys:
2095 if info_dict.get(key) is False:
2096 continue
2097 if info_dict.get(key):
2098 live_status = key
2099 break
2100 if all(info_dict.get(key) is False for key in live_keys):
2101 live_status = 'not_live'
2102 if live_status:
2103 info_dict['live_status'] = live_status
2104 for key in live_keys:
2105 if info_dict.get(key) is None:
2106 info_dict[key] = (live_status == key)
2107
2108 # Auto generate title fields corresponding to the *_number fields when missing
2109 # in order to always have clean titles. This is very common for TV series.
2110 for field in ('chapter', 'season', 'episode'):
2111 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2112 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2113
2114 for cc_kind in ('subtitles', 'automatic_captions'):
2115 cc = info_dict.get(cc_kind)
2116 if cc:
2117 for _, subtitle in cc.items():
2118 for subtitle_format in subtitle:
2119 if subtitle_format.get('url'):
2120 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2121 if subtitle_format.get('ext') is None:
2122 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2123
2124 automatic_captions = info_dict.get('automatic_captions')
2125 subtitles = info_dict.get('subtitles')
2126
2127 info_dict['requested_subtitles'] = self.process_subtitles(
2128 info_dict['id'], subtitles, automatic_captions)
2129
2130 # We now pick which formats have to be downloaded
2131 if info_dict.get('formats') is None:
2132 # There's only one format available
2133 formats = [info_dict]
2134 else:
2135 formats = info_dict['formats']
2136
2137 if not formats:
2138 if not self.params.get('ignore_no_formats_error'):
2139 raise ExtractorError('No video formats found!')
2140 else:
2141 self.report_warning('No video formats found!')
2142
2143 def is_wellformed(f):
2144 url = f.get('url')
2145 if not url:
2146 self.report_warning(
2147 '"url" field is missing or empty - skipping format, '
2148 'there is an error in extractor')
2149 return False
2150 if isinstance(url, bytes):
2151 sanitize_string_field(f, 'url')
2152 return True
2153
2154 # Filter out malformed formats for better extraction robustness
2155 formats = list(filter(is_wellformed, formats))
2156
2157 formats_dict = {}
2158
2159 # We check that all the formats have the format and format_id fields
2160 for i, format in enumerate(formats):
2161 sanitize_string_field(format, 'format_id')
2162 sanitize_numeric_fields(format)
2163 format['url'] = sanitize_url(format['url'])
2164 if not format.get('format_id'):
2165 format['format_id'] = compat_str(i)
2166 else:
2167 # Sanitize format_id from characters used in format selector expression
2168 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2169 format_id = format['format_id']
2170 if format_id not in formats_dict:
2171 formats_dict[format_id] = []
2172 formats_dict[format_id].append(format)
2173
2174 # Make sure all formats have unique format_id
2175 for format_id, ambiguous_formats in formats_dict.items():
2176 if len(ambiguous_formats) > 1:
2177 for i, format in enumerate(ambiguous_formats):
2178 format['format_id'] = '%s-%d' % (format_id, i)
2179
2180 for i, format in enumerate(formats):
2181 if format.get('format') is None:
2182 format['format'] = '{id} - {res}{note}'.format(
2183 id=format['format_id'],
2184 res=self.format_resolution(format),
2185 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
2186 )
2187 # Automatically determine file extension if missing
2188 if format.get('ext') is None:
2189 format['ext'] = determine_ext(format['url']).lower()
2190 # Automatically determine protocol if missing (useful for format
2191 # selection purposes)
2192 if format.get('protocol') is None:
2193 format['protocol'] = determine_protocol(format)
2194 # Add HTTP headers, so that external programs can use them from the
2195 # json output
2196 full_format_info = info_dict.copy()
2197 full_format_info.update(format)
2198 format['http_headers'] = self._calc_headers(full_format_info)
2199 # Remove private housekeeping stuff
2200 if '__x_forwarded_for_ip' in info_dict:
2201 del info_dict['__x_forwarded_for_ip']
2202
2203 # TODO Central sorting goes here
2204
2205 if formats and formats[0] is not info_dict:
2206 # only set the 'formats' fields if the original info_dict list them
2207 # otherwise we end up with a circular reference, the first (and unique)
2208 # element in the 'formats' field in info_dict is info_dict itself,
2209 # which can't be exported to json
2210 info_dict['formats'] = formats
2211
2212 info_dict, _ = self.pre_process(info_dict)
2213
2214 list_only = self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles')
2215 if list_only:
2216 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2217 if self.params.get('list_thumbnails'):
2218 self.list_thumbnails(info_dict)
2219 if self.params.get('listformats'):
2220 if not info_dict.get('formats'):
2221 raise ExtractorError('No video formats found', expected=True)
2222 self.list_formats(info_dict)
2223 if self.params.get('listsubtitles'):
2224 if 'automatic_captions' in info_dict:
2225 self.list_subtitles(
2226 info_dict['id'], automatic_captions, 'automatic captions')
2227 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2228 return
2229
2230 format_selector = self.format_selector
2231 if format_selector is None:
2232 req_format = self._default_format_spec(info_dict, download=download)
2233 self.write_debug('Default format spec: %s' % req_format)
2234 format_selector = self.build_format_selector(req_format)
2235
2236 # While in format selection we may need to have an access to the original
2237 # format set in order to calculate some metrics or do some processing.
2238 # For now we need to be able to guess whether original formats provided
2239 # by extractor are incomplete or not (i.e. whether extractor provides only
2240 # video-only or audio-only formats) for proper formats selection for
2241 # extractors with such incomplete formats (see
2242 # https://github.com/ytdl-org/youtube-dl/pull/5556).
2243 # Since formats may be filtered during format selection and may not match
2244 # the original formats the results may be incorrect. Thus original formats
2245 # or pre-calculated metrics should be passed to format selection routines
2246 # as well.
2247 # We will pass a context object containing all necessary additional data
2248 # instead of just formats.
2249 # This fixes incorrect format selection issue (see
2250 # https://github.com/ytdl-org/youtube-dl/issues/10083).
