]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[cleanup] Refactor updater
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 #!/usr/bin/env python
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 time
24 import tokenize
25 import traceback
26 import random
27
28 from string import ascii_letters
29 from zipimport import zipimporter
30
31 from .compat import (
32 compat_basestring,
33 compat_cookiejar,
34 compat_get_terminal_size,
35 compat_kwargs,
36 compat_numeric_types,
37 compat_os_name,
38 compat_str,
39 compat_tokenize_tokenize,
40 compat_urllib_error,
41 compat_urllib_request,
42 compat_urllib_request_DataHandler,
43 )
44 from .utils import (
45 age_restricted,
46 args_to_str,
47 ContentTooShortError,
48 date_from_str,
49 DateRange,
50 DEFAULT_OUTTMPL,
51 determine_ext,
52 determine_protocol,
53 DOT_DESKTOP_LINK_TEMPLATE,
54 DOT_URL_LINK_TEMPLATE,
55 DOT_WEBLOC_LINK_TEMPLATE,
56 DownloadError,
57 encode_compat_str,
58 encodeFilename,
59 EntryNotInPlaylist,
60 error_to_compat_str,
61 ExistingVideoReached,
62 expand_path,
63 ExtractorError,
64 float_or_none,
65 format_bytes,
66 format_field,
67 FORMAT_RE,
68 formatSeconds,
69 GeoRestrictedError,
70 int_or_none,
71 iri_to_uri,
72 ISO3166Utils,
73 locked_file,
74 make_dir,
75 make_HTTPS_handler,
76 MaxDownloadsReached,
77 network_exceptions,
78 orderedSet,
79 OUTTMPL_TYPES,
80 PagedList,
81 parse_filesize,
82 PerRequestProxyHandler,
83 platform_name,
84 PostProcessingError,
85 preferredencoding,
86 prepend_extension,
87 process_communicate_or_kill,
88 random_uuidv4,
89 register_socks_protocols,
90 RejectedVideoReached,
91 render_table,
92 replace_extension,
93 SameFileError,
94 sanitize_filename,
95 sanitize_path,
96 sanitize_url,
97 sanitized_Request,
98 std_headers,
99 str_or_none,
100 strftime_or_none,
101 subtitles_filename,
102 to_high_limit_path,
103 traverse_dict,
104 UnavailableVideoError,
105 url_basename,
106 version_tuple,
107 write_json_file,
108 write_string,
109 YoutubeDLCookieJar,
110 YoutubeDLCookieProcessor,
111 YoutubeDLHandler,
112 YoutubeDLRedirectHandler,
113 )
114 from .cache import Cache
115 from .extractor import (
116 gen_extractor_classes,
117 get_info_extractor,
118 _LAZY_LOADER,
119 _PLUGIN_CLASSES
120 )
121 from .extractor.openload import PhantomJSwrapper
122 from .downloader import (
123 get_suitable_downloader,
124 shorten_protocol_name
125 )
126 from .downloader.rtmp import rtmpdump_version
127 from .postprocessor import (
128 FFmpegFixupM3u8PP,
129 FFmpegFixupM4aPP,
130 FFmpegFixupStretchedPP,
131 FFmpegMergerPP,
132 FFmpegPostProcessor,
133 # FFmpegSubtitlesConvertorPP,
134 get_postprocessor,
135 MoveFilesAfterDownloadPP,
136 )
137 from .version import __version__
138
139 if compat_os_name == 'nt':
140 import ctypes
141
142
143 class YoutubeDL(object):
144 """YoutubeDL class.
145
146 YoutubeDL objects are the ones responsible of downloading the
147 actual video file and writing it to disk if the user has requested
148 it, among some other tasks. In most cases there should be one per
149 program. As, given a video URL, the downloader doesn't know how to
150 extract all the needed information, task that InfoExtractors do, it
151 has to pass the URL to one of them.
152
153 For this, YoutubeDL objects have a method that allows
154 InfoExtractors to be registered in a given order. When it is passed
155 a URL, the YoutubeDL object handles it to the first InfoExtractor it
156 finds that reports being able to handle it. The InfoExtractor extracts
157 all the information about the video or videos the URL refers to, and
158 YoutubeDL process the extracted information, possibly using a File
159 Downloader to download the video.
160
161 YoutubeDL objects accept a lot of parameters. In order not to saturate
162 the object constructor with arguments, it receives a dictionary of
163 options instead. These options are available through the params
164 attribute for the InfoExtractors to use. The YoutubeDL also
165 registers itself as the downloader in charge for the InfoExtractors
166 that are added to it, so this is a "mutual registration".
167
168 Available options:
169
170 username: Username for authentication purposes.
171 password: Password for authentication purposes.
172 videopassword: Password for accessing a video.
173 ap_mso: Adobe Pass multiple-system operator identifier.
174 ap_username: Multiple-system operator account username.
175 ap_password: Multiple-system operator account password.
176 usenetrc: Use netrc for authentication instead.
177 verbose: Print additional info to stdout.
178 quiet: Do not print messages to stdout.
179 no_warnings: Do not print out anything for warnings.
180 forceprint: A list of templates to force print
181 forceurl: Force printing final URL. (Deprecated)
182 forcetitle: Force printing title. (Deprecated)
183 forceid: Force printing ID. (Deprecated)
184 forcethumbnail: Force printing thumbnail URL. (Deprecated)
185 forcedescription: Force printing description. (Deprecated)
186 forcefilename: Force printing final filename. (Deprecated)
187 forceduration: Force printing duration. (Deprecated)
188 forcejson: Force printing info_dict as JSON.
189 dump_single_json: Force printing the info_dict of the whole playlist
190 (or video) as a single JSON line.
191 force_write_download_archive: Force writing download archive regardless
192 of 'skip_download' or 'simulate'.
193 simulate: Do not download the video files.
194 format: Video format code. see "FORMAT SELECTION" for more details.
195 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
196 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
197 extracting metadata even if the video is not actually
198 available for download (experimental)
199 format_sort: How to sort the video formats. see "Sorting Formats"
200 for more details.
201 format_sort_force: Force the given format_sort. see "Sorting Formats"
202 for more details.
203 allow_multiple_video_streams: Allow multiple video streams to be merged
204 into a single file
205 allow_multiple_audio_streams: Allow multiple audio streams to be merged
206 into a single file
207 paths: Dictionary of output paths. The allowed keys are 'home'
208 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
209 outtmpl: Dictionary of templates for output names. Allowed keys
210 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
211 A string a also accepted for backward compatibility
212 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
213 restrictfilenames: Do not allow "&" and spaces in file names
214 trim_file_name: Limit length of filename (extension excluded)
215 windowsfilenames: Force the filenames to be windows compatible
216 ignoreerrors: Do not stop on download errors
217 (Default True when running yt-dlp,
218 but False when directly accessing YoutubeDL class)
219 skip_playlist_after_errors: Number of allowed failures until the rest of
220 the playlist is skipped
221 force_generic_extractor: Force downloader to use the generic extractor
222 overwrites: Overwrite all video and metadata files if True,
223 overwrite only non-video files if None
224 and don't overwrite any file if False
225 playliststart: Playlist item to start at.
226 playlistend: Playlist item to end at.
227 playlist_items: Specific indices of playlist to download.
228 playlistreverse: Download playlist items in reverse order.
229 playlistrandom: Download playlist items in random order.
230 matchtitle: Download only matching titles.
231 rejecttitle: Reject downloads for matching titles.
232 logger: Log messages to a logging.Logger instance.
233 logtostderr: Log messages to stderr instead of stdout.
234 writedescription: Write the video description to a .description file
235 writeinfojson: Write the video description to a .info.json file
236 clean_infojson: Remove private fields from the infojson
237 writecomments: Extract video comments. This will not be written to disk
238 unless writeinfojson is also given
239 writeannotations: Write the video annotations to a .annotations.xml file
240 writethumbnail: Write the thumbnail image to a file
241 allow_playlist_files: Whether to write playlists' description, infojson etc
242 also to disk when using the 'write*' options
243 write_all_thumbnails: Write all thumbnail formats to files
244 writelink: Write an internet shortcut file, depending on the
245 current platform (.url/.webloc/.desktop)
246 writeurllink: Write a Windows internet shortcut file (.url)
247 writewebloclink: Write a macOS internet shortcut file (.webloc)
248 writedesktoplink: Write a Linux internet shortcut file (.desktop)
249 writesubtitles: Write the video subtitles to a file
250 writeautomaticsub: Write the automatically generated subtitles to a file
251 allsubtitles: Deprecated - Use subtitlelangs = ['all']
252 Downloads all the subtitles of the video
253 (requires writesubtitles or writeautomaticsub)
254 listsubtitles: Lists all available subtitles for the video
255 subtitlesformat: The format code for subtitles
256 subtitleslangs: List of languages of the subtitles to download (can be regex).
257 The list may contain "all" to refer to all the available
258 subtitles. The language can be prefixed with a "-" to
259 exclude it from the requested languages. Eg: ['all', '-live_chat']
260 keepvideo: Keep the video file after post-processing
261 daterange: A DateRange object, download only if the upload_date is in the range.
262 skip_download: Skip the actual download of the video file
263 cachedir: Location of the cache files in the filesystem.
264 False to disable filesystem cache.
265 noplaylist: Download single video instead of a playlist if in doubt.
266 age_limit: An integer representing the user's age in years.
267 Unsuitable videos for the given age are skipped.
268 min_views: An integer representing the minimum view count the video
269 must have in order to not be skipped.
270 Videos without view count information are always
271 downloaded. None for no limit.
272 max_views: An integer representing the maximum view count.
273 Videos that are more popular than that are not
274 downloaded.
275 Videos without view count information are always
276 downloaded. None for no limit.
277 download_archive: File name of a file where all downloads are recorded.
278 Videos already present in the file are not downloaded
279 again.
280 break_on_existing: Stop the download process after attempting to download a
281 file that is in the archive.
282 break_on_reject: Stop the download process when encountering a video that
283 has been filtered out.
284 cookiefile: File name where cookies should be read from and dumped to
285 nocheckcertificate:Do not verify SSL certificates
286 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
287 At the moment, this is only supported by YouTube.
288 proxy: URL of the proxy server to use
289 geo_verification_proxy: URL of the proxy to use for IP address verification
290 on geo-restricted sites.
291 socket_timeout: Time to wait for unresponsive hosts, in seconds
292 bidi_workaround: Work around buggy terminals without bidirectional text
293 support, using fridibi
294 debug_printtraffic:Print out sent and received HTTP traffic
295 include_ads: Download ads as well
296 default_search: Prepend this string if an input url is not valid.
297 'auto' for elaborate guessing
298 encoding: Use this encoding instead of the system-specified.
299 extract_flat: Do not resolve URLs, return the immediate result.
300 Pass in 'in_playlist' to only show this behavior for
301 playlist items.
302 postprocessors: A list of dictionaries, each with an entry
303 * key: The name of the postprocessor. See
304 yt_dlp/postprocessor/__init__.py for a list.
305 * when: When to run the postprocessor. Can be one of
306 pre_process|before_dl|post_process|after_move.
307 Assumed to be 'post_process' if not given
308 post_hooks: A list of functions that get called as the final step
309 for each video file, after all postprocessors have been
310 called. The filename will be passed as the only argument.
311 progress_hooks: A list of functions that get called on download
312 progress, with a dictionary with the entries
313 * status: One of "downloading", "error", or "finished".
314 Check this first and ignore unknown values.
315
316 If status is one of "downloading", or "finished", the
317 following properties may also be present:
318 * filename: The final filename (always present)
319 * tmpfilename: The filename we're currently writing to
320 * downloaded_bytes: Bytes on disk
321 * total_bytes: Size of the whole file, None if unknown
322 * total_bytes_estimate: Guess of the eventual file size,
323 None if unavailable.
324 * elapsed: The number of seconds since download started.
325 * eta: The estimated time in seconds, None if unknown
326 * speed: The download speed in bytes/second, None if
327 unknown
328 * fragment_index: The counter of the currently
329 downloaded video fragment.
330 * fragment_count: The number of fragments (= individual
331 files that will be merged)
332
333 Progress hooks are guaranteed to be called at least once
334 (with status "finished") if the download is successful.
335 merge_output_format: Extension to use when merging formats.
336 final_ext: Expected final extension; used to detect when the file was
337 already downloaded and converted. "merge_output_format" is
338 replaced by this extension when given
339 fixup: Automatically correct known faults of the file.
340 One of:
341 - "never": do nothing
342 - "warn": only emit a warning
343 - "detect_or_warn": check whether we can do anything
344 about it, warn otherwise (default)
345 source_address: Client-side IP address to bind to.
346 call_home: Boolean, true iff we are allowed to contact the
347 yt-dlp servers for debugging. (BROKEN)
348 sleep_interval_requests: Number of seconds to sleep between requests
349 during extraction
350 sleep_interval: Number of seconds to sleep before each download when
351 used alone or a lower bound of a range for randomized
352 sleep before each download (minimum possible number
353 of seconds to sleep) when used along with
354 max_sleep_interval.
355 max_sleep_interval:Upper bound of a range for randomized sleep before each
356 download (maximum possible number of seconds to sleep).
357 Must only be used along with sleep_interval.
358 Actual sleep time will be a random float from range
359 [sleep_interval; max_sleep_interval].
360 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
361 listformats: Print an overview of available video formats and exit.
362 list_thumbnails: Print a table of all thumbnails and exit.
363 match_filter: A function that gets called with the info_dict of
364 every video.
365 If it returns a message, the video is ignored.
366 If it returns None, the video is downloaded.
367 match_filter_func in utils.py is one example for this.
368 no_color: Do not emit color codes in output.
369 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
370 HTTP header
371 geo_bypass_country:
372 Two-letter ISO 3166-2 country code that will be used for
373 explicit geographic restriction bypassing via faking
374 X-Forwarded-For HTTP header
375 geo_bypass_ip_block:
376 IP range in CIDR notation that will be used similarly to
377 geo_bypass_country
378
379 The following options determine which downloader is picked:
380 external_downloader: A dictionary of protocol keys and the executable of the
381 external downloader to use for it. The allowed protocols
382 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
383 Set the value to 'native' to use the native downloader
384 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
385 or {'m3u8': 'ffmpeg'} instead.
386 Use the native HLS downloader instead of ffmpeg/avconv
387 if True, otherwise use ffmpeg/avconv if False, otherwise
388 use downloader suggested by extractor if None.
389 compat_opts: Compatibility options. See "Differences in default behavior".
390 Note that only format-sort, format-spec, no-live-chat,
391 no-attach-info-json, playlist-index, list-formats,
392 no-direct-merge, no-youtube-channel-redirect,
393 and no-youtube-unavailable-videos works when used via the API
394
395 The following parameters are not used by YoutubeDL itself, they are used by
396 the downloader (see yt_dlp/downloader/common.py):
397 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
398 noresizebuffer, retries, continuedl, noprogress, consoletitle,
399 xattr_set_filesize, external_downloader_args, hls_use_mpegts,
400 http_chunk_size.
