]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
Completely change project name to yt-dlp (#85)
[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 socket
23 import sys
24 import time
25 import tokenize
26 import traceback
27 import random
28
29 from string import ascii_letters
30 from zipimport import zipimporter
31
32 from .compat import (
33 compat_basestring,
34 compat_cookiejar,
35 compat_get_terminal_size,
36 compat_http_client,
37 compat_kwargs,
38 compat_numeric_types,
39 compat_os_name,
40 compat_str,
41 compat_tokenize_tokenize,
42 compat_urllib_error,
43 compat_urllib_request,
44 compat_urllib_request_DataHandler,
45 )
46 from .utils import (
47 age_restricted,
48 args_to_str,
49 ContentTooShortError,
50 date_from_str,
51 DateRange,
52 DEFAULT_OUTTMPL,
53 OUTTMPL_TYPES,
54 determine_ext,
55 determine_protocol,
56 DOT_DESKTOP_LINK_TEMPLATE,
57 DOT_URL_LINK_TEMPLATE,
58 DOT_WEBLOC_LINK_TEMPLATE,
59 DownloadError,
60 encode_compat_str,
61 encodeFilename,
62 error_to_compat_str,
63 ExistingVideoReached,
64 expand_path,
65 ExtractorError,
66 float_or_none,
67 format_bytes,
68 format_field,
69 formatSeconds,
70 GeoRestrictedError,
71 int_or_none,
72 iri_to_uri,
73 ISO3166Utils,
74 locked_file,
75 make_dir,
76 make_HTTPS_handler,
77 MaxDownloadsReached,
78 orderedSet,
79 PagedList,
80 parse_filesize,
81 PerRequestProxyHandler,
82 platform_name,
83 PostProcessingError,
84 preferredencoding,
85 prepend_extension,
86 register_socks_protocols,
87 render_table,
88 replace_extension,
89 RejectedVideoReached,
90 SameFileError,
91 sanitize_filename,
92 sanitize_path,
93 sanitize_url,
94 sanitized_Request,
95 std_headers,
96 str_or_none,
97 strftime_or_none,
98 subtitles_filename,
99 to_high_limit_path,
100 UnavailableVideoError,
101 url_basename,
102 version_tuple,
103 write_json_file,
104 write_string,
105 YoutubeDLCookieJar,
106 YoutubeDLCookieProcessor,
107 YoutubeDLHandler,
108 YoutubeDLRedirectHandler,
109 process_communicate_or_kill,
110 )
111 from .cache import Cache
112 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER, _PLUGIN_CLASSES
113 from .extractor.openload import PhantomJSwrapper
114 from .downloader import get_suitable_downloader
115 from .downloader.rtmp import rtmpdump_version
116 from .postprocessor import (
117 FFmpegFixupM3u8PP,
118 FFmpegFixupM4aPP,
119 FFmpegFixupStretchedPP,
120 FFmpegMergerPP,
121 FFmpegPostProcessor,
122 # FFmpegSubtitlesConvertorPP,
123 get_postprocessor,
124 MoveFilesAfterDownloadPP,
125 )
126 from .version import __version__
127
128 if compat_os_name == 'nt':
129 import ctypes
130
131
132 class YoutubeDL(object):
133 """YoutubeDL class.
134
135 YoutubeDL objects are the ones responsible of downloading the
136 actual video file and writing it to disk if the user has requested
137 it, among some other tasks. In most cases there should be one per
138 program. As, given a video URL, the downloader doesn't know how to
139 extract all the needed information, task that InfoExtractors do, it
140 has to pass the URL to one of them.
141
142 For this, YoutubeDL objects have a method that allows
143 InfoExtractors to be registered in a given order. When it is passed
144 a URL, the YoutubeDL object handles it to the first InfoExtractor it
145 finds that reports being able to handle it. The InfoExtractor extracts
146 all the information about the video or videos the URL refers to, and
147 YoutubeDL process the extracted information, possibly using a File
148 Downloader to download the video.
149
150 YoutubeDL objects accept a lot of parameters. In order not to saturate
151 the object constructor with arguments, it receives a dictionary of
152 options instead. These options are available through the params
153 attribute for the InfoExtractors to use. The YoutubeDL also
154 registers itself as the downloader in charge for the InfoExtractors
155 that are added to it, so this is a "mutual registration".