2251 incomplete_formats = (
2252 # All formats are video-only or
2253 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2254 # all formats are audio-only
2255 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
2256
2257 ctx = {
2258 'formats': formats,
2259 'incomplete_formats': incomplete_formats,
2260 }
2261
2262 formats_to_download = list(format_selector(ctx))
2263 if not formats_to_download:
2264 if not self.params.get('ignore_no_formats_error'):
2265 raise ExtractorError('Requested format is not available', expected=True)
2266 else:
2267 self.report_warning('Requested format is not available')
2268 # Process what we can, even without any available formats.
2269 self.process_info(dict(info_dict))
2270 elif download:
2271 self.to_screen(
2272 '[info] %s: Downloading %d format(s): %s' % (
2273 info_dict['id'], len(formats_to_download),
2274 ", ".join([f['format_id'] for f in formats_to_download])))
2275 for fmt in formats_to_download:
2276 new_info = dict(info_dict)
2277 # Save a reference to the original info_dict so that it can be modified in process_info if needed
2278 new_info['__original_infodict'] = info_dict
2279 new_info.update(fmt)
2280 self.process_info(new_info)
2281 # We update the info dict with the best quality format (backwards compatibility)
2282 if formats_to_download:
2283 info_dict.update(formats_to_download[-1])
2284 return info_dict
2285
2286 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2287 """Select the requested subtitles and their format"""
2288 available_subs = {}
2289 if normal_subtitles and self.params.get('writesubtitles'):
2290 available_subs.update(normal_subtitles)
2291 if automatic_captions and self.params.get('writeautomaticsub'):
2292 for lang, cap_info in automatic_captions.items():
2293 if lang not in available_subs:
2294 available_subs[lang] = cap_info
2295
2296 if (not self.params.get('writesubtitles') and not
2297 self.params.get('writeautomaticsub') or not
2298 available_subs):
2299 return None
2300
2301 all_sub_langs = available_subs.keys()
2302 if self.params.get('allsubtitles', False):
2303 requested_langs = all_sub_langs
2304 elif self.params.get('subtitleslangs', False):
2305 requested_langs = set()
2306 for lang in self.params.get('subtitleslangs'):
2307 if lang == 'all':
2308 requested_langs.update(all_sub_langs)
2309 continue
2310 discard = lang[0] == '-'
2311 if discard:
2312 lang = lang[1:]
2313 current_langs = filter(re.compile(lang + '$').match, all_sub_langs)
2314 if discard:
2315 for lang in current_langs:
2316 requested_langs.discard(lang)
2317 else:
2318 requested_langs.update(current_langs)
2319 elif 'en' in available_subs:
2320 requested_langs = ['en']
2321 else:
2322 requested_langs = [list(all_sub_langs)[0]]
2323 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2324
2325 formats_query = self.params.get('subtitlesformat', 'best')
2326 formats_preference = formats_query.split('/') if formats_query else []
2327 subs = {}
2328 for lang in requested_langs:
2329 formats = available_subs.get(lang)
2330 if formats is None:
2331 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
2332 continue
2333 for ext in formats_preference:
2334 if ext == 'best':
2335 f = formats[-1]
2336 break
2337 matches = list(filter(lambda f: f['ext'] == ext, formats))
2338 if matches:
2339 f = matches[-1]
2340 break
2341 else:
2342 f = formats[-1]
2343 self.report_warning(
2344 'No subtitle format found matching "%s" for language %s, '
2345 'using %s' % (formats_query, lang, f['ext']))
2346 subs[lang] = f
2347 return subs
2348
2349 def __forced_printings(self, info_dict, filename, incomplete):
2350 def print_mandatory(field, actual_field=None):
2351 if actual_field is None:
2352 actual_field = field
2353 if (self.params.get('force%s' % field, False)
2354 and (not incomplete or info_dict.get(actual_field) is not None)):
2355 self.to_stdout(info_dict[actual_field])
2356
2357 def print_optional(field):
2358 if (self.params.get('force%s' % field, False)
2359 and info_dict.get(field) is not None):
2360 self.to_stdout(info_dict[field])
2361
2362 info_dict = info_dict.copy()
2363 if filename is not None:
2364 info_dict['filename'] = filename
2365 if info_dict.get('requested_formats') is not None:
2366 # For RTMP URLs, also include the playpath
2367 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2368 elif 'url' in info_dict:
2369 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2370
2371 for tmpl in self.params.get('forceprint', []):
2372 if re.match(r'\w+$', tmpl):
2373 tmpl = '%({})s'.format(tmpl)
2374 tmpl, info_copy = self.prepare_outtmpl(tmpl, info_dict)
2375 self.to_stdout(self.escape_outtmpl(tmpl) % info_copy)
2376
2377 print_mandatory('title')
2378 print_mandatory('id')
2379 print_mandatory('url', 'urls')
2380 print_optional('thumbnail')
2381 print_optional('description')
2382 print_optional('filename')
2383 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
2384 self.to_stdout(formatSeconds(info_dict['duration']))
2385 print_mandatory('format')
2386
2387 if self.params.get('forcejson', False):
2388 self.post_extract(info_dict)
2389 self.to_stdout(json.dumps(info_dict, default=repr))
2390
2391 def dl(self, name, info, subtitle=False, test=False):
2392
2393 if test:
2394 verbose = self.params.get('verbose')
2395 params = {
2396 'test': True,
2397 'quiet': not verbose,
2398 'verbose': verbose,
2399 'noprogress': not verbose,
2400 'nopart': True,
2401 'skip_unavailable_fragments': False,
2402 'keep_fragments': False,
2403 'overwrites': True,
2404 '_no_ytdl_file': True,
2405 }
2406 else:
2407 params = self.params
2408 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2409 if not test:
2410 for ph in self._progress_hooks:
2411 fd.add_progress_hook(ph)
2412 urls = '", "'.join([f['url'] for f in info.get('requested_formats', [])] or [info['url']])
2413 self.write_debug('Invoking downloader on "%s"' % urls)
2414 new_info = dict(info)
2415 if new_info.get('http_headers') is None:
2416 new_info['http_headers'] = self._calc_headers(new_info)
2417 return fd.download(name, new_info, subtitle)
2418
2419 def process_info(self, info_dict):
2420 """Process a single resolved IE result."""