401
402 The following options are used by the post processors:
403 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
404 otherwise prefer ffmpeg. (avconv support is deprecated)
405 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
406 to the binary or its containing directory.
407 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
408 and a list of additional command-line arguments for the
409 postprocessor/executable. The dict can also have "PP+EXE" keys
410 which are used when the given exe is used by the given PP.
411 Use 'default' as the name for arguments to passed to all PP
412
413 The following options are used by the extractors:
414 extractor_retries: Number of times to retry for known errors
415 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
416 hls_split_discontinuity: Split HLS playlists to different formats at
417 discontinuities such as ad breaks (default: False)
418 youtube_include_dash_manifest: If True (default), DASH manifests and related
419 data will be downloaded and processed by extractor.
420 You can reduce network I/O by disabling it if you don't
421 care about DASH. (only for youtube)
422 youtube_include_hls_manifest: If True (default), HLS manifests and related
423 data will be downloaded and processed by extractor.
424 You can reduce network I/O by disabling it if you don't
425 care about HLS. (only for youtube)
426 """
427
428 _NUMERIC_FIELDS = set((
429 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
430 'timestamp', 'upload_year', 'upload_month', 'upload_day',
431 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
432 'average_rating', 'comment_count', 'age_limit',
433 'start_time', 'end_time',
434 'chapter_number', 'season_number', 'episode_number',
435 'track_number', 'disc_number', 'release_year',
436 'playlist_index',
437 ))
438
439 params = None
440 _ies = []
441 _pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
442 __prepare_filename_warned = False
443 _first_webpage_request = True
444 _download_retcode = None
445 _num_downloads = None
446 _playlist_level = 0
447 _playlist_urls = set()
448 _screen_file = None
449
450 def __init__(self, params=None, auto_init=True):
451 """Create a FileDownloader object with the given options."""
452 if params is None:
453 params = {}
454 self._ies = []
455 self._ies_instances = {}
456 self._pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
457 self.__prepare_filename_warned = False
458 self._first_webpage_request = True
459 self._post_hooks = []
460 self._progress_hooks = []
461 self._download_retcode = 0
462 self._num_downloads = 0
463 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
464 self._err_file = sys.stderr
465 self.params = {
466 # Default parameters
467 'nocheckcertificate': False,
468 }
469 self.params.update(params)
470 self.cache = Cache(self)
471
472 if sys.version_info < (3, 6):
473 self.report_warning(
474 'Support for Python version %d.%d have been deprecated and will break in future versions of yt-dlp! '
475 'Update to Python 3.6 or above' % sys.version_info[:2])
476
477 def check_deprecated(param, option, suggestion):
478 if self.params.get(param) is not None:
479 self.report_warning('%s is deprecated. Use %s instead' % (option, suggestion))
480 return True
481 return False
482
483 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
484 if self.params.get('geo_verification_proxy') is None:
485 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
486
487 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
488 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
489 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
490
491 for msg in self.params.get('warnings', []):
492 self.report_warning(msg)
493
494 if self.params.get('final_ext'):
495 if self.params.get('merge_output_format'):
496 self.report_warning('--merge-output-format will be ignored since --remux-video or --recode-video is given')
497 self.params['merge_output_format'] = self.params['final_ext']
498
499 if 'overwrites' in self.params and self.params['overwrites'] is None:
500 del self.params['overwrites']
501
502 if params.get('bidi_workaround', False):
503 try:
504 import pty
505 master, slave = pty.openpty()
506 width = compat_get_terminal_size().columns
507 if width is None:
508 width_args = []
509 else:
510 width_args = ['-w', str(width)]
511 sp_kwargs = dict(
512 stdin=subprocess.PIPE,
513 stdout=slave,
514 stderr=self._err_file)
515 try:
516 self._output_process = subprocess.Popen(
517 ['bidiv'] + width_args, **sp_kwargs
518 )
519 except OSError:
520 self._output_process = subprocess.Popen(
521 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
522 self._output_channel = os.fdopen(master, 'rb')
523 except OSError as ose:
524 if ose.errno == errno.ENOENT:
525 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 $PATH.')
526 else:
527 raise
528
529 if (sys.platform != 'win32'
530 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
531 and not params.get('restrictfilenames', False)):
532 # Unicode filesystem API will throw errors (#1474, #13027)
533 self.report_warning(
534 'Assuming --restrict-filenames since file system encoding '
535 'cannot encode all characters. '
536 'Set the LC_ALL environment variable to fix this.')
537 self.params['restrictfilenames'] = True
538
539 self.outtmpl_dict = self.parse_outtmpl()
540
541 self._setup_opener()
542
543 """Preload the archive, if any is specified"""
544 def preload_download_archive(fn):
545 if fn is None:
546 return False
547 self.write_debug('Loading archive file %r\n' % fn)
548 try:
549 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
550 for line in archive_file:
551 self.archive.add(line.strip())
552 except IOError as ioe:
553 if ioe.errno != errno.ENOENT:
554 raise
555 return False
556 return True
557
558 self.archive = set()
559 preload_download_archive(self.params.get('download_archive'))
560
561 if auto_init:
562 self.print_debug_header()
563 self.add_default_info_extractors()
564
565 for pp_def_raw in self.params.get('postprocessors', []):
566 pp_class = get_postprocessor(pp_def_raw['key'])
567 pp_def = dict(pp_def_raw)
568 del pp_def['key']
569 if 'when' in pp_def:
570 when = pp_def['when']
571 del pp_def['when']
572 else:
573 when = 'post_process'
574 pp = pp_class(self, **compat_kwargs(pp_def))
575 self.add_post_processor(pp, when=when)
576
577 for ph in self.params.get('post_hooks', []):
578 self.add_post_hook(ph)
579
580 for ph in self.params.get('progress_hooks', []):
581 self.add_progress_hook(ph)
582
583 register_socks_protocols()
584
585 def warn_if_short_id(self, argv):
586 # short YouTube ID starting with dash?
587 idxs = [
588 i for i, a in enumerate(argv)
589 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
590 if idxs:
591 correct_argv = (
592 ['yt-dlp']
593 + [a for i, a in enumerate(argv) if i not in idxs]
594 + ['--'] + [argv[i] for i in idxs]
595 )
596 self.report_warning(
597 'Long argument string detected. '
598 'Use -- to separate parameters and URLs, like this:\n%s\n' %
599 args_to_str(correct_argv))
600
601 def add_info_extractor(self, ie):
602 """Add an InfoExtractor object to the end of the list."""
603 self._ies.append(ie)
604 if not isinstance(ie, type):
605 self._ies_instances[ie.ie_key()] = ie
606 ie.set_downloader(self)
607
608 def get_info_extractor(self, ie_key):
609 """
610 Get an instance of an IE with name ie_key, it will try to get one from
611 the _ies list, if there's no instance it will create a new one and add
612 it to the extractor list.
613 """
614 ie = self._ies_instances.get(ie_key)
615 if ie is None:
616 ie = get_info_extractor(ie_key)()
617 self.add_info_extractor(ie)
618 return ie
619
620 def add_default_info_extractors(self):
621 """
622 Add the InfoExtractors returned by gen_extractors to the end of the list
623 """
624 for ie in gen_extractor_classes():
625 self.add_info_extractor(ie)
626
627 def add_post_processor(self, pp, when='post_process'):
628 """Add a PostProcessor object to the end of the chain."""
629 self._pps[when].append(pp)
630 pp.set_downloader(self)
631
632 def add_post_hook(self, ph):
633 """Add the post hook"""
634 self._post_hooks.append(ph)
635
636 def add_progress_hook(self, ph):
637 """Add the progress hook (currently only for the file downloader)"""
638 self._progress_hooks.append(ph)
639
640 def _bidi_workaround(self, message):
641 if not hasattr(self, '_output_channel'):
642 return message
643
644 assert hasattr(self, '_output_process')
645 assert isinstance(message, compat_str)
646 line_count = message.count('\n') + 1
647 self._output_process.stdin.write((message + '\n').encode('utf-8'))
648 self._output_process.stdin.flush()
649 res = ''.join(self._output_channel.readline().decode('utf-8')
650 for _ in range(line_count))
651 return res[:-len('\n')]
652
653 def _write_string(self, s, out=None):
654 write_string(s, out=out, encoding=self.params.get('encoding'))
655
656 def to_stdout(self, message, skip_eol=False, quiet=False):
657 """Print message to stdout"""
658 if self.params.get('logger'):
659 self.params['logger'].debug(message)
660 elif not quiet:
661 message = self._bidi_workaround(message)
662 terminator = ['\n', ''][skip_eol]
663 output = message + terminator
664
665 self._write_string(output, self._screen_file)
666
667 def to_stderr(self, message):
668 """Print message to stderr"""
669 assert isinstance(message, compat_str)
670 if self.params.get('logger'):
671 self.params['logger'].error(message)
672 else:
673 message = self._bidi_workaround(message)
674 output = message + '\n'
675 self._write_string(output, self._err_file)
676
677 def to_console_title(self, message):
678 if not self.params.get('consoletitle', False):
679 return
680 if compat_os_name == 'nt':
681 if ctypes.windll.kernel32.GetConsoleWindow():
682 # c_wchar_p() might not be necessary if `message` is
683 # already of type unicode()
684 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
685 elif 'TERM' in os.environ:
686 self._write_string('\033]0;%s\007' % message, self._screen_file)
687
688 def save_console_title(self):
689 if not self.params.get('consoletitle', False):
690 return
691 if self.params.get('simulate', False):
692 return
693 if compat_os_name != 'nt' and 'TERM' in os.environ:
694 # Save the title on stack
695 self._write_string('\033[22;0t', self._screen_file)
696
697 def restore_console_title(self):
698 if not self.params.get('consoletitle', False):
699 return
700 if self.params.get('simulate', False):
701 return
702 if compat_os_name != 'nt' and 'TERM' in os.environ:
703 # Restore the title from stack
704 self._write_string('\033[23;0t', self._screen_file)
705
706 def __enter__(self):
707 self.save_console_title()
708 return self
709
710 def __exit__(self, *args):
711 self.restore_console_title()
712
713 if self.params.get('cookiefile') is not None:
714 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
715
716 def trouble(self, message=None, tb=None):
717 """Determine action to take when a download problem appears.
718
719 Depending on if the downloader has been configured to ignore
720 download errors or not, this method may throw an exception or
721 not when errors are found, after printing the message.
722
723 tb, if given, is additional traceback information.
724 """
725 if message is not None:
726 self.to_stderr(message)
727 if self.params.get('verbose'):
728 if tb is None:
729 if sys.exc_info()[0]: # if .trouble has been called from an except block
730 tb = ''
731 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
732 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
733 tb += encode_compat_str(traceback.format_exc())
734 else:
735 tb_data = traceback.format_list(traceback.extract_stack())
736 tb = ''.join(tb_data)
737 if tb:
738 self.to_stderr(tb)
739 if not self.params.get('ignoreerrors', False):
740 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
741 exc_info = sys.exc_info()[1].exc_info
742 else:
743 exc_info = sys.exc_info()
744 raise DownloadError(message, exc_info)
745 self._download_retcode = 1
746
747 def to_screen(self, message, skip_eol=False):
748 """Print message to stdout if not in quiet mode"""
749 self.to_stdout(
750 message, skip_eol, quiet=self.params.get('quiet', False))
751
752 def report_warning(self, message):
753 '''
754 Print the message to stderr, it will be prefixed with 'WARNING:'
755 If stderr is a tty file the 'WARNING:' will be colored
756 '''
757 if self.params.get('logger') is not None:
758 self.params['logger'].warning(message)
759 else:
760 if self.params.get('no_warnings'):
761 return
762 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
763 _msg_header = '\033[0;33mWARNING:\033[0m'
764 else:
765 _msg_header = 'WARNING:'
766 warning_message = '%s %s' % (_msg_header, message)
767 self.to_stderr(warning_message)
768
769 def report_error(self, message, tb=None):
770 '''
771 Do the same as trouble, but prefixes the message with 'ERROR:', colored
772 in red if stderr is a tty file.
773 '''
774 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
775 _msg_header = '\033[0;31mERROR:\033[0m'
776 else:
777 _msg_header = 'ERROR:'
778 error_message = '%s %s' % (_msg_header, message)
779 self.trouble(error_message, tb)
780
781 def write_debug(self, message):
782 '''Log debug message or Print message to stderr'''
783 if not self.params.get('verbose', False):
784 return
785 message = '[debug] %s' % message
786 if self.params.get('logger'):
787 self.params['logger'].debug(message)
788 else:
789 self._write_string('%s\n' % message)
790
791 def report_file_already_downloaded(self, file_name):
792 """Report file has already been fully downloaded."""
793 try:
794 self.to_screen('[download] %s has already been downloaded' % file_name)
795 except UnicodeEncodeError:
796 self.to_screen('[download] The file has already been downloaded')
797
798 def report_file_delete(self, file_name):
799 """Report that existing file will be deleted."""
800 try:
801 self.to_screen('Deleting existing file %s' % file_name)
802 except UnicodeEncodeError:
803 self.to_screen('Deleting existing file')
804
805 def parse_outtmpl(self):
806 outtmpl_dict = self.params.get('outtmpl', {})
807 if not isinstance(outtmpl_dict, dict):
808 outtmpl_dict = {'default': outtmpl_dict}
809 outtmpl_dict.update({
810 k: v for k, v in DEFAULT_OUTTMPL.items()
811 if not outtmpl_dict.get(k)})
812 for key, val in outtmpl_dict.items():
813 if isinstance(val, bytes):
814 self.report_warning(
815 'Parameter outtmpl is bytes, but should be a unicode string. '
816 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
817 return outtmpl_dict
818
819 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=None):
820 """ Make the template and info_dict suitable for substitution (outtmpl % info_dict)"""
821 template_dict = dict(info_dict)
822 na = self.params.get('outtmpl_na_placeholder', 'NA')
823
824 # duration_string
825 template_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
826 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
827 if info_dict.get('duration', None) is not None
828 else None)
829
830 # epoch
831 template_dict['epoch'] = int(time.time())
832
833 # autonumber
834 autonumber_size = self.params.get('autonumber_size')
835 if autonumber_size is None:
836 autonumber_size = 5
837 template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
838
839 # resolution if not defined
840 if template_dict.get('resolution') is None:
841 if template_dict.get('width') and template_dict.get('height'):
842 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
843 elif template_dict.get('height'):
844 template_dict['resolution'] = '%sp' % template_dict['height']
845 elif template_dict.get('width'):
846 template_dict['resolution'] = '%dx?' % template_dict['width']
847
848 # For fields playlist_index and autonumber convert all occurrences
849 # of %(field)s to %(field)0Nd for backward compatibility
850 field_size_compat_map = {
851 'playlist_index': len(str(template_dict.get('_last_playlist_index') or '')),
852 'autonumber': autonumber_size,
853 }
854 FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
855 mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
856 if mobj:
857 outtmpl = re.sub(
858 FIELD_SIZE_COMPAT_RE,
859 r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
860 outtmpl)
861
862 numeric_fields = list(self._NUMERIC_FIELDS)
863 if sanitize is None:
864 sanitize = lambda k, v: v
865
866 EXTERNAL_FORMAT_RE = FORMAT_RE.format('(?P<key>[^)]*)')
867 # Field is of the form key1.key2...