156
157 Available options:
158
159 username: Username for authentication purposes.
160 password: Password for authentication purposes.
161 videopassword: Password for accessing a video.
162 ap_mso: Adobe Pass multiple-system operator identifier.
163 ap_username: Multiple-system operator account username.
164 ap_password: Multiple-system operator account password.
165 usenetrc: Use netrc for authentication instead.
166 verbose: Print additional info to stdout.
167 quiet: Do not print messages to stdout.
168 no_warnings: Do not print out anything for warnings.
169 forceurl: Force printing final URL.
170 forcetitle: Force printing title.
171 forceid: Force printing ID.
172 forcethumbnail: Force printing thumbnail URL.
173 forcedescription: Force printing description.
174 forcefilename: Force printing final filename.
175 forceduration: Force printing duration.
176 forcejson: Force printing info_dict as JSON.
177 dump_single_json: Force printing the info_dict of the whole playlist
178 (or video) as a single JSON line.
179 force_write_download_archive: Force writing download archive regardless
180 of 'skip_download' or 'simulate'.
181 simulate: Do not download the video files.
182 format: Video format code. see "FORMAT SELECTION" for more details.
183 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
184 format_sort: How to sort the video formats. see "Sorting Formats"
185 for more details.
186 format_sort_force: Force the given format_sort. see "Sorting Formats"
187 for more details.
188 allow_multiple_video_streams: Allow multiple video streams to be merged
189 into a single file
190 allow_multiple_audio_streams: Allow multiple audio streams to be merged
191 into a single file
192 paths: Dictionary of output paths. The allowed keys are 'home'
193 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
194 outtmpl: Dictionary of templates for output names. Allowed keys
195 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
196 A string a also accepted for backward compatibility
197 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
198 restrictfilenames: Do not allow "&" and spaces in file names
199 trim_file_name: Limit length of filename (extension excluded)
200 windowsfilenames: Force the filenames to be windows compatible
201 ignoreerrors: Do not stop on download errors
202 (Default True when running yt-dlp,
203 but False when directly accessing YoutubeDL class)
204 force_generic_extractor: Force downloader to use the generic extractor
205 overwrites: Overwrite all video and metadata files if True,
206 overwrite only non-video files if None
207 and don't overwrite any file if False
208 playliststart: Playlist item to start at.
209 playlistend: Playlist item to end at.
210 playlist_items: Specific indices of playlist to download.
211 playlistreverse: Download playlist items in reverse order.
212 playlistrandom: Download playlist items in random order.
213 matchtitle: Download only matching titles.
214 rejecttitle: Reject downloads for matching titles.
215 logger: Log messages to a logging.Logger instance.
216 logtostderr: Log messages to stderr instead of stdout.
217 writedescription: Write the video description to a .description file
218 writeinfojson: Write the video description to a .info.json file
219 writecomments: Extract video comments. This will not be written to disk
220 unless writeinfojson is also given
221 writeannotations: Write the video annotations to a .annotations.xml file
222 writethumbnail: Write the thumbnail image to a file
223 allow_playlist_files: Whether to write playlists' description, infojson etc
224 also to disk when using the 'write*' options
225 write_all_thumbnails: Write all thumbnail formats to files
226 writelink: Write an internet shortcut file, depending on the
227 current platform (.url/.webloc/.desktop)
228 writeurllink: Write a Windows internet shortcut file (.url)
229 writewebloclink: Write a macOS internet shortcut file (.webloc)
230 writedesktoplink: Write a Linux internet shortcut file (.desktop)
231 writesubtitles: Write the video subtitles to a file
232 writeautomaticsub: Write the automatically generated subtitles to a file
233 allsubtitles: Downloads all the subtitles of the video
234 (requires writesubtitles or writeautomaticsub)
235 listsubtitles: Lists all available subtitles for the video
236 subtitlesformat: The format code for subtitles
237 subtitleslangs: List of languages of the subtitles to download
238 keepvideo: Keep the video file after post-processing
239 daterange: A DateRange object, download only if the upload_date is in the range.
240 skip_download: Skip the actual download of the video file
241 cachedir: Location of the cache files in the filesystem.