2421
2422 assert info_dict.get('_type', 'video') == 'video'
2423
2424 info_dict.setdefault('__postprocessors', [])
2425
2426 max_downloads = self.params.get('max_downloads')
2427 if max_downloads is not None:
2428 if self._num_downloads >= int(max_downloads):
2429 raise MaxDownloadsReached()
2430
2431 # TODO: backward compatibility, to be removed
2432 info_dict['fulltitle'] = info_dict['title']
2433
2434 if 'format' not in info_dict and 'ext' in info_dict:
2435 info_dict['format'] = info_dict['ext']
2436
2437 if self._match_entry(info_dict) is not None:
2438 return
2439
2440 self.post_extract(info_dict)
2441 self._num_downloads += 1
2442
2443 # info_dict['_filename'] needs to be set for backward compatibility
2444 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2445 temp_filename = self.prepare_filename(info_dict, 'temp')
2446 files_to_move = {}
2447
2448 # Forced printings
2449 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2450
2451 if self.params.get('simulate', False):
2452 if self.params.get('force_write_download_archive', False):
2453 self.record_download_archive(info_dict)
2454
2455 # Do nothing else if in simulate mode
2456 return
2457
2458 if full_filename is None:
2459 return
2460
2461 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2462 return
2463 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2464 return
2465
2466 if self.params.get('writedescription', False):
2467 descfn = self.prepare_filename(info_dict, 'description')
2468 if not self._ensure_dir_exists(encodeFilename(descfn)):
2469 return
2470 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
2471 self.to_screen('[info] Video description is already present')
2472 elif info_dict.get('description') is None:
2473 self.report_warning('There\'s no description to write.')
2474 else:
2475 try:
2476 self.to_screen('[info] Writing video description to: ' + descfn)
2477 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
2478 descfile.write(info_dict['description'])
2479 except (OSError, IOError):
2480 self.report_error('Cannot write description file ' + descfn)
2481 return
2482
2483 if self.params.get('writeannotations', False):
2484 annofn = self.prepare_filename(info_dict, 'annotation')
2485 if not self._ensure_dir_exists(encodeFilename(annofn)):
2486 return
2487 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2488 self.to_screen('[info] Video annotations are already present')
2489 elif not info_dict.get('annotations'):
2490 self.report_warning('There are no annotations to write.')
2491 else:
2492 try:
2493 self.to_screen('[info] Writing video annotations to: ' + annofn)
2494 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2495 annofile.write(info_dict['annotations'])
2496 except (KeyError, TypeError):
2497 self.report_warning('There are no annotations to write.')
2498 except (OSError, IOError):
2499 self.report_error('Cannot write annotations file: ' + annofn)
2500 return
2501
2502 subtitles_are_requested = any([self.params.get('writesubtitles', False),
2503 self.params.get('writeautomaticsub')])
2504
2505 if subtitles_are_requested and info_dict.get('requested_subtitles'):
2506 # subtitles download errors are already managed as troubles in relevant IE
2507 # that way it will silently go on when used with unsupporting IE
2508 subtitles = info_dict['requested_subtitles']
2509 # ie = self.get_info_extractor(info_dict['extractor_key'])
2510 for sub_lang, sub_info in subtitles.items():
2511 sub_format = sub_info['ext']
2512 sub_filename = subtitles_filename(temp_filename, sub_lang, sub_format, info_dict.get('ext'))
2513 sub_filename_final = subtitles_filename(
2514 self.prepare_filename(info_dict, 'subtitle'), sub_lang, sub_format, info_dict.get('ext'))
2515 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(sub_filename)):
2516 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
2517 sub_info['filepath'] = sub_filename
2518 files_to_move[sub_filename] = sub_filename_final
2519 else:
2520 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
2521 if sub_info.get('data') is not None:
2522 try:
2523 # Use newline='' to prevent conversion of newline characters
2524 # See https://github.com/ytdl-org/youtube-dl/issues/10268
2525 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
2526 subfile.write(sub_info['data'])
2527 sub_info['filepath'] = sub_filename
2528 files_to_move[sub_filename] = sub_filename_final
2529 except (OSError, IOError):
2530 self.report_error('Cannot write subtitles file ' + sub_filename)
2531 return
2532 else:
2533 try:
2534 self.dl(sub_filename, sub_info.copy(), subtitle=True)
2535 sub_info['filepath'] = sub_filename
2536 files_to_move[sub_filename] = sub_filename_final
2537 except (ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
2538 self.report_warning('Unable to download subtitle for "%s": %s' %
2539 (sub_lang, error_to_compat_str(err)))
2540 continue
2541
2542 if self.params.get('writeinfojson', False):
2543 infofn = self.prepare_filename(info_dict, 'infojson')
2544 if not self._ensure_dir_exists(encodeFilename(infofn)):
2545 return
2546 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
2547 self.to_screen('[info] Video metadata is already present')
2548 else:
2549 self.to_screen('[info] Writing video metadata as JSON to: ' + infofn)
2550 try:
2551 write_json_file(self.filter_requested_info(info_dict, self.params.get('clean_infojson', True)), infofn)
2552 except (OSError, IOError):
2553 self.report_error('Cannot write video metadata to JSON file ' + infofn)
2554 return
2555 info_dict['__infojson_filename'] = infofn
2556
2557 for thumb_ext in self._write_thumbnails(info_dict, temp_filename):
2558 thumb_filename_temp = replace_extension(temp_filename, thumb_ext, info_dict.get('ext'))
2559 thumb_filename = replace_extension(
2560 self.prepare_filename(info_dict, 'thumbnail'), thumb_ext, info_dict.get('ext'))
2561 files_to_move[thumb_filename_temp] = thumb_filename
2562
2563 # Write internet shortcut files
2564 url_link = webloc_link = desktop_link = False
2565 if self.params.get('writelink', False):
2566 if sys.platform == "darwin": # macOS.