868 # where keys (except first) can be string, int or slice
869 FIELD_RE = r'\w+(?:\.(?:\w+|[-\d]*(?::[-\d]*){0,2}))*'
870 INTERNAL_FORMAT_RE = re.compile(r'''(?x)
871 (?P<negate>-)?
872 (?P<fields>{0})
873 (?P<maths>(?:[-+]-?(?:\d+(?:\.\d+)?|{0}))*)
874 (?:>(?P<strf_format>.+?))?
875 (?:\|(?P<default>.*?))?
876 $'''.format(FIELD_RE))
877 MATH_OPERATORS_RE = re.compile(r'(?<![-+])([-+])')
878 MATH_FUNCTIONS = {
879 '+': float.__add__,
880 '-': float.__sub__,
881 }
882 for outer_mobj in re.finditer(EXTERNAL_FORMAT_RE, outtmpl):
883 final_key = outer_mobj.group('key')
884 str_type = outer_mobj.group('type')
885 value = None
886 mobj = re.match(INTERNAL_FORMAT_RE, final_key)
887 if mobj is not None:
888 mobj = mobj.groupdict()
889 # Object traversal
890 fields = mobj['fields'].split('.')
891 value = traverse_dict(template_dict, fields)
892 # Negative
893 if mobj['negate']:
894 value = float_or_none(value)
895 if value is not None:
896 value *= -1
897 # Do maths
898 if mobj['maths']:
899 value = float_or_none(value)
900 operator = None
901 for item in MATH_OPERATORS_RE.split(mobj['maths'])[1:]:
902 if item == '':
903 value = None
904 if value is None:
905 break
906 if operator:
907 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
908 offset = float_or_none(item)
909 if offset is None:
910 offset = float_or_none(traverse_dict(template_dict, item.split('.')))
911 try:
912 value = operator(value, multiplier * offset)
913 except (TypeError, ZeroDivisionError):
914 value = None
915 operator = None
916 else:
917 operator = MATH_FUNCTIONS[item]
918 # Datetime formatting
919 if mobj['strf_format']:
920 value = strftime_or_none(value, mobj['strf_format'])
921 # Set default
922 if value is None and mobj['default'] is not None:
923 value = mobj['default']
924 # Sanitize
925 if str_type in 'crs' and value is not None: # string
926 value = sanitize('%{}'.format(str_type) % fields[-1], value)
927 else: # numeric
928 numeric_fields.append(final_key)
929 value = float_or_none(value)
930 if value is not None:
931 template_dict[final_key] = value
932
933 # Missing numeric fields used together with integer presentation types
934 # in format specification will break the argument substitution since
935 # string NA placeholder is returned for missing fields. We will patch
936 # output template for missing fields to meet string presentation type.
937 for numeric_field in numeric_fields:
938 if template_dict.get(numeric_field) is None:
939 outtmpl = re.sub(
940 FORMAT_RE.format(re.escape(numeric_field)),
941 r'%({0})s'.format(numeric_field), outtmpl)
942
943 template_dict = collections.defaultdict(lambda: na, (
944 (k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
945 for k, v in template_dict.items() if v is not None))
946 return outtmpl, template_dict
947
948 def _prepare_filename(self, info_dict, tmpl_type='default'):
949 try:
950 sanitize = lambda k, v: sanitize_filename(
951 compat_str(v),
952 restricted=self.params.get('restrictfilenames'),
953 is_id=(k == 'id' or k.endswith('_id')))
954 outtmpl = self.outtmpl_dict.get(tmpl_type, self.outtmpl_dict['default'])
955 outtmpl, template_dict = self.prepare_outtmpl(outtmpl, info_dict, sanitize)
956
957 # expand_path translates '%%' into '%' and '$$' into '$'
958 # correspondingly that is not what we want since we need to keep
959 # '%%' intact for template dict substitution step. Working around
960 # with boundary-alike separator hack.
961 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
962 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
963
964 # outtmpl should be expand_path'ed before template dict substitution
965 # because meta fields may contain env variables we don't want to
966 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
967 # title "Hello $PATH", we don't want `$PATH` to be expanded.
968 filename = expand_path(outtmpl).replace(sep, '') % template_dict
969
970 force_ext = OUTTMPL_TYPES.get(tmpl_type)
971 if force_ext is not None:
972 filename = replace_extension(filename, force_ext, template_dict.get('ext'))
973
974 # https://github.com/blackjack4494/youtube-dlc/issues/85
975 trim_file_name = self.params.get('trim_file_name', False)
976 if trim_file_name:
977 fn_groups = filename.rsplit('.')
978 ext = fn_groups[-1]
979 sub_ext = ''
980 if len(fn_groups) > 2:
981 sub_ext = fn_groups[-2]
982 filename = '.'.join(filter(None, [fn_groups[0][:trim_file_name], sub_ext, ext]))
983
984 return filename
985 except ValueError as err:
986 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
987 return None
988
989 def prepare_filename(self, info_dict, dir_type='', warn=False):
990 """Generate the output filename."""
991 paths = self.params.get('paths', {})
992 assert isinstance(paths, dict)
993 filename = self._prepare_filename(info_dict, dir_type or 'default')
994
995 if warn and not self.__prepare_filename_warned:
996 if not paths:
997 pass
998 elif filename == '-':
999 self.report_warning('--paths is ignored when an outputting to stdout')
1000 elif os.path.isabs(filename):
1001 self.report_warning('--paths is ignored since an absolute path is given in output template')
1002 self.__prepare_filename_warned = True
1003 if filename == '-' or not filename:
1004 return filename
1005
1006 homepath = expand_path(paths.get('home', '').strip())
1007 assert isinstance(homepath, compat_str)
1008 subdir = expand_path(paths.get(dir_type, '').strip()) if dir_type else ''
1009 assert isinstance(subdir, compat_str)
1010 path = os.path.join(homepath, subdir, filename)
1011
1012 # Temporary fix for #4787
1013 # 'Treat' all problem characters by passing filename through preferredencoding
1014 # to workaround encoding issues with subprocess on python2 @ Windows
1015 if sys.version_info < (3, 0) and sys.platform == 'win32':
1016 path = encodeFilename(path, True).decode(preferredencoding())
1017 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1018
1019 def _match_entry(self, info_dict, incomplete):
1020 """ Returns None if the file should be downloaded """
1021
1022 def check_filter():
1023 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1024 if 'title' in info_dict:
1025 # This can happen when we're just evaluating the playlist
1026 title = info_dict['title']
1027 matchtitle = self.params.get('matchtitle', False)
1028 if matchtitle:
1029 if not re.search(matchtitle, title, re.IGNORECASE):
1030 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1031 rejecttitle = self.params.get('rejecttitle', False)
1032 if rejecttitle:
1033 if re.search(rejecttitle, title, re.IGNORECASE):
1034 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1035 date = info_dict.get('upload_date')
1036 if date is not None:
1037 dateRange = self.params.get('daterange', DateRange())
1038 if date not in dateRange:
1039 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
1040 view_count = info_dict.get('view_count')
1041 if view_count is not None:
1042 min_views = self.params.get('min_views')
1043 if min_views is not None and view_count < min_views:
1044 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1045 max_views = self.params.get('max_views')
1046 if max_views is not None and view_count > max_views:
1047 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1048 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1049 return 'Skipping "%s" because it is age restricted' % video_title
1050 if self.in_download_archive(info_dict):
1051 return '%s has already been recorded in archive' % video_title
1052
1053 if not incomplete:
1054 match_filter = self.params.get('match_filter')
1055 if match_filter is not None:
1056 ret = match_filter(info_dict)
1057 if ret is not None:
1058 return ret
1059 return None
1060
1061 reason = check_filter()
1062 if reason is not None:
1063 self.to_screen('[download] ' + reason)
1064 if reason.endswith('has already been recorded in the archive') and self.params.get('break_on_existing', False):
1065 raise ExistingVideoReached()
1066 elif self.params.get('break_on_reject', False):
1067 raise RejectedVideoReached()
1068 return reason
1069
1070 @staticmethod
1071 def add_extra_info(info_dict, extra_info):
1072 '''Set the keys from extra_info in info dict if they are missing'''
1073 for key, value in extra_info.items():
1074 info_dict.setdefault(key, value)
1075
1076 def extract_info(self, url, download=True, ie_key=None, extra_info={},
1077 process=True, force_generic_extractor=False):
1078 """
1079 Return a list with a dictionary for each video extracted.
1080
1081 Arguments:
1082 url -- URL to extract
1083
1084 Keyword arguments:
1085 download -- whether to download videos during extraction
1086 ie_key -- extractor key hint
1087 extra_info -- dictionary containing the extra values to add to each result
1088 process -- whether to resolve all unresolved references (URLs, playlist items),
1089 must be True for download to work.
1090 force_generic_extractor -- force using the generic extractor
1091 """
1092
1093 if not ie_key and force_generic_extractor:
1094 ie_key = 'Generic'
1095
1096 if ie_key:
1097 ies = [self.get_info_extractor(ie_key)]
1098 else:
1099 ies = self._ies
1100
1101 for ie in ies:
1102 if not ie.suitable(url):
1103 continue
1104
1105 ie_key = ie.ie_key()
1106 ie = self.get_info_extractor(ie_key)
1107 if not ie.working():
1108 self.report_warning('The program functionality for this site has been marked as broken, '
1109 'and will probably not work.')
1110
1111 try:
1112 temp_id = str_or_none(
1113 ie.extract_id(url) if callable(getattr(ie, 'extract_id', None))
1114 else ie._match_id(url))
1115 except (AssertionError, IndexError, AttributeError):
1116 temp_id = None
1117 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1118 self.to_screen("[%s] %s: has already been recorded in archive" % (
1119 ie_key, temp_id))
1120 break
1121 return self.__extract_info(url, ie, download, extra_info, process)
1122 else:
1123 self.report_error('no suitable InfoExtractor for URL %s' % url)
1124
1125 def __handle_extraction_exceptions(func):
1126 def wrapper(self, *args, **kwargs):
1127 try:
1128 return func(self, *args, **kwargs)
1129 except GeoRestrictedError as e:
1130 msg = e.msg
1131 if e.countries:
1132 msg += '\nThis video is available in %s.' % ', '.join(
1133 map(ISO3166Utils.short2full, e.countries))
1134 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1135 self.report_error(msg)
1136 except ExtractorError as e: # An error we somewhat expected
1137 self.report_error(compat_str(e), e.format_traceback())
1138 except (MaxDownloadsReached, ExistingVideoReached, RejectedVideoReached):
1139 raise
1140 except Exception as e:
1141 if self.params.get('ignoreerrors', False):
1142 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
1143 else:
1144 raise
1145 return wrapper
1146
1147 @__handle_extraction_exceptions
1148 def __extract_info(self, url, ie, download, extra_info, process):
1149 ie_result = ie.extract(url)
1150 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1151 return
1152 if isinstance(ie_result, list):
1153 # Backwards compatibility: old IE result format
1154 ie_result = {
1155 '_type': 'compat_list',
1156 'entries': ie_result,
1157 }
1158 self.add_default_extra_info(ie_result, ie, url)
1159 if process:
1160 return self.process_ie_result(ie_result, download, extra_info)
1161 else:
1162 return ie_result
1163
1164 def add_default_extra_info(self, ie_result, ie, url):
1165 self.add_extra_info(ie_result, {
1166 'extractor': ie.IE_NAME,
1167 'webpage_url': url,
1168 'webpage_url_basename': url_basename(url),
1169 'extractor_key': ie.ie_key(),
1170 })
1171
1172 def process_ie_result(self, ie_result, download=True, extra_info={}):
1173 """
1174 Take the result of the ie(may be modified) and resolve all unresolved
1175 references (URLs, playlist items).
1176
1177 It will also download the videos if 'download'.
1178 Returns the resolved ie_result.
1179 """
1180 result_type = ie_result.get('_type', 'video')
1181
1182 if result_type in ('url', 'url_transparent'):
1183 ie_result['url'] = sanitize_url(ie_result['url'])
1184 extract_flat = self.params.get('extract_flat', False)
1185 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1186 or extract_flat is True):
1187 self.__forced_printings(ie_result, self.prepare_filename(ie_result), incomplete=True)
1188 return ie_result
1189
1190 if result_type == 'video':
1191 self.add_extra_info(ie_result, extra_info)
1192 ie_result = self.process_video_result(ie_result, download=download)
1193 additional_urls = (ie_result or {}).get('additional_urls')
1194 if additional_urls:
1195 # TODO: Improve MetadataFromFieldPP to allow setting a list
1196 if isinstance(additional_urls, compat_str):
1197 additional_urls = [additional_urls]
1198 self.to_screen(
1199 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1200 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1201 ie_result['additional_entries'] = [
1202 self.extract_info(
1203 url, download, extra_info,
1204 force_generic_extractor=self.params.get('force_generic_extractor'))
1205 for url in additional_urls
1206 ]
1207 return ie_result
1208 elif result_type == 'url':
1209 # We have to add extra_info to the results because it may be
1210 # contained in a playlist
1211 return self.extract_info(
1212 ie_result['url'], download,
1213 ie_key=ie_result.get('ie_key'),
1214 extra_info=extra_info)
1215 elif result_type == 'url_transparent':
1216 # Use the information from the embedding page
1217 info = self.extract_info(
1218 ie_result['url'], ie_key=ie_result.get('ie_key'),
1219 extra_info=extra_info, download=False, process=False)
1220
1221 # extract_info may return None when ignoreerrors is enabled and
1222 # extraction failed with an error, don't crash and return early
1223 # in this case
1224 if not info:
1225 return info
1226
1227 force_properties = dict(
1228 (k, v) for k, v in ie_result.items() if v is not None)
1229 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1230 if f in force_properties:
1231 del force_properties[f]
1232 new_result = info.copy()
1233 new_result.update(force_properties)
1234
1235 # Extracted info may not be a video result (i.e.