242 False to disable filesystem cache.
243 noplaylist: Download single video instead of a playlist if in doubt.
244 age_limit: An integer representing the user's age in years.
245 Unsuitable videos for the given age are skipped.
246 min_views: An integer representing the minimum view count the video
247 must have in order to not be skipped.
248 Videos without view count information are always
249 downloaded. None for no limit.
250 max_views: An integer representing the maximum view count.
251 Videos that are more popular than that are not
252 downloaded.
253 Videos without view count information are always
254 downloaded. None for no limit.
255 download_archive: File name of a file where all downloads are recorded.
256 Videos already present in the file are not downloaded
257 again.
258 break_on_existing: Stop the download process after attempting to download a
259 file that is in the archive.
260 break_on_reject: Stop the download process when encountering a video that
261 has been filtered out.
262 cookiefile: File name where cookies should be read from and dumped to
263 nocheckcertificate:Do not verify SSL certificates
264 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
265 At the moment, this is only supported by YouTube.
266 proxy: URL of the proxy server to use
267 geo_verification_proxy: URL of the proxy to use for IP address verification
268 on geo-restricted sites.
269 socket_timeout: Time to wait for unresponsive hosts, in seconds
270 bidi_workaround: Work around buggy terminals without bidirectional text
271 support, using fridibi
272 debug_printtraffic:Print out sent and received HTTP traffic
273 include_ads: Download ads as well
274 default_search: Prepend this string if an input url is not valid.
275 'auto' for elaborate guessing
276 encoding: Use this encoding instead of the system-specified.
277 extract_flat: Do not resolve URLs, return the immediate result.
278 Pass in 'in_playlist' to only show this behavior for
279 playlist items.
280 postprocessors: A list of dictionaries, each with an entry
281 * key: The name of the postprocessor. See
282 yt_dlp/postprocessor/__init__.py for a list.
283 * _after_move: Optional. If True, run this post_processor
284 after 'MoveFilesAfterDownload'
285 as well as any further keyword arguments for the
286 postprocessor.
287 post_hooks: A list of functions that get called as the final step
288 for each video file, after all postprocessors have been
289 called. The filename will be passed as the only argument.
290 progress_hooks: A list of functions that get called on download
291 progress, with a dictionary with the entries
292 * status: One of "downloading", "error", or "finished".
293 Check this first and ignore unknown values.
294
295 If status is one of "downloading", or "finished", the
296 following properties may also be present:
297 * filename: The final filename (always present)
298 * tmpfilename: The filename we're currently writing to
299 * downloaded_bytes: Bytes on disk
300 * total_bytes: Size of the whole file, None if unknown
301 * total_bytes_estimate: Guess of the eventual file size,
302 None if unavailable.
303 * elapsed: The number of seconds since download started.
304 * eta: The estimated time in seconds, None if unknown
305 * speed: The download speed in bytes/second, None if
306 unknown
307 * fragment_index: The counter of the currently
308 downloaded video fragment.
309 * fragment_count: The number of fragments (= individual
310 files that will be merged)
311
312 Progress hooks are guaranteed to be called at least once
313 (with status "finished") if the download is successful.
314 merge_output_format: Extension to use when merging formats.
315 final_ext: Expected final extension; used to detect when the file was
316 already downloaded and converted. "merge_output_format" is
317 replaced by this extension when given
318 fixup: Automatically correct known faults of the file.
319 One of:
320 - "never": do nothing
321 - "warn": only emit a warning
322 - "detect_or_warn": check whether we can do anything
323 about it, warn otherwise (default)
324 source_address: Client-side IP address to bind to.
325 call_home: Boolean, true iff we are allowed to contact the
326 yt-dlp servers for debugging. (BROKEN)
327 sleep_interval: Number of seconds to sleep before each download when
328 used alone or a lower bound of a range for randomized
329 sleep before each download (minimum possible number
330 of seconds to sleep) when used along with
331 max_sleep_interval.
332 max_sleep_interval:Upper bound of a range for randomized sleep before each
333 download (maximum possible number of seconds to sleep).
334 Must only be used along with sleep_interval.
335 Actual sleep time will be a random float from range
336 [sleep_interval; max_sleep_interval].