2567 webloc_link = True
2568 elif sys.platform.startswith("linux"):
2569 desktop_link = True
2570 else: # if sys.platform in ['win32', 'cygwin']:
2571 url_link = True
2572 if self.params.get('writeurllink', False):
2573 url_link = True
2574 if self.params.get('writewebloclink', False):
2575 webloc_link = True
2576 if self.params.get('writedesktoplink', False):
2577 desktop_link = True
2578
2579 if url_link or webloc_link or desktop_link:
2580 if 'webpage_url' not in info_dict:
2581 self.report_error('Cannot write internet shortcut file because the "webpage_url" field is missing in the media information')
2582 return
2583 ascii_url = iri_to_uri(info_dict['webpage_url'])
2584
2585 def _write_link_file(extension, template, newline, embed_filename):
2586 linkfn = replace_extension(full_filename, extension, info_dict.get('ext'))
2587 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2588 self.to_screen('[info] Internet shortcut is already present')
2589 else:
2590 try:
2591 self.to_screen('[info] Writing internet shortcut to: ' + linkfn)
2592 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', newline=newline) as linkfile:
2593 template_vars = {'url': ascii_url}
2594 if embed_filename:
2595 template_vars['filename'] = linkfn[:-(len(extension) + 1)]
2596 linkfile.write(template % template_vars)
2597 except (OSError, IOError):
2598 self.report_error('Cannot write internet shortcut ' + linkfn)
2599 return False
2600 return True
2601
2602 if url_link:
2603 if not _write_link_file('url', DOT_URL_LINK_TEMPLATE, '\r\n', embed_filename=False):
2604 return
2605 if webloc_link:
2606 if not _write_link_file('webloc', DOT_WEBLOC_LINK_TEMPLATE, '\n', embed_filename=False):
2607 return
2608 if desktop_link:
2609 if not _write_link_file('desktop', DOT_DESKTOP_LINK_TEMPLATE, '\n', embed_filename=True):
2610 return
2611
2612 try:
2613 info_dict, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2614 except PostProcessingError as err:
2615 self.report_error('Preprocessing: %s' % str(err))
2616 return
2617
2618 must_record_download_archive = False
2619 if self.params.get('skip_download', False):
2620 info_dict['filepath'] = temp_filename
2621 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2622 info_dict['__files_to_move'] = files_to_move
2623 info_dict = self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict)
2624 else:
2625 # Download
2626 try:
2627
2628 def existing_file(*filepaths):
2629 ext = info_dict.get('ext')
2630 final_ext = self.params.get('final_ext', ext)
2631 existing_files = []
2632 for file in orderedSet(filepaths):
2633 if final_ext != ext:
2634 converted = replace_extension(file, final_ext, ext)
2635 if os.path.exists(encodeFilename(converted)):
2636 existing_files.append(converted)
2637 if os.path.exists(encodeFilename(file)):
2638 existing_files.append(file)
2639
2640 if not existing_files or self.params.get('overwrites', False):
2641 for file in orderedSet(existing_files):
2642 self.report_file_delete(file)
2643 os.remove(encodeFilename(file))
2644 return None
2645
2646 self.report_file_already_downloaded(existing_files[0])
2647 info_dict['ext'] = os.path.splitext(existing_files[0])[1][1:]
2648 return existing_files[0]
2649
2650 success = True
2651 if info_dict.get('requested_formats') is not None:
2652
2653 def compatible_formats(formats):
2654 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2655 video_formats = [format for format in formats if format.get('vcodec') != 'none']
2656 audio_formats = [format for format in formats if format.get('acodec') != 'none']
2657 if len(video_formats) > 2 or len(audio_formats) > 2:
2658 return False
2659
2660 # Check extension
2661 exts = set(format.get('ext') for format in formats)
2662 COMPATIBLE_EXTS = (
2663 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
2664 set(('webm',)),
2665 )
2666 for ext_sets in COMPATIBLE_EXTS:
2667 if ext_sets.issuperset(exts):
2668 return True
2669 # TODO: Check acodec/vcodec
2670 return False
2671
2672 requested_formats = info_dict['requested_formats']
2673 old_ext = info_dict['ext']
2674 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
2675 info_dict['ext'] = 'mkv'
2676 self.report_warning(
2677 'Requested formats are incompatible for merge and will be merged into mkv.')
2678
2679 def correct_ext(filename):
2680 if filename == '-':
2681 return filename
2682 filename_real_ext = os.path.splitext(filename)[1][1:]
2683 filename_wo_ext = (
2684 os.path.splitext(filename)[0]
2685 if filename_real_ext == old_ext
2686 else filename)
2687 return '%s.%s' % (filename_wo_ext, info_dict['ext'])
2688
2689 # Ensure filename always has a correct extension for successful merge
2690 full_filename = correct_ext(full_filename)
2691 temp_filename = correct_ext(temp_filename)
2692 dl_filename = existing_file(full_filename, temp_filename)
2693 info_dict['__real_download'] = False
2694
2695 _protocols = set(determine_protocol(f) for f in requested_formats)
2696 if len(_protocols) == 1: # All requested formats have same protocol
2697 info_dict['protocol'] = _protocols.pop()
2698 directly_mergable = FFmpegFD.can_merge_formats(info_dict)
2699 if dl_filename is not None:
2700 pass
2701 elif (directly_mergable and get_suitable_downloader(
2702 info_dict, self.params, to_stdout=(temp_filename== '-')) == FFmpegFD):
2703 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
2704 success, real_download = self.dl(temp_filename, info_dict)
2705 info_dict['__real_download'] = real_download
2706 else:
2707 downloaded = []
2708 merger = FFmpegMergerPP(self)
2709 if self.params.get('allow_unplayable_formats'):
2710 self.report_warning(
2711 'You have requested merging of multiple formats '
2712 'while also allowing unplayable formats to be downloaded. '
2713 'The formats won\'t be merged to prevent data corruption.')
2714 elif not merger.available:
2715 self.report_warning(
2716 'You have requested merging of multiple formats but ffmpeg is not installed. '
2717 'The formats won\'t be merged.')