1236 # info.get('_type', 'video') != video) but rather an url or
1237 # url_transparent. In such cases outer metadata (from ie_result)
1238 # should be propagated to inner one (info). For this to happen
1239 # _type of info should be overridden with url_transparent. This
1240 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1241 if new_result.get('_type') == 'url':
1242 new_result['_type'] = 'url_transparent'
1243
1244 return self.process_ie_result(
1245 new_result, download=download, extra_info=extra_info)
1246 elif result_type in ('playlist', 'multi_video'):
1247 # Protect from infinite recursion due to recursively nested playlists
1248 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1249 webpage_url = ie_result['webpage_url']
1250 if webpage_url in self._playlist_urls:
1251 self.to_screen(
1252 '[download] Skipping already downloaded playlist: %s'
1253 % ie_result.get('title') or ie_result.get('id'))
1254 return
1255
1256 self._playlist_level += 1
1257 self._playlist_urls.add(webpage_url)
1258 self._sanitize_thumbnails(ie_result)
1259 try:
1260 return self.__process_playlist(ie_result, download)
1261 finally:
1262 self._playlist_level -= 1
1263 if not self._playlist_level:
1264 self._playlist_urls.clear()
1265 elif result_type == 'compat_list':
1266 self.report_warning(
1267 'Extractor %s returned a compat_list result. '
1268 'It needs to be updated.' % ie_result.get('extractor'))
1269
1270 def _fixup(r):
1271 self.add_extra_info(
1272 r,
1273 {
1274 'extractor': ie_result['extractor'],
1275 'webpage_url': ie_result['webpage_url'],
1276 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1277 'extractor_key': ie_result['extractor_key'],
1278 }
1279 )
1280 return r
1281 ie_result['entries'] = [
1282 self.process_ie_result(_fixup(r), download, extra_info)
1283 for r in ie_result['entries']
1284 ]
1285 return ie_result
1286 else:
1287 raise Exception('Invalid result type: %s' % result_type)
1288
1289 def _ensure_dir_exists(self, path):
1290 return make_dir(path, self.report_error)
1291
1292 def __process_playlist(self, ie_result, download):
1293 # We process each entry in the playlist
1294 playlist = ie_result.get('title') or ie_result.get('id')
1295 self.to_screen('[download] Downloading playlist: %s' % playlist)
1296
1297 if 'entries' not in ie_result:
1298 raise EntryNotInPlaylist()
1299 incomplete_entries = bool(ie_result.get('requested_entries'))
1300 if incomplete_entries:
1301 def fill_missing_entries(entries, indexes):
1302 ret = [None] * max(*indexes)
1303 for i, entry in zip(indexes, entries):
1304 ret[i - 1] = entry
1305 return ret
1306 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1307
1308 playlist_results = []
1309
1310 playliststart = self.params.get('playliststart', 1) - 1
1311 playlistend = self.params.get('playlistend')
1312 # For backwards compatibility, interpret -1 as whole list
1313 if playlistend == -1:
1314 playlistend = None
1315
1316 playlistitems_str = self.params.get('playlist_items')
1317 playlistitems = None
1318 if playlistitems_str is not None:
1319 def iter_playlistitems(format):
1320 for string_segment in format.split(','):
1321 if '-' in string_segment:
1322 start, end = string_segment.split('-')
1323 for item in range(int(start), int(end) + 1):
1324 yield int(item)
1325 else:
1326 yield int(string_segment)
1327 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1328
1329 ie_entries = ie_result['entries']
1330
1331 def make_playlistitems_entries(list_ie_entries):
1332 num_entries = len(list_ie_entries)
1333 for i in playlistitems:
1334 if -num_entries < i <= num_entries:
1335 yield list_ie_entries[i - 1]
1336 elif incomplete_entries:
1337 raise EntryNotInPlaylist()
1338
1339 if isinstance(ie_entries, list):
1340 n_all_entries = len(ie_entries)
1341 if playlistitems:
1342 entries = list(make_playlistitems_entries(ie_entries))
1343 else:
1344 entries = ie_entries[playliststart:playlistend]
1345 n_entries = len(entries)
1346 msg = 'Collected %d videos; downloading %d of them' % (n_all_entries, n_entries)
1347 elif isinstance(ie_entries, PagedList):
1348 if playlistitems:
1349 entries = []
1350 for item in playlistitems:
1351 entries.extend(ie_entries.getslice(
1352 item - 1, item
1353 ))
1354 else:
1355 entries = ie_entries.getslice(
1356 playliststart, playlistend)
1357 n_entries = len(entries)
1358 msg = 'Downloading %d videos' % n_entries
1359 else: # iterable
1360 if playlistitems:
1361 entries = list(make_playlistitems_entries(list(itertools.islice(
1362 ie_entries, 0, max(playlistitems)))))
1363 else:
1364 entries = list(itertools.islice(
1365 ie_entries, playliststart, playlistend))
1366 n_entries = len(entries)
1367 msg = 'Downloading %d videos' % n_entries
1368
1369 if any((entry is None for entry in entries)):
1370 raise EntryNotInPlaylist()
1371 if not playlistitems and (playliststart or playlistend):
1372 playlistitems = list(range(1 + playliststart, 1 + playliststart + len(entries)))
1373 ie_result['entries'] = entries
1374 ie_result['requested_entries'] = playlistitems
1375
1376 if self.params.get('allow_playlist_files', True):
1377 ie_copy = {
1378 'playlist': playlist,
1379 'playlist_id': ie_result.get('id'),
1380 'playlist_title': ie_result.get('title'),
1381 'playlist_uploader': ie_result.get('uploader'),
1382 'playlist_uploader_id': ie_result.get('uploader_id'),
1383 'playlist_index': 0,
1384 }
1385 ie_copy.update(dict(ie_result))
1386
1387 if self.params.get('writeinfojson', False):
1388 infofn = self.prepare_filename(ie_copy, 'pl_infojson')
1389 if not self._ensure_dir_exists(encodeFilename(infofn)):
1390 return
1391 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
1392 self.to_screen('[info] Playlist metadata is already present')
1393 else:
1394 self.to_screen('[info] Writing playlist metadata as JSON to: ' + infofn)
1395 try:
1396 write_json_file(self.filter_requested_info(ie_result, self.params.get('clean_infojson', True)), infofn)
1397 except (OSError, IOError):
1398 self.report_error('Cannot write playlist metadata to JSON file ' + infofn)
1399
1400 # TODO: This should be passed to ThumbnailsConvertor if necessary
1401 self._write_thumbnails(ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1402
1403 if self.params.get('writedescription', False):
1404 descfn = self.prepare_filename(ie_copy, 'pl_description')
1405 if not self._ensure_dir_exists(encodeFilename(descfn)):
1406 return
1407 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
1408 self.to_screen('[info] Playlist description is already present')
1409 elif ie_result.get('description') is None:
1410 self.report_warning('There\'s no playlist description to write.')
1411 else:
1412 try:
1413 self.to_screen('[info] Writing playlist description to: ' + descfn)
1414 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1415 descfile.write(ie_result['description'])
1416 except (OSError, IOError):
1417 self.report_error('Cannot write playlist description file ' + descfn)
1418 return
1419
1420 # Save playlist_index before re-ordering
1421 entries = [
1422 ((playlistitems[i - 1] if playlistitems else i), entry)
1423 for i, entry in enumerate(entries, 1)]
1424
1425 if self.params.get('playlistreverse', False):
1426 entries = entries[::-1]
1427 if self.params.get('playlistrandom', False):
1428 random.shuffle(entries)
1429
1430 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1431
1432 self.to_screen('[%s] playlist %s: %s' % (ie_result['extractor'], playlist, msg))
1433 failures = 0
1434 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1435 for i, entry_tuple in enumerate(entries, 1):
1436 playlist_index, entry = entry_tuple
1437 if 'playlist_index' in self.params.get('compat_options', []):
1438 playlist_index = playlistitems[i - 1] if playlistitems else i
1439 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1440 # This __x_forwarded_for_ip thing is a bit ugly but requires
1441 # minimal changes
1442 if x_forwarded_for:
1443 entry['__x_forwarded_for_ip'] = x_forwarded_for
1444 extra = {
1445 'n_entries': n_entries,
1446 '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1447 'playlist_index': playlist_index,
1448 'playlist_autonumber': i,
1449 'playlist': playlist,
1450 'playlist_id': ie_result.get('id'),
1451 'playlist_title': ie_result.get('title'),
1452 'playlist_uploader': ie_result.get('uploader'),
1453 'playlist_uploader_id': ie_result.get('uploader_id'),
1454 'extractor': ie_result['extractor'],
1455 'webpage_url': ie_result['webpage_url'],
1456 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1457 'extractor_key': ie_result['extractor_key'],
1458 }
1459
1460 if self._match_entry(entry, incomplete=True) is not None:
1461 continue
1462
1463 entry_result = self.__process_iterable_entry(entry, download, extra)
1464 if not entry_result:
1465 failures += 1
1466 if failures >= max_failures:
1467 self.report_error(
1468 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1469 break
1470 # TODO: skip failed (empty) entries?
1471 playlist_results.append(entry_result)
1472 ie_result['entries'] = playlist_results
1473 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1474 return ie_result
1475
1476 @__handle_extraction_exceptions
1477 def __process_iterable_entry(self, entry, download, extra_info):
1478 return self.process_ie_result(
1479 entry, download=download, extra_info=extra_info)
1480
1481 def _build_format_filter(self, filter_spec):
1482 " Returns a function to filter the formats according to the filter_spec "
1483
1484 OPERATORS = {
1485 '<': operator.lt,
1486 '<=': operator.le,
1487 '>': operator.gt,
1488 '>=': operator.ge,
1489 '=': operator.eq,
1490 '!=': operator.ne,
1491 }
1492 operator_rex = re.compile(r'''(?x)\s*
1493 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)
1494 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1495 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
1496 $
1497 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1498 m = operator_rex.search(filter_spec)
1499 if m:
1500 try:
1501 comparison_value = int(m.group('value'))
1502 except ValueError:
1503 comparison_value = parse_filesize(m.group('value'))
1504 if comparison_value is None:
1505 comparison_value = parse_filesize(m.group('value') + 'B')
1506 if comparison_value is None:
1507 raise ValueError(
1508 'Invalid value %r in format specification %r' % (
1509 m.group('value'), filter_spec))
1510 op = OPERATORS[m.group('op')]
1511
1512 if not m:
1513 STR_OPERATORS = {
1514 '=': operator.eq,
1515 '^=': lambda attr, value: attr.startswith(value),
1516 '$=': lambda attr, value: attr.endswith(value),
1517 '*=': lambda attr, value: value in attr,
1518 }
1519 str_operator_rex = re.compile(r'''(?x)
1520 \s*(?P<key>[a-zA-Z0-9._-]+)
1521 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?
1522 \s*(?P<value>[a-zA-Z0-9._-]+)
1523 \s*$
1524 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1525 m = str_operator_rex.search(filter_spec)
1526 if m:
1527 comparison_value = m.group('value')
1528 str_op = STR_OPERATORS[m.group('op')]
1529 if m.group('negation'):
1530 op = lambda attr, value: not str_op(attr, value)
1531 else:
1532 op = str_op
1533
1534 if not m:
1535 raise ValueError('Invalid filter specification %r' % filter_spec)
1536
1537 def _filter(f):
1538 actual_value = f.get(m.group('key'))
1539 if actual_value is None:
1540 return m.group('none_inclusive')
1541 return op(actual_value, comparison_value)
1542 return _filter
1543
1544 def _default_format_spec(self, info_dict, download=True):
1545
1546 def can_merge():
1547 merger = FFmpegMergerPP(self)
1548 return merger.available and merger.can_merge()
1549
1550 prefer_best = (
1551 not self.params.get('simulate', False)
1552 and download
1553 and (
1554 not can_merge()
1555 or info_dict.get('is_live', False)
1556 or self.outtmpl_dict['default'] == '-'))
1557 compat = (
1558 prefer_best
1559 or self.params.get('allow_multiple_audio_streams', False)
1560 or 'format-spec' in self.params.get('compat_opts', []))
1561
1562 return (
1563 'best/bestvideo+bestaudio' if prefer_best
1564 else 'bestvideo*+bestaudio/best' if not compat
1565 else 'bestvideo+bestaudio/best')
1566
1567 def build_format_selector(self, format_spec):
1568 def syntax_error(note, start):
1569 message = (
1570 'Invalid format specification: '
1571 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1572 return SyntaxError(message)
1573
1574 PICKFIRST = 'PICKFIRST'
1575 MERGE = 'MERGE'
1576 SINGLE = 'SINGLE'
1577 GROUP = 'GROUP'
1578 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1579
1580 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1581 'video': self.params.get('allow_multiple_video_streams', False)}
1582
1583 check_formats = self.params.get('check_formats')
1584
1585 def _parse_filter(tokens):
1586 filter_parts = []
1587 for type, string, start, _, _ in tokens:
1588 if type == tokenize.OP and string == ']':
1589 return ''.join(filter_parts)
1590 else:
1591 filter_parts.append(string)
1592
1593 def _remove_unused_ops(tokens):
1594 # Remove operators that we don't use and join them with the surrounding strings
1595 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1596 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1597 last_string, last_start, last_end, last_line = None, None, None, None
1598 for type, string, start, end, line in tokens:
1599 if type == tokenize.OP and string == '[':
1600 if last_string:
1601 yield tokenize.NAME, last_string, last_start, last_end, last_line
1602 last_string = None
1603 yield type, string, start, end, line
1604 # everything inside brackets will be handled by _parse_filter
1605 for type, string, start, end, line in tokens:
1606 yield type, string, start, end, line
1607 if type == tokenize.OP and string == ']':
1608 break
1609 elif type == tokenize.OP and string in ALLOWED_OPS:
1610 if last_string:
1611 yield tokenize.NAME, last_string, last_start, last_end, last_line
1612 last_string = None
1613 yield type, string, start, end, line
1614 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1615 if not last_string:
1616 last_string = string
1617 last_start = start
1618 last_end = end
1619 else:
1620 last_string += string
1621 if last_string:
1622 yield tokenize.NAME, last_string, last_start, last_end, last_line
1623
1624 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1625 selectors = []
1626 current_selector = None
1627 for type, string, start, _, _ in tokens:
1628 # ENCODING is only defined in python 3.x
1629 if type == getattr(tokenize, 'ENCODING', None):
1630 continue
1631 elif type in [tokenize.NAME, tokenize.NUMBER]:
1632 current_selector = FormatSelector(SINGLE, string, [])
1633 elif type == tokenize.OP:
1634 if string == ')':
1635 if not inside_group:
1636 # ')' will be handled by the parentheses group
1637 tokens.restore_last_token()
1638 break
1639 elif inside_merge and string in ['/', ',']:
1640 tokens.restore_last_token()
1641 break
1642 elif inside_choice and string == ',':
1643 tokens.restore_last_token()
1644 break
1645 elif string == ',':
1646 if not current_selector:
1647 raise syntax_error('"," must follow a format selector', start)
1648 selectors.append(current_selector)
1649 current_selector = None
1650 elif string == '/':
1651 if not current_selector:
1652 raise syntax_error('"/" must follow a format selector', start)
1653 first_choice = current_selector
1654 second_choice = _parse_format_selection(tokens, inside_choice=True)
1655 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1656 elif string == '[':
1657 if not current_selector:
1658 current_selector = FormatSelector(SINGLE, 'best', [])
1659 format_filter = _parse_filter(tokens)
1660 current_selector.filters.append(format_filter)
1661 elif string == '(':
1662 if current_selector:
1663 raise syntax_error('Unexpected "("', start)
1664 group = _parse_format_selection(tokens, inside_group=True)
1665 current_selector = FormatSelector(GROUP, group, [])
1666 elif string == '+':
1667 if not current_selector:
1668 raise syntax_error('Unexpected "+"', start)
1669 selector_1 = current_selector
1670 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1671 if not selector_2:
1672 raise syntax_error('Expected a selector', start)
1673 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1674 else:
1675 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1676 elif type == tokenize.ENDMARKER:
1677 break
1678 if current_selector:
1679 selectors.append(current_selector)
1680 return selectors
1681
1682 def _merge(formats_pair):
1683 format_1, format_2 = formats_pair
1684
1685 formats_info = []
1686 formats_info.extend(format_1.get('requested_formats', (format_1,)))
1687 formats_info.extend(format_2.get('requested_formats', (format_2,)))
1688
1689 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
1690 get_no_more = {"video": False, "audio": False}
1691 for (i, fmt_info) in enumerate(formats_info):
1692 for aud_vid in ["audio", "video"]:
1693 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
1694 if get_no_more[aud_vid]:
1695 formats_info.pop(i)
1696 get_no_more[aud_vid] = True
1697
1698 if len(formats_info) == 1:
1699 return formats_info[0]
1700
1701 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
1702 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
1703
1704 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
1705 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
1706
1707 output_ext = self.params.get('merge_output_format')
1708 if not output_ext:
1709 if the_only_video:
1710 output_ext = the_only_video['ext']
1711 elif the_only_audio and not video_fmts:
1712 output_ext = the_only_audio['ext']
1713 else:
1714 output_ext = 'mkv'
1715
1716 new_dict = {
1717 'requested_formats': formats_info,
1718 'format': '+'.join(fmt_info.get('format') for fmt_info in formats_info),
1719 'format_id': '+'.join(fmt_info.get('format_id') for fmt_info in formats_info),
1720 'ext': output_ext,
1721 }
1722
1723 if the_only_video:
1724 new_dict.update({
1725 'width': the_only_video.get('width'),
1726 'height': the_only_video.get('height'),
1727 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
1728 'fps': the_only_video.get('fps'),
1729 'vcodec': the_only_video.get('vcodec'),
1730 'vbr': the_only_video.get('vbr'),
1731 'stretched_ratio': the_only_video.get('stretched_ratio'),
1732 })
1733
1734 if the_only_audio:
1735 new_dict.update({
1736 'acodec': the_only_audio.get('acodec'),
1737 'abr': the_only_audio.get('abr'),
1738 })
1739
1740 return new_dict
1741
1742 def _check_formats(formats):
1743 for f in formats:
1744 self.to_screen('[info] Testing format %s' % f['format_id'])
1745 paths = self.params.get('paths', {})
1746 temp_file = os.path.join(
1747 expand_path(paths.get('home', '').strip()),
1748 expand_path(paths.get('temp', '').strip()),
1749 'ytdl.%s.f%s.check-format' % (random_uuidv4(), f['format_id']))
1750 try:
1751 dl, _ = self.dl(temp_file, f, test=True)
1752 except (ExtractorError, IOError, OSError, ValueError) + network_exceptions:
1753 dl = False
1754 finally:
1755 if os.path.exists(temp_file):
1756 os.remove(temp_file)
1757 if dl:
1758 yield f
1759 else:
1760 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1761
1762 def _build_selector_function(selector):
1763 if isinstance(selector, list): # ,
1764 fs = [_build_selector_function(s) for s in selector]
1765
1766 def selector_function(ctx):
1767 for f in fs:
1768 for format in f(ctx):
1769 yield format
1770 return selector_function
1771
1772 elif selector.type == GROUP: # ()
1773 selector_function = _build_selector_function(selector.selector)
1774
1775 elif selector.type == PICKFIRST: # /
1776 fs = [_build_selector_function(s) for s in selector.selector]
1777
1778 def selector_function(ctx):
1779 for f in fs:
1780 picked_formats = list(f(ctx))
1781 if picked_formats:
1782 return picked_formats
1783 return []
1784
1785 elif selector.type == SINGLE: # atom
1786 format_spec = selector.selector or 'best'
1787
1788 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
1789 if format_spec == 'all':
1790 def selector_function(ctx):
1791 formats = list(ctx['formats'])
1792 if check_formats:
1793 formats = _check_formats(formats)
1794 for f in formats:
1795 yield f
1796 elif format_spec == 'mergeall':
1797 def selector_function(ctx):
1798 formats = list(_check_formats(ctx['formats']))
1799 if not formats:
1800 return
1801 merged_format = formats[-1]
1802 for f in formats[-2::-1]:
1803 merged_format = _merge((merged_format, f))
1804 yield merged_format
1805
1806 else:
1807 format_fallback, format_reverse, format_idx = False, True, 1
1808 mobj = re.match(
1809 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
1810 format_spec)
1811 if mobj is not None:
1812 format_idx = int_or_none(mobj.group('n'), default=1)
1813 format_reverse = mobj.group('bw')[0] == 'b'
1814 format_type = (mobj.group('type') or [None])[0]
1815 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
1816 format_modified = mobj.group('mod') is not None
1817
1818 format_fallback = not format_type and not format_modified # for b, w
1819 filter_f = (
1820 (lambda f: f.get('%scodec' % format_type) != 'none')
1821 if format_type and format_modified # bv*, ba*, wv*, wa*
1822 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
1823 if format_type # bv, ba, wv, wa
1824 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
1825 if not format_modified # b, w
1826 else None) # b*, w*
1827 else:
1828 filter_f = ((lambda f: f.get('ext') == format_spec)
1829 if format_spec in ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] # extension
1830 else (lambda f: f.get('format_id') == format_spec)) # id
1831
1832 def selector_function(ctx):
1833 formats = list(ctx['formats'])
1834 if not formats:
1835 return
1836 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
1837 if format_fallback and ctx['incomplete_formats'] and not matches:
1838 # for extractors with incomplete formats (audio only (soundcloud)
1839 # or video only (imgur)) best/worst will fallback to
1840 # best/worst {video,audio}-only format
1841 matches = formats
1842 if format_reverse:
1843 matches = matches[::-1]
1844 if check_formats:
1845 matches = list(itertools.islice(_check_formats(matches), format_idx))
1846 n = len(matches)
1847 if -n <= format_idx - 1 < n:
1848 yield matches[format_idx - 1]
1849
1850 elif selector.type == MERGE: # +
1851 selector_1, selector_2 = map(_build_selector_function, selector.selector)
1852
1853 def selector_function(ctx):
1854 for pair in itertools.product(
1855 selector_1(copy.deepcopy(ctx)), selector_2(copy.deepcopy(ctx))):
1856 yield _merge(pair)
1857
1858 filters = [self._build_format_filter(f) for f in selector.filters]
1859
1860 def final_selector(ctx):
1861 ctx_copy = copy.deepcopy(ctx)
1862 for _filter in filters:
1863 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1864 return selector_function(ctx_copy)
1865 return final_selector
1866
1867 stream = io.BytesIO(format_spec.encode('utf-8'))
1868 try:
1869 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1870 except tokenize.TokenError:
1871 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1872
1873 class TokenIterator(object):
1874 def __init__(self, tokens):
1875 self.tokens = tokens
1876 self.counter = 0
1877
1878 def __iter__(self):
1879 return self
1880
1881 def __next__(self):
1882 if self.counter >= len(self.tokens):
1883 raise StopIteration()
1884 value = self.tokens[self.counter]
1885 self.counter += 1
1886 return value
1887
1888 next = __next__
1889
1890 def restore_last_token(self):
1891 self.counter -= 1
1892
1893 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1894 return _build_selector_function(parsed_selector)
1895
1896 def _calc_headers(self, info_dict):
1897 res = std_headers.copy()
1898
1899 add_headers = info_dict.get('http_headers')
1900 if add_headers:
1901 res.update(add_headers)
1902
1903 cookies = self._calc_cookies(info_dict)
1904 if cookies:
1905 res['Cookie'] = cookies
1906
1907 if 'X-Forwarded-For' not in res:
1908 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1909 if x_forwarded_for_ip:
1910 res['X-Forwarded-For'] = x_forwarded_for_ip
1911
1912 return res
1913
1914 def _calc_cookies(self, info_dict):
1915 pr = sanitized_Request(info_dict['url'])
1916 self.cookiejar.add_cookie_header(pr)
1917 return pr.get_header('Cookie')
1918
1919 @staticmethod
1920 def _sanitize_thumbnails(info_dict):
1921 thumbnails = info_dict.get('thumbnails')
1922 if thumbnails is None:
1923 thumbnail = info_dict.get('thumbnail')
1924 if thumbnail:
1925 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1926 if thumbnails:
1927 thumbnails.sort(key=lambda t: (
1928 t.get('preference') if t.get('preference') is not None else -1,
1929 t.get('width') if t.get('width') is not None else -1,
1930 t.get('height') if t.get('height') is not None else -1,
1931 t.get('id') if t.get('id') is not None else '',
1932 t.get('url')))
1933 for i, t in enumerate(thumbnails):
1934 t['url'] = sanitize_url(t['url'])
1935 if t.get('width') and t.get('height'):
1936 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1937 if t.get('id') is None:
1938 t['id'] = '%d' % i
1939
1940 def process_video_result(self, info_dict, download=True):
1941 assert info_dict.get('_type', 'video') == 'video'
1942
1943 if 'id' not in info_dict:
1944 raise ExtractorError('Missing "id" field in extractor result')
1945 if 'title' not in info_dict:
1946 raise ExtractorError('Missing "title" field in extractor result')
1947
1948 def report_force_conversion(field, field_not, conversion):
1949 self.report_warning(
1950 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
1951 % (field, field_not, conversion))
1952
1953 def sanitize_string_field(info, string_field):
1954 field = info.get(string_field)
1955 if field is None or isinstance(field, compat_str):
1956 return
1957 report_force_conversion(string_field, 'a string', 'string')
1958 info[string_field] = compat_str(field)
1959
1960 def sanitize_numeric_fields(info):
1961 for numeric_field in self._NUMERIC_FIELDS:
1962 field = info.get(numeric_field)
1963 if field is None or isinstance(field, compat_numeric_types):
1964 continue
1965 report_force_conversion(numeric_field, 'numeric', 'int')
1966 info[numeric_field] = int_or_none(field)
1967
1968 sanitize_string_field(info_dict, 'id')
1969 sanitize_numeric_fields(info_dict)
1970
1971 if 'playlist' not in info_dict:
1972 # It isn't part of a playlist
1973 info_dict['playlist'] = None
1974 info_dict['playlist_index'] = None
1975
1976 self._sanitize_thumbnails(info_dict)
1977
1978 if self.params.get('list_thumbnails'):
1979 self.list_thumbnails(info_dict)
1980 return
1981
1982 thumbnail = info_dict.get('thumbnail')
1983 thumbnails = info_dict.get('thumbnails')
1984 if thumbnail:
1985 info_dict['thumbnail'] = sanitize_url(thumbnail)
1986 elif thumbnails:
1987 info_dict['thumbnail'] = thumbnails[-1]['url']
1988
1989 if 'display_id' not in info_dict and 'id' in info_dict:
1990 info_dict['display_id'] = info_dict['id']
1991
1992 for ts_key, date_key in (
1993 ('timestamp', 'upload_date'),
1994 ('release_timestamp', 'release_date'),
1995 ):
1996 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
1997 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1998 # see http://bugs.python.org/issue1646728)
1999 try:
2000 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2001 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2002 except (ValueError, OverflowError, OSError):
2003 pass
2004
2005 # Auto generate title fields corresponding to the *_number fields when missing
2006 # in order to always have clean titles. This is very common for TV series.
2007 for field in ('chapter', 'season', 'episode'):
2008 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2009 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2010
2011 for cc_kind in ('subtitles', 'automatic_captions'):
2012 cc = info_dict.get(cc_kind)
2013 if cc:
2014 for _, subtitle in cc.items():
2015 for subtitle_format in subtitle:
2016 if subtitle_format.get('url'):
2017 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2018 if subtitle_format.get('ext') is None:
2019 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2020
2021 automatic_captions = info_dict.get('automatic_captions')
2022 subtitles = info_dict.get('subtitles')
2023
2024 if self.params.get('listsubtitles', False):
2025 if 'automatic_captions' in info_dict:
2026 self.list_subtitles(
2027 info_dict['id'], automatic_captions, 'automatic captions')
2028 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2029 return
2030
2031 info_dict['requested_subtitles'] = self.process_subtitles(
2032 info_dict['id'], subtitles, automatic_captions)
2033
2034 # We now pick which formats have to be downloaded
2035 if info_dict.get('formats') is None:
2036 # There's only one format available
2037 formats = [info_dict]
2038 else:
2039 formats = info_dict['formats']
2040
2041 if not formats:
2042 if not self.params.get('ignore_no_formats_error'):
2043 raise ExtractorError('No video formats found!')
2044 else:
2045 self.report_warning('No video formats found!')
2046
2047 def is_wellformed(f):
2048 url = f.get('url')
2049 if not url:
2050 self.report_warning(
2051 '"url" field is missing or empty - skipping format, '
2052 'there is an error in extractor')
2053 return False
2054 if isinstance(url, bytes):
2055 sanitize_string_field(f, 'url')
2056 return True
2057
2058 # Filter out malformed formats for better extraction robustness
2059 formats = list(filter(is_wellformed, formats))
2060
2061 formats_dict = {}
2062
2063 # We check that all the formats have the format and format_id fields
2064 for i, format in enumerate(formats):
2065 sanitize_string_field(format, 'format_id')
2066 sanitize_numeric_fields(format)
2067 format['url'] = sanitize_url(format['url'])
2068 if not format.get('format_id'):
2069 format['format_id'] = compat_str(i)
2070 else:
2071 # Sanitize format_id from characters used in format selector expression
2072 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2073 format_id = format['format_id']
2074 if format_id not in formats_dict:
2075 formats_dict[format_id] = []
2076 formats_dict[format_id].append(format)
2077
2078 # Make sure all formats have unique format_id
2079 for format_id, ambiguous_formats in formats_dict.items():
2080 if len(ambiguous_formats) > 1:
2081 for i, format in enumerate(ambiguous_formats):
2082 format['format_id'] = '%s-%d' % (format_id, i)
2083
2084 for i, format in enumerate(formats):
2085 if format.get('format') is None:
2086 format['format'] = '{id} - {res}{note}'.format(
2087 id=format['format_id'],
2088 res=self.format_resolution(format),
2089 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
2090 )
2091 # Automatically determine file extension if missing
2092 if format.get('ext') is None:
2093 format['ext'] = determine_ext(format['url']).lower()
2094 # Automatically determine protocol if missing (useful for format
2095 # selection purposes)
2096 if format.get('protocol') is None:
2097 format['protocol'] = determine_protocol(format)
2098 # Add HTTP headers, so that external programs can use them from the
2099 # json output
2100 full_format_info = info_dict.copy()
2101 full_format_info.update(format)
2102 format['http_headers'] = self._calc_headers(full_format_info)
2103 # Remove private housekeeping stuff
2104 if '__x_forwarded_for_ip' in info_dict:
2105 del info_dict['__x_forwarded_for_ip']
2106
2107 # TODO Central sorting goes here
2108
2109 if formats and formats[0] is not info_dict:
2110 # only set the 'formats' fields if the original info_dict list them
2111 # otherwise we end up with a circular reference, the first (and unique)
2112 # element in the 'formats' field in info_dict is info_dict itself,
2113 # which can't be exported to json
2114 info_dict['formats'] = formats
2115
2116 info_dict, _ = self.pre_process(info_dict)
2117
2118 if self.params.get('listformats'):
2119 if not info_dict.get('formats'):
2120 raise ExtractorError('No video formats found', expected=True)
2121 self.list_formats(info_dict)
2122 return
2123
2124 req_format = self.params.get('format')
2125 if req_format is None:
2126 req_format = self._default_format_spec(info_dict, download=download)
2127 self.write_debug('Default format spec: %s' % req_format)
2128
2129 format_selector = self.build_format_selector(req_format)
2130
2131 # While in format selection we may need to have an access to the original
2132 # format set in order to calculate some metrics or do some processing.