337 listformats: Print an overview of available video formats and exit.
338 list_thumbnails: Print a table of all thumbnails and exit.
339 match_filter: A function that gets called with the info_dict of
340 every video.
341 If it returns a message, the video is ignored.
342 If it returns None, the video is downloaded.
343 match_filter_func in utils.py is one example for this.
344 no_color: Do not emit color codes in output.
345 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
346 HTTP header
347 geo_bypass_country:
348 Two-letter ISO 3166-2 country code that will be used for
349 explicit geographic restriction bypassing via faking
350 X-Forwarded-For HTTP header
351 geo_bypass_ip_block:
352 IP range in CIDR notation that will be used similarly to
353 geo_bypass_country
354
355 The following options determine which downloader is picked:
356 external_downloader: Executable of the external downloader to call.
357 None or unset for standard (built-in) downloader.
358 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
359 if True, otherwise use ffmpeg/avconv if False, otherwise
360 use downloader suggested by extractor if None.
361
362 The following parameters are not used by YoutubeDL itself, they are used by
363 the downloader (see yt_dlp/downloader/common.py):
364 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
365 noresizebuffer, retries, continuedl, noprogress, consoletitle,
366 xattr_set_filesize, external_downloader_args, hls_use_mpegts,
367 http_chunk_size.
368
369 The following options are used by the post processors:
370 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
371 otherwise prefer ffmpeg. (avconv support is deprecated)
372 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
373 to the binary or its containing directory.
374 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
375 and a list of additional command-line arguments for the
376 postprocessor/executable. The dict can also have "PP+EXE" keys
377 which are used when the given exe is used by the given PP.
378 Use 'default' as the name for arguments to passed to all PP
379
380 The following options are used by the extractors:
381 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
382 hls_split_discontinuity: Split HLS playlists to different formats at
383 discontinuities such as ad breaks (default: False)
384 youtube_include_dash_manifest: If True (default), DASH manifests and related
385 data will be downloaded and processed by extractor.
386 You can reduce network I/O by disabling it if you don't
387 care about DASH. (only for youtube)
388 youtube_include_hls_manifest: If True (default), HLS manifests and related
389 data will be downloaded and processed by extractor.
390 You can reduce network I/O by disabling it if you don't
391 care about HLS. (only for youtube)
392 """
393
394 _NUMERIC_FIELDS = set((
395 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
396 'timestamp', 'upload_year', 'upload_month', 'upload_day',
397 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
398 'average_rating', 'comment_count', 'age_limit',
399 'start_time', 'end_time',
400 'chapter_number', 'season_number', 'episode_number',
401 'track_number', 'disc_number', 'release_year',
402 'playlist_index',
403 ))
404
405 params = None
406 _ies = []
407 _pps = {'beforedl': [], 'aftermove': [], 'normal': []}
408 __prepare_filename_warned = False
409 _download_retcode = None
410 _num_downloads = None
411 _playlist_level = 0
412 _playlist_urls = set()
413 _screen_file = None
414
415 def __init__(self, params=None, auto_init=True):
416 """Create a FileDownloader object with the given options."""