2718
2719 if temp_filename == '-':
2720 reason = ('using a downloader other than ffmpeg' if directly_mergable
2721 else 'but the formats are incompatible for simultaneous download' if merger.available
2722 else 'but ffmpeg is not installed')
2723 self.report_warning(
2724 f'You have requested downloading multiple formats to stdout {reason}. '
2725 'The formats will be streamed one after the other')
2726 fname = temp_filename
2727 for f in requested_formats:
2728 new_info = dict(info_dict)
2729 del new_info['requested_formats']
2730 new_info.update(f)
2731 if temp_filename != '-':
2732 fname = prepend_extension(temp_filename, 'f%s' % f['format_id'], new_info['ext'])
2733 if not self._ensure_dir_exists(fname):
2734 return
2735 downloaded.append(fname)
2736 partial_success, real_download = self.dl(fname, new_info)
2737 info_dict['__real_download'] = info_dict['__real_download'] or real_download
2738 success = success and partial_success
2739 if merger.available and not self.params.get('allow_unplayable_formats'):
2740 info_dict['__postprocessors'].append(merger)
2741 info_dict['__files_to_merge'] = downloaded
2742 # Even if there were no downloads, it is being merged only now
2743 info_dict['__real_download'] = True
2744 else:
2745 for file in downloaded:
2746 files_to_move[file] = None
2747 else:
2748 # Just a single file
2749 dl_filename = existing_file(full_filename, temp_filename)
2750 if dl_filename is None:
2751 success, real_download = self.dl(temp_filename, info_dict)
2752 info_dict['__real_download'] = real_download
2753
2754 dl_filename = dl_filename or temp_filename
2755 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2756
2757 except network_exceptions as err:
2758 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
2759 return
2760 except (OSError, IOError) as err:
2761 raise UnavailableVideoError(err)
2762 except (ContentTooShortError, ) as err:
2763 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
2764 return
2765
2766 if success and full_filename != '-':
2767
2768 def fixup():
2769 do_fixup = True
2770 fixup_policy = self.params.get('fixup')
2771 vid = info_dict['id']
2772
2773 if fixup_policy in ('ignore', 'never'):
2774 return
2775 elif fixup_policy == 'warn':
2776 do_fixup = False
2777 elif fixup_policy != 'force':
2778 assert fixup_policy in ('detect_or_warn', None)
2779 if not info_dict.get('__real_download'):
2780 do_fixup = False
2781
2782 def ffmpeg_fixup(cndn, msg, cls):
2783 if not cndn:
2784 return
2785 if not do_fixup:
2786 self.report_warning(f'{vid}: {msg}')
2787 return
2788 pp = cls(self)
2789 if pp.available:
2790 info_dict['__postprocessors'].append(pp)
2791 else:
2792 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
2793
2794 stretched_ratio = info_dict.get('stretched_ratio')
2795 ffmpeg_fixup(
2796 stretched_ratio not in (1, None),
2797 f'Non-uniform pixel ratio {stretched_ratio}',
2798 FFmpegFixupStretchedPP)
2799
2800 ffmpeg_fixup(
2801 (info_dict.get('requested_formats') is None
2802 and info_dict.get('container') == 'm4a_dash'
2803 and info_dict.get('ext') == 'm4a'),
2804 'writing DASH m4a. Only some players support this container',
2805 FFmpegFixupM4aPP)
2806
2807 downloader = (get_suitable_downloader(info_dict, self.params).__name__
2808 if 'protocol' in info_dict else None)
2809 ffmpeg_fixup(downloader == 'HlsFD', 'malformed AAC bitstream detected', FFmpegFixupM3u8PP)
2810 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'malformed timestamps detected', FFmpegFixupTimestampPP)
2811 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'malformed duration detected', FFmpegFixupDurationPP)
2812
2813 fixup()
2814 try:
2815 info_dict = self.post_process(dl_filename, info_dict, files_to_move)
2816 except PostProcessingError as err:
2817 self.report_error('Postprocessing: %s' % str(err))
2818 return
2819 try:
2820 for ph in self._post_hooks:
2821 ph(info_dict['filepath'])
2822 except Exception as err:
2823 self.report_error('post hooks: %s' % str(err))
2824 return
2825 must_record_download_archive = True
2826
2827 if must_record_download_archive or self.params.get('force_write_download_archive', False):
2828 self.record_download_archive(info_dict)
2829 max_downloads = self.params.get('max_downloads')
2830 if max_downloads is not None and self._num_downloads >= int(max_downloads):
2831 raise MaxDownloadsReached()
2832
2833 def download(self, url_list):
2834 """Download a given list of URLs."""