2133 # For now we need to be able to guess whether original formats provided
2134 # by extractor are incomplete or not (i.e. whether extractor provides only
2135 # video-only or audio-only formats) for proper formats selection for
2136 # extractors with such incomplete formats (see
2137 # https://github.com/ytdl-org/youtube-dl/pull/5556).
2138 # Since formats may be filtered during format selection and may not match
2139 # the original formats the results may be incorrect. Thus original formats
2140 # or pre-calculated metrics should be passed to format selection routines
2141 # as well.
2142 # We will pass a context object containing all necessary additional data
2143 # instead of just formats.
2144 # This fixes incorrect format selection issue (see
2145 # https://github.com/ytdl-org/youtube-dl/issues/10083).
2146 incomplete_formats = (
2147 # All formats are video-only or
2148 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2149 # all formats are audio-only
2150 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
2151
2152 ctx = {
2153 'formats': formats,
2154 'incomplete_formats': incomplete_formats,
2155 }
2156
2157 formats_to_download = list(format_selector(ctx))
2158 if not formats_to_download:
2159 if not self.params.get('ignore_no_formats_error'):
2160 raise ExtractorError('Requested format is not available', expected=True)
2161 else:
2162 self.report_warning('Requested format is not available')
2163 elif download:
2164 self.to_screen(
2165 '[info] %s: Downloading %d format(s): %s' % (
2166 info_dict['id'], len(formats_to_download),
2167 ", ".join([f['format_id'] for f in formats_to_download])))
2168 for fmt in formats_to_download:
2169 new_info = dict(info_dict)
2170 # Save a reference to the original info_dict so that it can be modified in process_info if needed
2171 new_info['__original_infodict'] = info_dict
2172 new_info.update(fmt)
2173 self.process_info(new_info)
2174 # We update the info dict with the best quality format (backwards compatibility)
2175 if formats_to_download:
2176 info_dict.update(formats_to_download[-1])
2177 return info_dict
2178
2179 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2180 """Select the requested subtitles and their format"""
2181 available_subs = {}
2182 if normal_subtitles and self.params.get('writesubtitles'):
2183 available_subs.update(normal_subtitles)
2184 if automatic_captions and self.params.get('writeautomaticsub'):
2185 for lang, cap_info in automatic_captions.items():
2186 if lang not in available_subs:
2187 available_subs[lang] = cap_info
2188
2189 if (not self.params.get('writesubtitles') and not
2190 self.params.get('writeautomaticsub') or not
2191 available_subs):
2192 return None
2193
2194 all_sub_langs = available_subs.keys()
2195 if self.params.get('allsubtitles', False):
2196 requested_langs = all_sub_langs
2197 elif self.params.get('subtitleslangs', False):
2198 requested_langs = set()
2199 for lang in self.params.get('subtitleslangs'):
2200 if lang == 'all':
2201 requested_langs.update(all_sub_langs)
2202 continue
2203 discard = lang[0] == '-'
2204 if discard:
2205 lang = lang[1:]
2206 current_langs = filter(re.compile(lang + '$').match, all_sub_langs)
2207 if discard:
2208 for lang in current_langs:
2209 requested_langs.discard(lang)
2210 else:
2211 requested_langs.update(current_langs)
2212 elif 'en' in available_subs:
2213 requested_langs = ['en']
2214 else:
2215 requested_langs = [list(all_sub_langs)[0]]
2216 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2217
2218 formats_query = self.params.get('subtitlesformat', 'best')
2219 formats_preference = formats_query.split('/') if formats_query else []
2220 subs = {}
2221 for lang in requested_langs:
2222 formats = available_subs.get(lang)
2223 if formats is None:
2224 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
2225 continue
2226 for ext in formats_preference:
2227 if ext == 'best':
2228 f = formats[-1]
2229 break
2230 matches = list(filter(lambda f: f['ext'] == ext, formats))
2231 if matches:
2232 f = matches[-1]
2233 break
2234 else:
2235 f = formats[-1]
2236 self.report_warning(
2237 'No subtitle format found matching "%s" for language %s, '
2238 'using %s' % (formats_query, lang, f['ext']))
2239 subs[lang] = f
2240 return subs
2241
2242 def __forced_printings(self, info_dict, filename, incomplete):
2243 def print_mandatory(field, actual_field=None):
2244 if actual_field is None:
2245 actual_field = field
2246 if (self.params.get('force%s' % field, False)
2247 and (not incomplete or info_dict.get(actual_field) is not None)):
2248 self.to_stdout(info_dict[actual_field])
2249
2250 def print_optional(field):
2251 if (self.params.get('force%s' % field, False)
2252 and info_dict.get(field) is not None):
2253 self.to_stdout(info_dict[field])
2254
2255 info_dict = info_dict.copy()
2256 if filename is not None:
2257 info_dict['filename'] = filename
2258 if info_dict.get('requested_formats') is not None:
2259 # For RTMP URLs, also include the playpath
2260 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2261 elif 'url' in info_dict:
2262 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2263
2264 for tmpl in self.params.get('forceprint', []):
2265 if re.match(r'\w+$', tmpl):
2266 tmpl = '%({})s'.format(tmpl)
2267 tmpl, info_copy = self.prepare_outtmpl(tmpl, info_dict)
2268 self.to_stdout(tmpl % info_copy)
2269
2270 print_mandatory('title')
2271 print_mandatory('id')
2272 print_mandatory('url', 'urls')
2273 print_optional('thumbnail')
2274 print_optional('description')
2275 print_optional('filename')
2276 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
2277 self.to_stdout(formatSeconds(info_dict['duration']))
2278 print_mandatory('format')
2279
2280 if self.params.get('forcejson', False):
2281 self.post_extract(info_dict)
2282 self.to_stdout(json.dumps(info_dict, default=repr))
2283
2284 def dl(self, name, info, subtitle=False, test=False):
2285
2286 if test:
2287 verbose = self.params.get('verbose')
2288 params = {
2289 'test': True,
2290 'quiet': not verbose,
2291 'verbose': verbose,
2292 'noprogress': not verbose,
2293 'nopart': True,
2294 'skip_unavailable_fragments': False,
2295 'keep_fragments': False,
2296 'overwrites': True,
2297 '_no_ytdl_file': True,
2298 }
2299 else:
2300 params = self.params
2301 fd = get_suitable_downloader(info, params)(self, params)
2302 if not test:
2303 for ph in self._progress_hooks:
2304 fd.add_progress_hook(ph)
2305 urls = '", "'.join([f['url'] for f in info.get('requested_formats', [])] or [info['url']])
2306 self.write_debug('Invoking downloader on "%s"' % urls)
2307 new_info = dict(info)
2308 if new_info.get('http_headers') is None:
2309 new_info['http_headers'] = self._calc_headers(new_info)
2310 return fd.download(name, new_info, subtitle)
2311
2312 def process_info(self, info_dict):
2313 """Process a single resolved IE result."""
2314
2315 assert info_dict.get('_type', 'video') == 'video'
2316
2317 info_dict.setdefault('__postprocessors', [])
2318
2319 max_downloads = self.params.get('max_downloads')
2320 if max_downloads is not None:
2321 if self._num_downloads >= int(max_downloads):
2322 raise MaxDownloadsReached()
2323
2324 # TODO: backward compatibility, to be removed
2325 info_dict['fulltitle'] = info_dict['title']
2326
2327 if 'format' not in info_dict:
2328 info_dict['format'] = info_dict['ext']
2329
2330 if self._match_entry(info_dict, incomplete=False) is not None:
2331 return
2332
2333 self.post_extract(info_dict)
2334 self._num_downloads += 1
2335
2336 # info_dict['_filename'] needs to be set for backward compatibility
2337 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2338 temp_filename = self.prepare_filename(info_dict, 'temp')
2339 files_to_move = {}
2340
2341 # Forced printings
2342 self.__forced_printings(info_dict, full_filename, incomplete=False)
2343
2344 if self.params.get('simulate', False):
2345 if self.params.get('force_write_download_archive', False):
2346 self.record_download_archive(info_dict)
2347
2348 # Do nothing else if in simulate mode
2349 return
2350
2351 if full_filename is None:
2352 return
2353
2354 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2355 return
2356 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2357 return
2358
2359 if self.params.get('writedescription', False):
2360 descfn = self.prepare_filename(info_dict, 'description')
2361 if not self._ensure_dir_exists(encodeFilename(descfn)):
2362 return
2363 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
2364 self.to_screen('[info] Video description is already present')
2365 elif info_dict.get('description') is None:
2366 self.report_warning('There\'s no description to write.')
2367 else:
2368 try:
2369 self.to_screen('[info] Writing video description to: ' + descfn)
2370 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
2371 descfile.write(info_dict['description'])
2372 except (OSError, IOError):
2373 self.report_error('Cannot write description file ' + descfn)
2374 return
2375
2376 if self.params.get('writeannotations', False):
2377 annofn = self.prepare_filename(info_dict, 'annotation')
2378 if not self._ensure_dir_exists(encodeFilename(annofn)):
2379 return
2380 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2381 self.to_screen('[info] Video annotations are already present')
2382 elif not info_dict.get('annotations'):
2383 self.report_warning('There are no annotations to write.')
2384 else:
2385 try:
2386 self.to_screen('[info] Writing video annotations to: ' + annofn)
2387 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2388 annofile.write(info_dict['annotations'])
2389 except (KeyError, TypeError):
2390 self.report_warning('There are no annotations to write.')
2391 except (OSError, IOError):
2392 self.report_error('Cannot write annotations file: ' + annofn)
2393 return
2394
2395 subtitles_are_requested = any([self.params.get('writesubtitles', False),
2396 self.params.get('writeautomaticsub')])
2397
2398 if subtitles_are_requested and info_dict.get('requested_subtitles'):
2399 # subtitles download errors are already managed as troubles in relevant IE
2400 # that way it will silently go on when used with unsupporting IE
2401 subtitles = info_dict['requested_subtitles']
2402 # ie = self.get_info_extractor(info_dict['extractor_key'])
2403 for sub_lang, sub_info in subtitles.items():
2404 sub_format = sub_info['ext']
2405 sub_filename = subtitles_filename(temp_filename, sub_lang, sub_format, info_dict.get('ext'))
2406 sub_filename_final = subtitles_filename(
2407 self.prepare_filename(info_dict, 'subtitle'), sub_lang, sub_format, info_dict.get('ext'))
2408 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(sub_filename)):
2409 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
2410 sub_info['filepath'] = sub_filename
2411 files_to_move[sub_filename] = sub_filename_final
2412 else:
2413 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
2414 if sub_info.get('data') is not None:
2415 try:
2416 # Use newline='' to prevent conversion of newline characters
2417 # See https://github.com/ytdl-org/youtube-dl/issues/10268
2418 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
2419 subfile.write(sub_info['data'])
2420 sub_info['filepath'] = sub_filename
2421 files_to_move[sub_filename] = sub_filename_final
2422 except (OSError, IOError):
2423 self.report_error('Cannot write subtitles file ' + sub_filename)
2424 return
2425 else:
2426 try:
2427 self.dl(sub_filename, sub_info.copy(), subtitle=True)
2428 sub_info['filepath'] = sub_filename
2429 files_to_move[sub_filename] = sub_filename_final
2430 except (ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
2431 self.report_warning('Unable to download subtitle for "%s": %s' %
2432 (sub_lang, error_to_compat_str(err)))
2433 continue
2434
2435 if self.params.get('writeinfojson', False):
2436 infofn = self.prepare_filename(info_dict, 'infojson')
2437 if not self._ensure_dir_exists(encodeFilename(infofn)):
2438 return
2439 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
2440 self.to_screen('[info] Video metadata is already present')
2441 else:
2442 self.to_screen('[info] Writing video metadata as JSON to: ' + infofn)
2443 try:
2444 write_json_file(self.filter_requested_info(info_dict, self.params.get('clean_infojson', True)), infofn)
2445 except (OSError, IOError):
2446 self.report_error('Cannot write video metadata to JSON file ' + infofn)
2447 return
2448 info_dict['__infojson_filename'] = infofn
2449
2450 for thumb_ext in self._write_thumbnails(info_dict, temp_filename):
2451 thumb_filename_temp = replace_extension(temp_filename, thumb_ext, info_dict.get('ext'))
2452 thumb_filename = replace_extension(
2453 self.prepare_filename(info_dict, 'thumbnail'), thumb_ext, info_dict.get('ext'))
2454 files_to_move[thumb_filename_temp] = thumb_filename
2455
2456 # Write internet shortcut files
2457 url_link = webloc_link = desktop_link = False
2458 if self.params.get('writelink', False):
2459 if sys.platform == "darwin": # macOS.