417 if params is None:
418 params = {}
419 self._ies = []
420 self._ies_instances = {}
421 self._pps = {'beforedl': [], 'aftermove': [], 'normal': []}
422 self.__prepare_filename_warned = False
423 self._post_hooks = []
424 self._progress_hooks = []
425 self._download_retcode = 0
426 self._num_downloads = 0
427 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
428 self._err_file = sys.stderr
429 self.params = {
430 # Default parameters
431 'nocheckcertificate': False,
432 }
433 self.params.update(params)
434 self.cache = Cache(self)
435 self.archive = set()
436
437 """Preload the archive, if any is specified"""
438 def preload_download_archive(self):
439 fn = self.params.get('download_archive')
440 if fn is None:
441 return False
442 try:
443 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
444 for line in archive_file:
445 self.archive.add(line.strip())
446 except IOError as ioe:
447 if ioe.errno != errno.ENOENT:
448 raise
449 return False
450 return True
451
452 def check_deprecated(param, option, suggestion):
453 if self.params.get(param) is not None:
454 self.report_warning(
455 '%s is deprecated. Use %s instead.' % (option, suggestion))
456 return True
457 return False
458
459 if self.params.get('verbose'):
460 self.to_stdout('[debug] Loading archive file %r' % self.params.get('download_archive'))
461
462 preload_download_archive(self)
463
464 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
465 if self.params.get('geo_verification_proxy') is None:
466 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
467
468 if self.params.get('final_ext'):
469 if self.params.get('merge_output_format'):
470 self.report_warning('--merge-output-format will be ignored since --remux-video or --recode-video is given')
471 self.params['merge_output_format'] = self.params['final_ext']
472
473 if 'overwrites' in self.params and self.params['overwrites'] is None:
474 del self.params['overwrites']
475
476 check_deprecated('autonumber_size', '--autonumber-size', 'output template with %(autonumber)0Nd, where N in the number of digits')
477 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
478 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
479
480 if params.get('bidi_workaround', False):
481 try:
482 import pty
483 master, slave = pty.openpty()
484 width = compat_get_terminal_size().columns
485 if width is None:
486 width_args = []
487 else:
488 width_args = ['-w', str(width)]
489 sp_kwargs = dict(
490 stdin=subprocess.PIPE,
491 stdout=slave,
492 stderr=self._err_file)
493 try:
494 self._output_process = subprocess.Popen(
495 ['bidiv'] + width_args, **sp_kwargs
496 )
497 except OSError:
498 self._output_process = subprocess.Popen(
499 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
500 self._output_channel = os.fdopen(master, 'rb')
501 except OSError as ose:
502 if ose.errno == errno.ENOENT:
503 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 $PATH.')
504 else:
505 raise
506
507 if (sys.platform != 'win32'
508 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
509 and not params.get('restrictfilenames', False)):
510 # Unicode filesystem API will throw errors (#1474, #13027)
511 self.report_warning(
512 'Assuming --restrict-filenames since file system encoding '
513 'cannot encode all characters. '
514 'Set the LC_ALL environment variable to fix this.')
515 self.params['restrictfilenames'] = True
516
517 self.outtmpl_dict = self.parse_outtmpl()
518
519 self._setup_opener()
520
521 if auto_init:
522 self.print_debug_header()
523 self.add_default_info_extractors()
524
525 for pp_def_raw in self.params.get('postprocessors', []):
526 pp_class = get_postprocessor(pp_def_raw['key'])
527 pp_def = dict(pp_def_raw)
528 del pp_def['key']
529 if 'when' in pp_def:
530 when = pp_def['when']
531 del pp_def['when']
532 else:
533 when = 'normal'
534 pp = pp_class(self, **compat_kwargs(pp_def))
535 self.add_post_processor(pp, when=when)
536
537 for ph in self.params.get('post_hooks', []):
538 self.add_post_hook(ph)
539
540 for ph in self.params.get('progress_hooks', []):
541 self.add_progress_hook(ph)
542
543 register_socks_protocols()
544
545 def warn_if_short_id(self, argv):
546 # short YouTube ID starting with dash?
547 idxs = [
548 i for i, a in enumerate(argv)
549 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
550 if idxs:
551 correct_argv = (
552 ['yt-dlp']
553 + [a for i, a in enumerate(argv) if i not in idxs]
554 + ['--'] + [argv[i] for i in idxs]
555 )
556 self.report_warning(
557 'Long argument string detected. '
558 'Use -- to separate parameters and URLs, like this:\n%s\n' %
559 args_to_str(correct_argv))
560
561 def add_info_extractor(self, ie):
562 """Add an InfoExtractor object to the end of the list."""
563 self._ies.append(ie)
564 if not isinstance(ie, type):
565 self._ies_instances[ie.ie_key()] = ie
566 ie.set_downloader(self)
567
568 def get_info_extractor(self, ie_key):
569 """
570 Get an instance of an IE with name ie_key, it will try to get one from
571 the _ies list, if there's no instance it will create a new one and add
572 it to the extractor list.