2835 outtmpl = self.outtmpl_dict['default']
2836 if (len(url_list) > 1
2837 and outtmpl != '-'
2838 and '%' not in outtmpl
2839 and self.params.get('max_downloads') != 1):
2840 raise SameFileError(outtmpl)
2841
2842 for url in url_list:
2843 try:
2844 # It also downloads the videos
2845 res = self.extract_info(
2846 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2847 except UnavailableVideoError:
2848 self.report_error('unable to download video')
2849 except MaxDownloadsReached:
2850 self.to_screen('[info] Maximum number of downloaded files reached')
2851 raise
2852 except ExistingVideoReached:
2853 self.to_screen('[info] Encountered a file that is already in the archive, stopping due to --break-on-existing')
2854 raise
2855 except RejectedVideoReached:
2856 self.to_screen('[info] Encountered a file that did not match filter, stopping due to --break-on-reject')
2857 raise
2858 else:
2859 if self.params.get('dump_single_json', False):
2860 self.post_extract(res)
2861 self.to_stdout(json.dumps(res, default=repr))
2862
2863 return self._download_retcode
2864
2865 def download_with_info_file(self, info_filename):
2866 with contextlib.closing(fileinput.FileInput(
2867 [info_filename], mode='r',
2868 openhook=fileinput.hook_encoded('utf-8'))) as f:
2869 # FileInput doesn't have a read method, we can't call json.load
2870 info = self.filter_requested_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
2871 try:
2872 self.process_ie_result(info, download=True)
2873 except (DownloadError, EntryNotInPlaylist, ThrottledDownload):
2874 webpage_url = info.get('webpage_url')
2875 if webpage_url is not None:
2876 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2877 return self.download([webpage_url])
2878 else:
2879 raise
2880 return self._download_retcode
2881
2882 @staticmethod
2883 def filter_requested_info(info_dict, actually_filter=True):
2884 remove_keys = ['__original_infodict'] # Always remove this since this may contain a copy of the entire dict
2885 keep_keys = ['_type'], # Always keep this to facilitate load-info-json
2886 if actually_filter:
2887 remove_keys += ('requested_formats', 'requested_subtitles', 'requested_entries', 'filepath', 'entries', 'original_url')
2888 empty_values = (None, {}, [], set(), tuple())
2889 reject = lambda k, v: k not in keep_keys and (
2890 k.startswith('_') or k in remove_keys or v in empty_values)
2891 else:
2892 info_dict['epoch'] = int(time.time())
2893 reject = lambda k, v: k in remove_keys
2894 filter_fn = lambda obj: (
2895 list(map(filter_fn, obj)) if isinstance(obj, (LazyList, list, tuple, set))
2896 else obj if not isinstance(obj, dict)
2897 else dict((k, filter_fn(v)) for k, v in obj.items() if not reject(k, v)))
2898 return filter_fn(info_dict)
2899
2900 def run_pp(self, pp, infodict):
2901 files_to_delete = []
2902 if '__files_to_move' not in infodict:
2903 infodict['__files_to_move'] = {}
2904 files_to_delete, infodict = pp.run(infodict)
2905 if not files_to_delete:
2906 return infodict
2907
2908 if self.params.get('keepvideo', False):
2909 for f in files_to_delete:
2910 infodict['__files_to_move'].setdefault(f, '')
2911 else:
2912 for old_filename in set(files_to_delete):
2913 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2914 try:
2915 os.remove(encodeFilename(old_filename))
2916 except (IOError, OSError):
2917 self.report_warning('Unable to remove downloaded original file')
2918 if old_filename in infodict['__files_to_move']:
2919 del infodict['__files_to_move'][old_filename]
2920 return infodict
2921
2922 @staticmethod
2923 def post_extract(info_dict):
2924 def actual_post_extract(info_dict):
2925 if info_dict.get('_type') in ('playlist', 'multi_video'):
2926 for video_dict in info_dict.get('entries', {}):
2927 actual_post_extract(video_dict or {})
2928 return
2929
2930 post_extractor = info_dict.get('__post_extractor') or (lambda: {})
2931 extra = post_extractor().items()
2932 info_dict.update(extra)
2933 info_dict.pop('__post_extractor', None)
2934
2935 original_infodict = info_dict.get('__original_infodict') or {}
2936 original_infodict.update(extra)
2937 original_infodict.pop('__post_extractor', None)
2938
2939 actual_post_extract(info_dict or {})
2940
2941 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
2942 info = dict(ie_info)
2943 info['__files_to_move'] = files_to_move or {}
2944 for pp in self._pps[key]:
2945 info = self.run_pp(pp, info)
2946 return info, info.pop('__files_to_move', None)
2947
2948 def post_process(self, filename, ie_info, files_to_move=None):
2949 """Run all the postprocessors on the given file."""
2950 info = dict(ie_info)
2951 info['filepath'] = filename
2952 info['__files_to_move'] = files_to_move or {}
2953
2954 for pp in ie_info.get('__postprocessors', []) + self._pps['post_process']:
2955 info = self.run_pp(pp, info)
2956 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
2957 del info['__files_to_move']
2958 for pp in self._pps['after_move']:
2959 info = self.run_pp(pp, info)
2960 return info
2961
2962 def _make_archive_id(self, info_dict):
2963 video_id = info_dict.get('id')
2964 if not video_id:
2965 return
2966 # Future-proof against any change in case
2967 # and backwards compatibility with prior versions
2968 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
2969 if extractor is None:
2970 url = str_or_none(info_dict.get('url'))
2971 if not url:
2972 return
2973 # Try to find matching extractor for the URL and take its ie_key
2974 for ie in self._ies:
2975 if ie.suitable(url):
2976 extractor = ie.ie_key()
2977 break
2978 else:
2979 return
2980 return '%s %s' % (extractor.lower(), video_id)
2981
2982 def in_download_archive(self, info_dict):
2983 fn = self.params.get('download_archive')
2984 if fn is None:
2985 return False
2986
2987 vid_id = self._make_archive_id(info_dict)
2988 if not vid_id:
2989 return False # Incomplete video information
2990
2991 return vid_id in self.archive
2992
2993 def record_download_archive(self, info_dict):
2994 fn = self.params.get('download_archive')
2995 if fn is None:
2996 return
2997 vid_id = self._make_archive_id(info_dict)
2998 assert vid_id
2999 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3000 archive_file.write(vid_id + '\n')
3001 self.archive.add(vid_id)
3002
3003 @staticmethod
3004 def format_resolution(format, default='unknown'):