2460 webloc_link = True
2461 elif sys.platform.startswith("linux"):
2462 desktop_link = True
2463 else: # if sys.platform in ['win32', 'cygwin']:
2464 url_link = True
2465 if self.params.get('writeurllink', False):
2466 url_link = True
2467 if self.params.get('writewebloclink', False):
2468 webloc_link = True
2469 if self.params.get('writedesktoplink', False):
2470 desktop_link = True
2471
2472 if url_link or webloc_link or desktop_link:
2473 if 'webpage_url' not in info_dict:
2474 self.report_error('Cannot write internet shortcut file because the "webpage_url" field is missing in the media information')
2475 return
2476 ascii_url = iri_to_uri(info_dict['webpage_url'])
2477
2478 def _write_link_file(extension, template, newline, embed_filename):
2479 linkfn = replace_extension(full_filename, extension, info_dict.get('ext'))
2480 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2481 self.to_screen('[info] Internet shortcut is already present')
2482 else:
2483 try:
2484 self.to_screen('[info] Writing internet shortcut to: ' + linkfn)
2485 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', newline=newline) as linkfile:
2486 template_vars = {'url': ascii_url}
2487 if embed_filename:
2488 template_vars['filename'] = linkfn[:-(len(extension) + 1)]
2489 linkfile.write(template % template_vars)
2490 except (OSError, IOError):
2491 self.report_error('Cannot write internet shortcut ' + linkfn)
2492 return False
2493 return True
2494
2495 if url_link:
2496 if not _write_link_file('url', DOT_URL_LINK_TEMPLATE, '\r\n', embed_filename=False):
2497 return
2498 if webloc_link:
2499 if not _write_link_file('webloc', DOT_WEBLOC_LINK_TEMPLATE, '\n', embed_filename=False):
2500 return
2501 if desktop_link:
2502 if not _write_link_file('desktop', DOT_DESKTOP_LINK_TEMPLATE, '\n', embed_filename=True):
2503 return
2504
2505 try:
2506 info_dict, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2507 except PostProcessingError as err:
2508 self.report_error('Preprocessing: %s' % str(err))
2509 return
2510
2511 must_record_download_archive = False
2512 if self.params.get('skip_download', False):
2513 info_dict['filepath'] = temp_filename
2514 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2515 info_dict['__files_to_move'] = files_to_move
2516 info_dict = self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict)
2517 else:
2518 # Download
2519 try:
2520
2521 def existing_file(*filepaths):
2522 ext = info_dict.get('ext')
2523 final_ext = self.params.get('final_ext', ext)
2524 existing_files = []
2525 for file in orderedSet(filepaths):
2526 if final_ext != ext:
2527 converted = replace_extension(file, final_ext, ext)
2528 if os.path.exists(encodeFilename(converted)):
2529 existing_files.append(converted)
2530 if os.path.exists(encodeFilename(file)):
2531 existing_files.append(file)
2532
2533 if not existing_files or self.params.get('overwrites', False):
2534 for file in orderedSet(existing_files):
2535 self.report_file_delete(file)
2536 os.remove(encodeFilename(file))
2537 return None
2538
2539 self.report_file_already_downloaded(existing_files[0])
2540 info_dict['ext'] = os.path.splitext(existing_files[0])[1][1:]
2541 return existing_files[0]
2542
2543 success = True
2544 if info_dict.get('requested_formats') is not None:
2545
2546 def compatible_formats(formats):
2547 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2548 video_formats = [format for format in formats if format.get('vcodec') != 'none']
2549 audio_formats = [format for format in formats if format.get('acodec') != 'none']
2550 if len(video_formats) > 2 or len(audio_formats) > 2:
2551 return False
2552
2553 # Check extension
2554 exts = set(format.get('ext') for format in formats)
2555 COMPATIBLE_EXTS = (
2556 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
2557 set(('webm',)),
2558 )
2559 for ext_sets in COMPATIBLE_EXTS:
2560 if ext_sets.issuperset(exts):
2561 return True
2562 # TODO: Check acodec/vcodec
2563 return False
2564
2565 requested_formats = info_dict['requested_formats']
2566 old_ext = info_dict['ext']
2567 if self.params.get('merge_output_format') is None:
2568 if not compatible_formats(requested_formats):
2569 info_dict['ext'] = 'mkv'
2570 self.report_warning(
2571 'Requested formats are incompatible for merge and will be merged into mkv.')
2572 if (info_dict['ext'] == 'webm'
2573 and self.params.get('writethumbnail', False)
2574 and info_dict.get('thumbnails')):
2575 info_dict['ext'] = 'mkv'
2576 self.report_warning(
2577 'webm doesn\'t support embedding a thumbnail, mkv will be used.')
2578
2579 def correct_ext(filename):
2580 filename_real_ext = os.path.splitext(filename)[1][1:]
2581 filename_wo_ext = (
2582 os.path.splitext(filename)[0]
2583 if filename_real_ext == old_ext
2584 else filename)
2585 return '%s.%s' % (filename_wo_ext, info_dict['ext'])
2586
2587 # Ensure filename always has a correct extension for successful merge
2588 full_filename = correct_ext(full_filename)
2589 temp_filename = correct_ext(temp_filename)
2590 dl_filename = existing_file(full_filename, temp_filename)
2591 info_dict['__real_download'] = False
2592
2593 _protocols = set(determine_protocol(f) for f in requested_formats)
2594 if len(_protocols) == 1:
2595 info_dict['protocol'] = _protocols.pop()
2596 directly_mergable = (
2597 'no-direct-merge' not in self.params.get('compat_opts', [])
2598 and info_dict.get('protocol') is not None # All requested formats have same protocol
2599 and not self.params.get('allow_unplayable_formats')
2600 and get_suitable_downloader(info_dict, self.params).__name__ == 'FFmpegFD')
2601 if directly_mergable:
2602 info_dict['url'] = requested_formats[0]['url']
2603 # Treat it as a single download
2604 dl_filename = existing_file(full_filename, temp_filename)
2605 if dl_filename is None:
2606 success, real_download = self.dl(temp_filename, info_dict)
2607 info_dict['__real_download'] = real_download
2608 else:
2609 downloaded = []
2610 merger = FFmpegMergerPP(self)
2611 if self.params.get('allow_unplayable_formats'):
2612 self.report_warning(
2613 'You have requested merging of multiple formats '
2614 'while also allowing unplayable formats to be downloaded. '
2615 'The formats won\'t be merged to prevent data corruption.')
2616 elif not merger.available:
2617 self.report_warning(
2618 'You have requested merging of multiple formats but ffmpeg is not installed. '
2619 'The formats won\'t be merged.')
2620
2621 if dl_filename is None:
2622 for f in requested_formats:
2623 new_info = dict(info_dict)
2624 del new_info['requested_formats']
2625 new_info.update(f)
2626 fname = prepend_extension(
2627 self.prepare_filename(new_info, 'temp'),
2628 'f%s' % f['format_id'], new_info['ext'])
2629 if not self._ensure_dir_exists(fname):
2630 return
2631 downloaded.append(fname)
2632 partial_success, real_download = self.dl(fname, new_info)
2633 info_dict['__real_download'] = info_dict['__real_download'] or real_download
2634 success = success and partial_success
2635 if merger.available and not self.params.get('allow_unplayable_formats'):
2636 info_dict['__postprocessors'].append(merger)
2637 info_dict['__files_to_merge'] = downloaded
2638 # Even if there were no downloads, it is being merged only now
2639 info_dict['__real_download'] = True
2640 else:
2641 for file in downloaded:
2642 files_to_move[file] = None
2643 else:
2644 # Just a single file
2645 dl_filename = existing_file(full_filename, temp_filename)
2646 if dl_filename is None:
2647 success, real_download = self.dl(temp_filename, info_dict)
2648 info_dict['__real_download'] = real_download
2649
2650 dl_filename = dl_filename or temp_filename
2651 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2652
2653 except network_exceptions as err:
2654 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
2655 return
2656 except (OSError, IOError) as err:
2657 raise UnavailableVideoError(err)
2658 except (ContentTooShortError, ) as err:
2659 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
2660 return
2661
2662 if success and full_filename != '-':
2663 # Fixup content
2664 fixup_policy = self.params.get('fixup')
2665 if fixup_policy is None:
2666 fixup_policy = 'detect_or_warn'
2667
2668 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg to fix this automatically.'
2669
2670 stretched_ratio = info_dict.get('stretched_ratio')
2671 if stretched_ratio is not None and stretched_ratio != 1:
2672 if fixup_policy == 'warn':
2673 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
2674 info_dict['id'], stretched_ratio))
2675 elif fixup_policy == 'detect_or_warn':
2676 stretched_pp = FFmpegFixupStretchedPP(self)
2677 if stretched_pp.available:
2678 info_dict['__postprocessors'].append(stretched_pp)
2679 else:
2680 self.report_warning(
2681 '%s: Non-uniform pixel ratio (%s). %s'
2682 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
2683 else:
2684 assert fixup_policy in ('ignore', 'never')
2685
2686 if (info_dict.get('requested_formats') is None
2687 and info_dict.get('container') == 'm4a_dash'
2688 and info_dict.get('ext') == 'm4a'):
2689 if fixup_policy == 'warn':
2690 self.report_warning(
2691 '%s: writing DASH m4a. '
2692 'Only some players support this container.'
2693 % info_dict['id'])
2694 elif fixup_policy == 'detect_or_warn':
2695 fixup_pp = FFmpegFixupM4aPP(self)
2696 if fixup_pp.available:
2697 info_dict['__postprocessors'].append(fixup_pp)
2698 else:
2699 self.report_warning(
2700 '%s: writing DASH m4a. '
2701 'Only some players support this container. %s'
2702 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
2703 else:
2704 assert fixup_policy in ('ignore', 'never')
2705
2706 if ('protocol' in info_dict
2707 and get_suitable_downloader(info_dict, self.params).__name__ == 'HlsFD'):
2708 if fixup_policy == 'warn':
2709 self.report_warning('%s: malformed AAC bitstream detected.' % (
2710 info_dict['id']))
2711 elif fixup_policy == 'detect_or_warn':
2712 fixup_pp = FFmpegFixupM3u8PP(self)
2713 if fixup_pp.available:
2714 info_dict['__postprocessors'].append(fixup_pp)
2715 else:
2716 self.report_warning(
2717 '%s: malformed AAC bitstream detected. %s'
2718 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
2719 else:
2720 assert fixup_policy in ('ignore', 'never')
2721
2722 try:
2723 info_dict = self.post_process(dl_filename, info_dict, files_to_move)
2724 except PostProcessingError as err:
2725 self.report_error('Postprocessing: %s' % str(err))
2726 return
2727 try:
2728 for ph in self._post_hooks:
2729 ph(info_dict['filepath'])
2730 except Exception as err:
2731 self.report_error('post hooks: %s' % str(err))
2732 return
2733 must_record_download_archive = True
2734
2735 if must_record_download_archive or self.params.get('force_write_download_archive', False):
2736 self.record_download_archive(info_dict)
2737 max_downloads = self.params.get('max_downloads')
2738 if max_downloads is not None and self._num_downloads >= int(max_downloads):
2739 raise MaxDownloadsReached()
2740
2741 def download(self, url_list):
2742 """Download a given list of URLs."""
2743 outtmpl = self.outtmpl_dict['default']
2744 if (len(url_list) > 1
2745 and outtmpl != '-'
2746 and '%' not in outtmpl
2747 and self.params.get('max_downloads') != 1):
2748 raise SameFileError(outtmpl)
2749
2750 for url in url_list:
2751 try:
2752 # It also downloads the videos
2753 res = self.extract_info(
2754 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2755 except UnavailableVideoError:
2756 self.report_error('unable to download video')
2757 except MaxDownloadsReached:
2758 self.to_screen('[info] Maximum number of downloaded files reached')
2759 raise
2760 except ExistingVideoReached:
2761 self.to_screen('[info] Encountered a file that is already in the archive, stopping due to --break-on-existing')
2762 raise
2763 except RejectedVideoReached:
2764 self.to_screen('[info] Encountered a file that did not match filter, stopping due to --break-on-reject')
2765 raise
2766 else:
2767 if self.params.get('dump_single_json', False):
2768 self.post_extract(res)
2769 self.to_stdout(json.dumps(res, default=repr))
2770
2771 return self._download_retcode
2772
2773 def download_with_info_file(self, info_filename):
2774 with contextlib.closing(fileinput.FileInput(
2775 [info_filename], mode='r',
2776 openhook=fileinput.hook_encoded('utf-8'))) as f:
2777 # FileInput doesn't have a read method, we can't call json.load
2778 info = self.filter_requested_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
2779 try:
2780 self.process_ie_result(info, download=True)
2781 except (DownloadError, EntryNotInPlaylist):
2782 webpage_url = info.get('webpage_url')
2783 if webpage_url is not None:
2784 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2785 return self.download([webpage_url])
2786 else:
2787 raise
2788 return self._download_retcode
2789
2790 @staticmethod
2791 def filter_requested_info(info_dict, actually_filter=True):
2792 info_dict.pop('__original_infodict', None) # Always remove this
2793 if not actually_filter:
2794 info_dict['epoch'] = int(time.time())
2795 return info_dict
2796 exceptions = {
2797 'remove': ['requested_formats', 'requested_subtitles', 'requested_entries', 'filepath', 'entries'],
2798 'keep': ['_type'],
2799 }
2800 keep_key = lambda k: k in exceptions['keep'] or not (k.startswith('_') or k in exceptions['remove'])
2801 filter_fn = lambda obj: (
2802 list(map(filter_fn, obj)) if isinstance(obj, (list, tuple))
2803 else obj if not isinstance(obj, dict)
2804 else dict((k, filter_fn(v)) for k, v in obj.items() if keep_key(k)))
2805 return filter_fn(info_dict)
2806
2807 def run_pp(self, pp, infodict):
2808 files_to_delete = []
2809 if '__files_to_move' not in infodict:
2810 infodict['__files_to_move'] = {}
2811 files_to_delete, infodict = pp.run(infodict)
2812 if not files_to_delete:
2813 return infodict
2814
2815 if self.params.get('keepvideo', False):
2816 for f in files_to_delete:
2817 infodict['__files_to_move'].setdefault(f, '')
2818 else:
2819 for old_filename in set(files_to_delete):
2820 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2821 try:
2822 os.remove(encodeFilename(old_filename))
2823 except (IOError, OSError):
2824 self.report_warning('Unable to remove downloaded original file')
2825 if old_filename in infodict['__files_to_move']:
2826 del infodict['__files_to_move'][old_filename]
2827 return infodict
2828
2829 @staticmethod
2830 def post_extract(info_dict):
2831 def actual_post_extract(info_dict):
2832 if info_dict.get('_type') in ('playlist', 'multi_video'):
2833 for video_dict in info_dict.get('entries', {}):
2834 actual_post_extract(video_dict or {})
2835 return
2836
2837 post_extractor = info_dict.get('__post_extractor') or (lambda: {})
2838 extra = post_extractor().items()
2839 info_dict.update(extra)
2840 info_dict.pop('__post_extractor', None)
2841
2842 original_infodict = info_dict.get('__original_infodict') or {}
2843 original_infodict.update(extra)
2844 original_infodict.pop('__post_extractor', None)
2845
2846 actual_post_extract(info_dict or {})
2847
2848 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
2849 info = dict(ie_info)
2850 info['__files_to_move'] = files_to_move or {}
2851 for pp in self._pps[key]:
2852 info = self.run_pp(pp, info)
2853 return info, info.pop('__files_to_move', None)
2854
2855 def post_process(self, filename, ie_info, files_to_move=None):
2856 """Run all the postprocessors on the given file."""