573 """
574 ie = self._ies_instances.get(ie_key)
575 if ie is None:
576 ie = get_info_extractor(ie_key)()
577 self.add_info_extractor(ie)
578 return ie
579
580 def add_default_info_extractors(self):
581 """
582 Add the InfoExtractors returned by gen_extractors to the end of the list
583 """
584 for ie in gen_extractor_classes():
585 self.add_info_extractor(ie)
586
587 def add_post_processor(self, pp, when='normal'):
588 """Add a PostProcessor object to the end of the chain."""
589 self._pps[when].append(pp)
590 pp.set_downloader(self)
591
592 def add_post_hook(self, ph):
593 """Add the post hook"""
594 self._post_hooks.append(ph)
595
596 def add_progress_hook(self, ph):
597 """Add the progress hook (currently only for the file downloader)"""
598 self._progress_hooks.append(ph)
599
600 def _bidi_workaround(self, message):
601 if not hasattr(self, '_output_channel'):
602 return message
603
604 assert hasattr(self, '_output_process')
605 assert isinstance(message, compat_str)
606 line_count = message.count('\n') + 1
607 self._output_process.stdin.write((message + '\n').encode('utf-8'))
608 self._output_process.stdin.flush()
609 res = ''.join(self._output_channel.readline().decode('utf-8')
610 for _ in range(line_count))
611 return res[:-len('\n')]
612
613 def to_screen(self, message, skip_eol=False):
614 """Print message to stdout if not in quiet mode."""
615 return self.to_stdout(message, skip_eol, check_quiet=True)
616
617 def _write_string(self, s, out=None):
618 write_string(s, out=out, encoding=self.params.get('encoding'))
619
620 def to_stdout(self, message, skip_eol=False, check_quiet=False):
621 """Print message to stdout if not in quiet mode."""
622 if self.params.get('logger'):
623 self.params['logger'].debug(message)
624 elif not check_quiet or not self.params.get('quiet', False):
625 message = self._bidi_workaround(message)
626 terminator = ['\n', ''][skip_eol]
627 output = message + terminator
628
629 self._write_string(output, self._screen_file)
630
631 def to_stderr(self, message):
632 """Print message to stderr."""
633 assert isinstance(message, compat_str)
634 if self.params.get('logger'):
635 self.params['logger'].error(message)
636 else:
637 message = self._bidi_workaround(message)
638 output = message + '\n'
639 self._write_string(output, self._err_file)
640
641 def to_console_title(self, message):
642 if not self.params.get('consoletitle', False):
643 return
644 if compat_os_name == 'nt':
645 if ctypes.windll.kernel32.GetConsoleWindow():
646 # c_wchar_p() might not be necessary if `message` is
647 # already of type unicode()
648 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
649 elif 'TERM' in os.environ:
650 self._write_string('\033]0;%s\007' % message, self._screen_file)
651
652 def save_console_title(self):
653 if not self.params.get('consoletitle', False):
654 return
655 if self.params.get('simulate', False):
656 return
657 if compat_os_name != 'nt' and 'TERM' in os.environ:
658 # Save the title on stack
659 self._write_string('\033[22;0t', self._screen_file)
660
661 def restore_console_title(self):
662 if not self.params.get('consoletitle', False):
663 return
664 if self.params.get('simulate', False):
665 return
666 if compat_os_name != 'nt' and 'TERM' in os.environ:
667 # Restore the title from stack
668 self._write_string('\033[23;0t', self._screen_file)
669
670 def __enter__(self):
671 self.save_console_title()
672 return self
673
674 def __exit__(self, *args):
675 self.restore_console_title()
676
677 if self.params.get('cookiefile') is not None:
678 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
679
680 def trouble(self, message=None, tb=None):
681 """Determine action to take when a download problem appears.
682
683 Depending on if the downloader has been configured to ignore
684 download errors or not, this method may throw an exception or
685 not when errors are found, after printing the message.
686
687 tb, if given, is additional traceback information.