3005 if format.get('vcodec') == 'none':
3006 if format.get('acodec') == 'none':
3007 return 'images'
3008 return 'audio only'
3009 if format.get('resolution') is not None:
3010 return format['resolution']
3011 if format.get('width') and format.get('height'):
3012 res = '%dx%d' % (format['width'], format['height'])
3013 elif format.get('height'):
3014 res = '%sp' % format['height']
3015 elif format.get('width'):
3016 res = '%dx?' % format['width']
3017 else:
3018 res = default
3019 return res
3020
3021 def _format_note(self, fdict):
3022 res = ''
3023 if fdict.get('ext') in ['f4f', 'f4m']:
3024 res += '(unsupported) '
3025 if fdict.get('language'):
3026 if res:
3027 res += ' '
3028 res += '[%s] ' % fdict['language']
3029 if fdict.get('format_note') is not None:
3030 res += fdict['format_note'] + ' '
3031 if fdict.get('tbr') is not None:
3032 res += '%4dk ' % fdict['tbr']
3033 if fdict.get('container') is not None:
3034 if res:
3035 res += ', '
3036 res += '%s container' % fdict['container']
3037 if (fdict.get('vcodec') is not None
3038 and fdict.get('vcodec') != 'none'):
3039 if res:
3040 res += ', '
3041 res += fdict['vcodec']
3042 if fdict.get('vbr') is not None:
3043 res += '@'
3044 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3045 res += 'video@'
3046 if fdict.get('vbr') is not None:
3047 res += '%4dk' % fdict['vbr']
3048 if fdict.get('fps') is not None:
3049 if res:
3050 res += ', '
3051 res += '%sfps' % fdict['fps']
3052 if fdict.get('acodec') is not None:
3053 if res:
3054 res += ', '
3055 if fdict['acodec'] == 'none':
3056 res += 'video only'
3057 else:
3058 res += '%-5s' % fdict['acodec']
3059 elif fdict.get('abr') is not None:
3060 if res:
3061 res += ', '
3062 res += 'audio'
3063 if fdict.get('abr') is not None:
3064 res += '@%3dk' % fdict['abr']
3065 if fdict.get('asr') is not None:
3066 res += ' (%5dHz)' % fdict['asr']
3067 if fdict.get('filesize') is not None:
3068 if res:
3069 res += ', '
3070 res += format_bytes(fdict['filesize'])
3071 elif fdict.get('filesize_approx') is not None:
3072 if res:
3073 res += ', '
3074 res += '~' + format_bytes(fdict['filesize_approx'])
3075 return res
3076
3077 def list_formats(self, info_dict):
3078 formats = info_dict.get('formats', [info_dict])
3079 new_format = (
3080 'list-formats' not in self.params.get('compat_opts', [])
3081 and self.params.get('listformats_table', True) is not False)
3082 if new_format:
3083 table = [
3084 [
3085 format_field(f, 'format_id'),
3086 format_field(f, 'ext'),
3087 self.format_resolution(f),
3088 format_field(f, 'fps', '%d'),
3089 '|',
3090 format_field(f, 'filesize', ' %s', func=format_bytes) + format_field(f, 'filesize_approx', '~%s', func=format_bytes),
3091 format_field(f, 'tbr', '%4dk'),
3092 shorten_protocol_name(f.get('protocol', '').replace("native", "n")),
3093 '|',
3094 format_field(f, 'vcodec', default='unknown').replace('none', ''),
3095 format_field(f, 'vbr', '%4dk'),
3096 format_field(f, 'acodec', default='unknown').replace('none', ''),
3097 format_field(f, 'abr', '%3dk'),
3098 format_field(f, 'asr', '%5dHz'),
3099 ', '.join(filter(None, (
3100 'UNSUPPORTED' if f.get('ext') in ('f4f', 'f4m') else '',
3101 format_field(f, 'language', '[%s]'),
3102 format_field(f, 'format_note'),
3103 format_field(f, 'container', ignore=(None, f.get('ext'))),
3104 ))),
3105 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3106 header_line = ['ID', 'EXT', 'RESOLUTION', 'FPS', '|', ' FILESIZE', ' TBR', 'PROTO',
3107 '|', 'VCODEC', ' VBR', 'ACODEC', ' ABR', ' ASR', 'MORE INFO']
3108 else:
3109 table = [
3110 [
3111 format_field(f, 'format_id'),
3112 format_field(f, 'ext'),
3113 self.format_resolution(f),
3114 self._format_note(f)]
3115 for f in formats
3116 if f.get('preference') is None or f['preference'] >= -1000]
3117 header_line = ['format code', 'extension', 'resolution', 'note']
3118
3119 self.to_screen(
3120 '[info] Available formats for %s:' % info_dict['id'])
3121 self.to_stdout(render_table(
3122 header_line, table, delim=new_format, extraGap=(0 if new_format else 1), hideEmpty=new_format))
3123
3124 def list_thumbnails(self, info_dict):
3125 thumbnails = list(info_dict.get('thumbnails'))
3126 if not thumbnails:
3127 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
3128 return
3129
3130 self.to_screen(
3131 '[info] Thumbnails for %s:' % info_dict['id'])
3132 self.to_stdout(render_table(
3133 ['ID', 'width', 'height', 'URL'],
3134 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
3135
3136 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3137 if not subtitles:
3138 self.to_screen('%s has no %s' % (video_id, name))
3139 return
3140 self.to_screen(
3141 'Available %s for %s:' % (name, video_id))
3142
3143 def _row(lang, formats):
3144 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3145 if len(set(names)) == 1:
3146 names = [] if names[0] == 'unknown' else names[:1]
3147 return [lang, ', '.join(names), ', '.join(exts)]
3148
3149 self.to_stdout(render_table(
3150 ['Language', 'Name', 'Formats'],
3151 [_row(lang, formats) for lang, formats in subtitles.items()],
3152 hideEmpty=True))
3153
3154 def urlopen(self, req):
3155 """ Start an HTTP download """
3156 if isinstance(req, compat_basestring):
3157 req = sanitized_Request(req)
3158 return self._opener.open(req, timeout=self._socket_timeout)
3159
3160 def print_debug_header(self):
3161 if not self.params.get('verbose'):
3162 return
3163
3164 if type('') is not compat_str:
3165 # Python 2.6 on SLES11 SP1 (https://github.com/ytdl-org/youtube-dl/issues/3326)
3166 self.report_warning(
3167 'Your Python is broken! Update to a newer and supported version')
3168
3169 stdout_encoding = getattr(
3170 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
3171 encoding_str = (
3172 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
3173 locale.getpreferredencoding(),
3174 sys.getfilesystemencoding(),
3175 stdout_encoding,
3176 self.get_encoding()))
3177 write_string(encoding_str, encoding=None)
3178
3179 source = (
3180 '(exe)' if hasattr(sys, 'frozen')
3181 else '(zip)' if isinstance(globals().get('__loader__'), zipimporter)
3182 else '(source)' if os.path.basename(sys.argv[0]) == '__main__.py'
3183 else '')
3184 self._write_string('[debug] yt-dlp version %s %s\n' % (__version__, source))
3185 if _LAZY_LOADER:
3186 self._write_string('[debug] Lazy loading extractors enabled\n')
3187 if _PLUGIN_CLASSES:
3188 self._write_string(
3189 '[debug] Plugin Extractors: %s\n' % [ie.ie_key() for ie in _PLUGIN_CLASSES])
3190 if self.params.get('compat_opts'):
3191 self._write_string(
3192 '[debug] Compatibility options: %s\n' % ', '.join(self.params.get('compat_opts')))
3193 try:
3194 sp = subprocess.Popen(
3195 ['git', 'rev-parse', '--short', 'HEAD'],
3196 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3197 cwd=os.path.dirname(os.path.abspath(__file__)))
3198 out, err = process_communicate_or_kill(sp)
3199 out = out.decode().strip()
3200 if re.match('[0-9a-f]+', out):
3201 self._write_string('[debug] Git HEAD: %s\n' % out)
3202 except Exception:
3203 try:
3204 sys.exc_clear()
3205 except Exception:
3206 pass
3207
3208 def python_implementation():
3209 impl_name = platform.python_implementation()
3210 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3211 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3212 return impl_name
3213
3214 self._write_string('[debug] Python version %s (%s %s) - %s\n' % (
3215 platform.python_version(),
3216 python_implementation(),
3217 platform.architecture()[0],
3218 platform_name()))
3219
3220 exe_versions = FFmpegPostProcessor.get_versions(self)
3221 exe_versions['rtmpdump'] = rtmpdump_version()
3222 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3223 exe_str = ', '.join(
3224 '%s %s' % (exe, v)
3225 for exe, v in sorted(exe_versions.items())
3226 if v
3227 )
3228 if not exe_str:
3229 exe_str = 'none'
3230 self._write_string('[debug] exe versions: %s\n' % exe_str)
3231
3232 proxy_map = {}
3233 for handler in self._opener.handlers:
3234 if hasattr(handler, 'proxies'):
3235 proxy_map.update(handler.proxies)
3236 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
3237
3238 if self.params.get('call_home', False):
3239 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3240 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
3241 return
3242 latest_version = self.urlopen(
3243 'https://yt-dl.org/latest/version').read().decode('utf-8')
3244 if version_tuple(latest_version) > version_tuple(__version__):
3245 self.report_warning(
3246 'You are using an outdated version (newest version: %s)! '
3247 'See https://yt-dl.org/update if you need help updating.' %
3248 latest_version)
3249
3250 def _setup_opener(self):
3251 timeout_val = self.params.get('socket_timeout')
3252 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
3253
3254 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3255 opts_cookiefile = self.params.get('cookiefile')
3256 opts_proxy = self.params.get('proxy')
3257
3258 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3259
3260 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3261 if opts_proxy is not None:
3262 if opts_proxy == '':
3263 proxies = {}
3264 else:
3265 proxies = {'http': opts_proxy, 'https': opts_proxy}
3266 else:
3267 proxies = compat_urllib_request.getproxies()
3268 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3269 if 'http' in proxies and 'https' not in proxies:
3270 proxies['https'] = proxies['http']
3271 proxy_handler = PerRequestProxyHandler(proxies)
3272
3273 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3274 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3275 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3276 redirect_handler = YoutubeDLRedirectHandler()
3277 data_handler = compat_urllib_request_DataHandler()
3278
3279 # When passing our own FileHandler instance, build_opener won't add the
3280 # default FileHandler and allows us to disable the file protocol, which
3281 # can be used for malicious purposes (see
3282 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3283 file_handler = compat_urllib_request.FileHandler()
3284
3285 def file_open(*args, **kwargs):
3286 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3287 file_handler.file_open = file_open
3288
3289 opener = compat_urllib_request.build_opener(
3290 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3291
3292 # Delete the default user-agent header, which would otherwise apply in
3293 # cases where our custom HTTP handler doesn't come into play
3294 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3295 opener.addheaders = []
3296 self._opener = opener
3297
3298 def encode(self, s):
3299 if isinstance(s, bytes):
3300 return s # Already encoded
3301
3302 try:
3303 return s.encode(self.get_encoding())
3304 except UnicodeEncodeError as err:
3305 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3306 raise
3307
3308 def get_encoding(self):
3309 encoding = self.params.get('encoding')
3310 if encoding is None:
3311 encoding = preferredencoding()
3312 return encoding
3313
3314 def _write_thumbnails(self, info_dict, filename): # return the extensions
3315 write_all = self.params.get('write_all_thumbnails', False)
3316 thumbnails = []
3317 if write_all or self.params.get('writethumbnail', False):
3318 thumbnails = info_dict.get('thumbnails') or []
3319 multiple = write_all and len(thumbnails) > 1
3320
3321 ret = []
3322 for t in thumbnails[::-1]:
3323 thumb_ext = determine_ext(t['url'], 'jpg')
3324 suffix = '%s.' % t['id'] if multiple else ''
3325 thumb_display_id = '%s ' % t['id'] if multiple else ''
3326 thumb_filename = replace_extension(filename, suffix + thumb_ext, info_dict.get('ext'))
3327
3328 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(thumb_filename)):
3329 ret.append(suffix + thumb_ext)
3330 t['filepath'] = thumb_filename
3331 self.to_screen('[%s] %s: Thumbnail %sis already present' %
3332 (info_dict['extractor'], info_dict['id'], thumb_display_id))
3333 else:
3334 self.to_screen('[%s] %s: Downloading thumbnail %s ...' %
3335 (info_dict['extractor'], info_dict['id'], thumb_display_id))
3336 try:
3337 uf = self.urlopen(t['url'])
3338 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3339 shutil.copyfileobj(uf, thumbf)
3340 ret.append(suffix + thumb_ext)
3341 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
3342 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
3343 t['filepath'] = thumb_filename
3344 except network_exceptions as err:
3345 self.report_warning('Unable to download thumbnail "%s": %s' %
3346 (t['url'], error_to_compat_str(err)))
3347 if ret and not write_all:
3348 break
3349 return ret