2857 info = dict(ie_info)
2858 info['filepath'] = filename
2859 info['__files_to_move'] = files_to_move or {}
2860
2861 for pp in ie_info.get('__postprocessors', []) + self._pps['post_process']:
2862 info = self.run_pp(pp, info)
2863 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
2864 del info['__files_to_move']
2865 for pp in self._pps['after_move']:
2866 info = self.run_pp(pp, info)
2867 return info
2868
2869 def _make_archive_id(self, info_dict):
2870 video_id = info_dict.get('id')
2871 if not video_id:
2872 return
2873 # Future-proof against any change in case
2874 # and backwards compatibility with prior versions
2875 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
2876 if extractor is None:
2877 url = str_or_none(info_dict.get('url'))
2878 if not url:
2879 return
2880 # Try to find matching extractor for the URL and take its ie_key
2881 for ie in self._ies:
2882 if ie.suitable(url):
2883 extractor = ie.ie_key()
2884 break
2885 else:
2886 return
2887 return '%s %s' % (extractor.lower(), video_id)
2888
2889 def in_download_archive(self, info_dict):
2890 fn = self.params.get('download_archive')
2891 if fn is None:
2892 return False
2893
2894 vid_id = self._make_archive_id(info_dict)
2895 if not vid_id:
2896 return False # Incomplete video information
2897
2898 return vid_id in self.archive
2899
2900 def record_download_archive(self, info_dict):
2901 fn = self.params.get('download_archive')
2902 if fn is None:
2903 return
2904 vid_id = self._make_archive_id(info_dict)
2905 assert vid_id
2906 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2907 archive_file.write(vid_id + '\n')
2908 self.archive.add(vid_id)
2909
2910 @staticmethod
2911 def format_resolution(format, default='unknown'):
2912 if format.get('vcodec') == 'none':
2913 return 'audio only'
2914 if format.get('resolution') is not None:
2915 return format['resolution']
2916 if format.get('width') and format.get('height'):
2917 res = '%dx%d' % (format['width'], format['height'])
2918 elif format.get('height'):
2919 res = '%sp' % format['height']
2920 elif format.get('width'):
2921 res = '%dx?' % format['width']
2922 else:
2923 res = default
2924 return res
2925
2926 def _format_note(self, fdict):
2927 res = ''
2928 if fdict.get('ext') in ['f4f', 'f4m']:
2929 res += '(unsupported) '
2930 if fdict.get('language'):
2931 if res:
2932 res += ' '
2933 res += '[%s] ' % fdict['language']
2934 if fdict.get('format_note') is not None:
2935 res += fdict['format_note'] + ' '
2936 if fdict.get('tbr') is not None:
2937 res += '%4dk ' % fdict['tbr']
2938 if fdict.get('container') is not None:
2939 if res:
2940 res += ', '
2941 res += '%s container' % fdict['container']
2942 if (fdict.get('vcodec') is not None
2943 and fdict.get('vcodec') != 'none'):
2944 if res:
2945 res += ', '
2946 res += fdict['vcodec']
2947 if fdict.get('vbr') is not None:
2948 res += '@'
2949 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2950 res += 'video@'
2951 if fdict.get('vbr') is not None:
2952 res += '%4dk' % fdict['vbr']
2953 if fdict.get('fps') is not None:
2954 if res:
2955 res += ', '
2956 res += '%sfps' % fdict['fps']
2957 if fdict.get('acodec') is not None:
2958 if res:
2959 res += ', '
2960 if fdict['acodec'] == 'none':
2961 res += 'video only'
2962 else:
2963 res += '%-5s' % fdict['acodec']
2964 elif fdict.get('abr') is not None:
2965 if res:
2966 res += ', '
2967 res += 'audio'
2968 if fdict.get('abr') is not None:
2969 res += '@%3dk' % fdict['abr']
2970 if fdict.get('asr') is not None:
2971 res += ' (%5dHz)' % fdict['asr']
2972 if fdict.get('filesize') is not None:
2973 if res:
2974 res += ', '
2975 res += format_bytes(fdict['filesize'])
2976 elif fdict.get('filesize_approx') is not None:
2977 if res:
2978 res += ', '
2979 res += '~' + format_bytes(fdict['filesize_approx'])
2980 return res
2981
2982 def _format_note_table(self, f):
2983 def join_fields(*vargs):
2984 return ', '.join((val for val in vargs if val != ''))
2985
2986 return join_fields(
2987 'UNSUPPORTED' if f.get('ext') in ('f4f', 'f4m') else '',
2988 format_field(f, 'language', '[%s]'),
2989 format_field(f, 'format_note'),
2990 format_field(f, 'container', ignore=(None, f.get('ext'))),
2991 format_field(f, 'asr', '%5dHz'))
2992
2993 def list_formats(self, info_dict):
2994 formats = info_dict.get('formats', [info_dict])
2995 new_format = (
2996 'list-formats' not in self.params.get('compat_opts', [])
2997 and self.params.get('list_formats_as_table', True) is not False)
2998 if new_format:
2999 table = [
3000 [
3001 format_field(f, 'format_id'),
3002 format_field(f, 'ext'),
3003 self.format_resolution(f),
3004 format_field(f, 'fps', '%d'),
3005 '|',
3006 format_field(f, 'filesize', ' %s', func=format_bytes) + format_field(f, 'filesize_approx', '~%s', func=format_bytes),
3007 format_field(f, 'tbr', '%4dk'),
3008 shorten_protocol_name(f.get('protocol', '').replace("native", "n")),
3009 '|',
3010 format_field(f, 'vcodec', default='unknown').replace('none', ''),
3011 format_field(f, 'vbr', '%4dk'),
3012 format_field(f, 'acodec', default='unknown').replace('none', ''),
3013 format_field(f, 'abr', '%3dk'),
3014 format_field(f, 'asr', '%5dHz'),
3015 self._format_note_table(f)]
3016 for f in formats
3017 if f.get('preference') is None or f['preference'] >= -1000]
3018 header_line = ['ID', 'EXT', 'RESOLUTION', 'FPS', '|', ' FILESIZE', ' TBR', 'PROTO',
3019 '|', 'VCODEC', ' VBR', 'ACODEC', ' ABR', ' ASR', 'NOTE']
3020 else:
3021 table = [
3022 [
3023 format_field(f, 'format_id'),
3024 format_field(f, 'ext'),
3025 self.format_resolution(f),
3026 self._format_note(f)]
3027 for f in formats
3028 if f.get('preference') is None or f['preference'] >= -1000]
3029 header_line = ['format code', 'extension', 'resolution', 'note']
3030
3031 self.to_screen(
3032 '[info] Available formats for %s:\n%s' % (info_dict['id'], render_table(
3033 header_line,
3034 table,
3035 delim=new_format,
3036 extraGap=(0 if new_format else 1),
3037 hideEmpty=new_format)))
3038
3039 def list_thumbnails(self, info_dict):
3040 thumbnails = info_dict.get('thumbnails')
3041 if not thumbnails:
3042 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
3043 return
3044
3045 self.to_screen(
3046 '[info] Thumbnails for %s:' % info_dict['id'])
3047 self.to_screen(render_table(
3048 ['ID', 'width', 'height', 'URL'],
3049 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
3050
3051 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3052 if not subtitles:
3053 self.to_screen('%s has no %s' % (video_id, name))
3054 return
3055 self.to_screen(
3056 'Available %s for %s:' % (name, video_id))
3057
3058 def _row(lang, formats):
3059 exts, names = zip(*((f['ext'], f.get('name', 'unknown')) for f in reversed(formats)))
3060 if len(set(names)) == 1:
3061 names = [] if names[0] == 'unknown' else names[:1]
3062 return [lang, ', '.join(names), ', '.join(exts)]
3063
3064 self.to_screen(render_table(
3065 ['Language', 'Name', 'Formats'],
3066 [_row(lang, formats) for lang, formats in subtitles.items()],
3067 hideEmpty=True))
3068
3069 def urlopen(self, req):
3070 """ Start an HTTP download """
3071 if isinstance(req, compat_basestring):
3072 req = sanitized_Request(req)
3073 return self._opener.open(req, timeout=self._socket_timeout)
3074
3075 def print_debug_header(self):
3076 if not self.params.get('verbose'):
3077 return
3078
3079 if type('') is not compat_str:
3080 # Python 2.6 on SLES11 SP1 (https://github.com/ytdl-org/youtube-dl/issues/3326)
3081 self.report_warning(
3082 'Your Python is broken! Update to a newer and supported version')
3083
3084 stdout_encoding = getattr(
3085 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
3086 encoding_str = (
3087 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
3088 locale.getpreferredencoding(),
3089 sys.getfilesystemencoding(),
3090 stdout_encoding,
3091 self.get_encoding()))
3092 write_string(encoding_str, encoding=None)
3093
3094 source = (
3095 '(exe)' if hasattr(sys, 'frozen')
3096 else '(zip)' if isinstance(globals().get('__loader__'), zipimporter)
3097 else '(source)' if os.path.basename(sys.argv[0]) == '__main__.py'
3098 else '')
3099 self._write_string('[debug] yt-dlp version %s %s\n' % (__version__, source))
3100 if _LAZY_LOADER:
3101 self._write_string('[debug] Lazy loading extractors enabled\n')
3102 if _PLUGIN_CLASSES:
3103 self._write_string(
3104 '[debug] Plugin Extractors: %s\n' % [ie.ie_key() for ie in _PLUGIN_CLASSES])
3105 if self.params.get('compat_opts'):
3106 self._write_string(
3107 '[debug] Compatibility options: %s\n' % ', '.join(self.params.get('compat_opts')))
3108 try:
3109 sp = subprocess.Popen(
3110 ['git', 'rev-parse', '--short', 'HEAD'],
3111 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3112 cwd=os.path.dirname(os.path.abspath(__file__)))
3113 out, err = process_communicate_or_kill(sp)
3114 out = out.decode().strip()
3115 if re.match('[0-9a-f]+', out):
3116 self._write_string('[debug] Git HEAD: %s\n' % out)
3117 except Exception:
3118 try:
3119 sys.exc_clear()
3120 except Exception:
3121 pass
3122
3123 def python_implementation():
3124 impl_name = platform.python_implementation()
3125 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3126 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3127 return impl_name
3128
3129 self._write_string('[debug] Python version %s (%s %s) - %s\n' % (
3130 platform.python_version(),
3131 python_implementation(),
3132 platform.architecture()[0],
3133 platform_name()))
3134
3135 exe_versions = FFmpegPostProcessor.get_versions(self)
3136 exe_versions['rtmpdump'] = rtmpdump_version()
3137 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3138 exe_str = ', '.join(
3139 '%s %s' % (exe, v)
3140 for exe, v in sorted(exe_versions.items())
3141 if v
3142 )
3143 if not exe_str:
3144 exe_str = 'none'
3145 self._write_string('[debug] exe versions: %s\n' % exe_str)
3146
3147 proxy_map = {}
3148 for handler in self._opener.handlers:
3149 if hasattr(handler, 'proxies'):
3150 proxy_map.update(handler.proxies)
3151 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
3152
3153 if self.params.get('call_home', False):
3154 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3155 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
3156 return
3157 latest_version = self.urlopen(
3158 'https://yt-dl.org/latest/version').read().decode('utf-8')
3159 if version_tuple(latest_version) > version_tuple(__version__):
3160 self.report_warning(
3161 'You are using an outdated version (newest version: %s)! '
3162 'See https://yt-dl.org/update if you need help updating.' %
3163 latest_version)
3164
3165 def _setup_opener(self):
3166 timeout_val = self.params.get('socket_timeout')
3167 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
3168
3169 opts_cookiefile = self.params.get('cookiefile')
3170 opts_proxy = self.params.get('proxy')
3171
3172 if opts_cookiefile is None:
3173 self.cookiejar = compat_cookiejar.CookieJar()
3174 else:
3175 opts_cookiefile = expand_path(opts_cookiefile)
3176 self.cookiejar = YoutubeDLCookieJar(opts_cookiefile)
3177 if os.access(opts_cookiefile, os.R_OK):
3178 self.cookiejar.load(ignore_discard=True, ignore_expires=True)
3179
3180 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3181 if opts_proxy is not None:
3182 if opts_proxy == '':
3183 proxies = {}
3184 else:
3185 proxies = {'http': opts_proxy, 'https': opts_proxy}
3186 else:
3187 proxies = compat_urllib_request.getproxies()
3188 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3189 if 'http' in proxies and 'https' not in proxies:
3190 proxies['https'] = proxies['http']
3191 proxy_handler = PerRequestProxyHandler(proxies)
3192
3193 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3194 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3195 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3196 redirect_handler = YoutubeDLRedirectHandler()
3197 data_handler = compat_urllib_request_DataHandler()
3198
3199 # When passing our own FileHandler instance, build_opener won't add the
3200 # default FileHandler and allows us to disable the file protocol, which
3201 # can be used for malicious purposes (see
3202 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3203 file_handler = compat_urllib_request.FileHandler()
3204
3205 def file_open(*args, **kwargs):
3206 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3207 file_handler.file_open = file_open
3208
3209 opener = compat_urllib_request.build_opener(
3210 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3211
3212 # Delete the default user-agent header, which would otherwise apply in
3213 # cases where our custom HTTP handler doesn't come into play
3214 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3215 opener.addheaders = []
3216 self._opener = opener
3217
3218 def encode(self, s):
3219 if isinstance(s, bytes):
3220 return s # Already encoded
3221
3222 try:
3223 return s.encode(self.get_encoding())
3224 except UnicodeEncodeError as err:
3225 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3226 raise
3227
3228 def get_encoding(self):
3229 encoding = self.params.get('encoding')
3230 if encoding is None:
3231 encoding = preferredencoding()
3232 return encoding
3233
3234 def _write_thumbnails(self, info_dict, filename): # return the extensions
3235 write_all = self.params.get('write_all_thumbnails', False)
3236 thumbnails = []
3237 if write_all or self.params.get('writethumbnail', False):
3238 thumbnails = info_dict.get('thumbnails') or []
3239 multiple = write_all and len(thumbnails) > 1
3240
3241 ret = []
3242 for t in thumbnails[::1 if write_all else -1]:
3243 thumb_ext = determine_ext(t['url'], 'jpg')
3244 suffix = '%s.' % t['id'] if multiple else ''
3245 thumb_display_id = '%s ' % t['id'] if multiple else ''
3246 t['filepath'] = thumb_filename = replace_extension(filename, suffix + thumb_ext, info_dict.get('ext'))
3247
3248 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(thumb_filename)):
3249 ret.append(suffix + thumb_ext)
3250 self.to_screen('[%s] %s: Thumbnail %sis already present' %
3251 (info_dict['extractor'], info_dict['id'], thumb_display_id))
3252 else:
3253 self.to_screen('[%s] %s: Downloading thumbnail %s ...' %
3254 (info_dict['extractor'], info_dict['id'], thumb_display_id))
3255 try:
3256 uf = self.urlopen(t['url'])
3257 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3258 shutil.copyfileobj(uf, thumbf)
3259 ret.append(suffix + thumb_ext)
3260 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
3261 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
3262 except network_exceptions as err:
3263 self.report_warning('Unable to download thumbnail "%s": %s' %
3264 (t['url'], error_to_compat_str(err)))
3265 if ret and not write_all:
3266 break
3267 return ret