688 """
689 if message is not None:
690 self.to_stderr(message)
691 if self.params.get('verbose'):
692 if tb is None:
693 if sys.exc_info()[0]: # if .trouble has been called from an except block
694 tb = ''
695 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
696 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
697 tb += encode_compat_str(traceback.format_exc())
698 else:
699 tb_data = traceback.format_list(traceback.extract_stack())
700 tb = ''.join(tb_data)
701 self.to_stderr(tb)
702 if not self.params.get('ignoreerrors', False):
703 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
704 exc_info = sys.exc_info()[1].exc_info
705 else:
706 exc_info = sys.exc_info()
707 raise DownloadError(message, exc_info)
708 self._download_retcode = 1
709
710 def report_warning(self, message):
711 '''
712 Print the message to stderr, it will be prefixed with 'WARNING:'
713 If stderr is a tty file the 'WARNING:' will be colored
714 '''
715 if self.params.get('logger') is not None:
716 self.params['logger'].warning(message)
717 else:
718 if self.params.get('no_warnings'):
719 return
720 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
721 _msg_header = '\033[0;33mWARNING:\033[0m'
722 else:
723 _msg_header = 'WARNING:'
724 warning_message = '%s %s' % (_msg_header, message)
725 self.to_stderr(warning_message)
726
727 def report_error(self, message, tb=None):
728 '''
729 Do the same as trouble, but prefixes the message with 'ERROR:', colored
730 in red if stderr is a tty file.
731 '''
732 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
733 _msg_header = '\033[0;31mERROR:\033[0m'
734 else:
735 _msg_header = 'ERROR:'
736 error_message = '%s %s' % (_msg_header, message)
737 self.trouble(error_message, tb)
738
739 def report_file_already_downloaded(self, file_name):
740 """Report file has already been fully downloaded."""
741 try:
742 self.to_screen('[download] %s has already been downloaded' % file_name)
743 except UnicodeEncodeError:
744 self.to_screen('[download] The file has already been downloaded')
745
746 def report_file_delete(self, file_name):
747 """Report that existing file will be deleted."""
748 try:
749 self.to_screen('Deleting existing file %s' % file_name)
750 except UnicodeEncodeError:
751 self.to_screen('Deleting existing file')
752
753 def parse_outtmpl(self):
754 outtmpl_dict = self.params.get('outtmpl', {})
755 if not isinstance(outtmpl_dict, dict):
756 outtmpl_dict = {'default': outtmpl_dict}
757 outtmpl_dict.update({
758 k: v for k, v in DEFAULT_OUTTMPL.items()
759 if not outtmpl_dict.get(k)})
760 for key, val in outtmpl_dict.items():
761 if isinstance(val, bytes):
762 self.report_warning(
763 'Parameter outtmpl is bytes, but should be a unicode string. '
764 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
765 return outtmpl_dict
766
767 def _prepare_filename(self, info_dict, tmpl_type='default'):
768 try:
769 template_dict = dict(info_dict)
770
771 template_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
772 formatSeconds(info_dict['duration'], '-')
773 if info_dict.get('duration', None) is not None
774 else None)
775
776 template_dict['epoch'] = int(time.time())
777 autonumber_size = self.params.get('autonumber_size')
778 if autonumber_size is None:
779 autonumber_size = 5
780 template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
781 if template_dict.get('resolution') is None:
782 if template_dict.get('width') and template_dict.get('height'):
783 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
784 elif template_dict.get('height'):
785 template_dict['resolution'] = '%sp' % template_dict['height']
786 elif template_dict.get('width'):
787 template_dict['resolution'] = '%dx?' % template_dict['width']
788
789 sanitize = lambda k, v: sanitize_filename(
790 compat_str(v),
791 restricted=self.params.get('restrictfilenames'),
792 is_id=(k == 'id' or k.endswith('_id')))
793 template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
794 for k, v in template_dict.items()
795 if v is not None and not isinstance(v, (list, tuple, dict)))
796 na = self.params.get('outtmpl_na_placeholder', 'NA')
797 template_dict = collections.defaultdict(lambda: na, template_dict)
798
799 outtmpl = self.outtmpl_dict.get(tmpl_type, self.outtmpl_dict['default'])
800 force_ext = OUTTMPL_TYPES.get(tmpl_type)
801
802 # For fields playlist_index and autonumber convert all occurrences
803 # of %(field)s to %(field)0Nd for backward compatibility
804 field_size_compat_map = {
805 'playlist_index': len(str(template_dict['n_entries'])),
806 'autonumber': autonumber_size,
807 }
808 FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
809 mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
810 if mobj:
811 outtmpl = re.sub(
812 FIELD_SIZE_COMPAT_RE,
813 r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
814 outtmpl)
815
816 # As of [1] format syntax is:
817 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
818 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
819 FORMAT_RE = r'''(?x)
820 (?<!%)
821 %
822 \({0}\) # mapping key
823 (?:[#0\-+ ]+)? # conversion flags (optional)
824 (?:\d+)? # minimum field width (optional)
825 (?:\.\d+)? # precision (optional)
826 [hlL]? # length modifier (optional)
827 (?P<type>[diouxXeEfFgGcrs%]) # conversion type
828 '''
829
830 numeric_fields = list(self._NUMERIC_FIELDS)
831
832 # Format date
833 FORMAT_DATE_RE = FORMAT_RE.format(r'(?P<key>(?P<field>\w+)>(?P<format>.+?))')
834 for mobj in re.finditer(FORMAT_DATE_RE, outtmpl):
835 conv_type, field, frmt, key = mobj.group('type', 'field', 'format', 'key')
836 if key in template_dict:
837 continue
838 value = strftime_or_none(template_dict.get(field), frmt, na)
839 if conv_type in 'crs': # string
840 value = sanitize(field, value)
841 else: # number
842 numeric_fields.append(key)
843 value = float_or_none(value, default=None)
844 if value is not None:
845 template_dict[key] = value
846
847 # Missing numeric fields used together with integer presentation types
848 # in format specification will break the argument substitution since
849 # string NA placeholder is returned for missing fields. We will patch
850 # output template for missing fields to meet string presentation type.
851 for numeric_field in numeric_fields:
852 if numeric_field not in template_dict:
853 outtmpl = re.sub(
854 FORMAT_RE.format(re.escape(numeric_field)),
855 r'%({0})s'.format(numeric_field), outtmpl)
856
857 # expand_path translates '%%' into '%' and '$$' into '$'
858 # correspondingly that is not what we want since we need to keep
859 # '%%' intact for template dict substitution step. Working around
860 # with boundary-alike separator hack.
861 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
862 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
863
864 # outtmpl should be expand_path'ed before template dict substitution
865 # because meta fields may contain env variables we don't want to
866 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
867 # title "Hello $PATH", we don't want `$PATH` to be expanded.
868 filename = expand_path(outtmpl).replace(sep, '') % template_dict
869
870 if force_ext is not None:
871 filename = replace_extension(filename, force_ext, template_dict.get('ext'))
872
873 # https://github.com/blackjack4494/youtube-dlc/issues/85
874 trim_file_name = self.params.get('trim_file_name', False)
875 if trim_file_name:
876 fn_groups = filename.rsplit('.')
877 ext = fn_groups[-1]
878 sub_ext = ''
879 if len(fn_groups) > 2:
880 sub_ext = fn_groups[-2]
881 filename = '.'.join(filter(None, [fn_groups[0][:trim_file_name], sub_ext, ext]))
882
883 return filename
884 except ValueError as err:
885 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
886 return None
887
888 def prepare_filename(self, info_dict, dir_type='', warn=False):
889 """Generate the output filename."""
890 paths = self.params.get('paths', {})
891 assert isinstance(paths, dict)
892 filename = self._prepare_filename(info_dict, dir_type or 'default')
893
894 if warn and not self.__prepare_filename_warned:
895 if not paths:
896 pass
897 elif filename == '-':
898 self.report_warning('--paths is ignored when an outputting to stdout')
899 elif os.path.isabs(filename):
900 self.report_warning('--paths is ignored since an absolute path is given in output template')
901 self.__prepare_filename_warned = True
902 if filename == '-' or not filename:
903 return filename
904
905 homepath = expand_path(paths.get('home', '').strip())
906 assert isinstance(homepath, compat_str)
907 subdir = expand_path(paths.get(dir_type, '').strip()) if dir_type else ''
908 assert isinstance(subdir, compat_str)
909 path = os.path.join(homepath, subdir, filename)
910
911 # Temporary fix for #4787
912 # 'Treat' all problem characters by passing filename through preferredencoding
913 # to workaround encoding issues with subprocess on python2 @ Windows
914 if sys.version_info < (3, 0) and sys.platform == 'win32':
915 path = encodeFilename(path, True).decode(preferredencoding())
916 return sanitize_path(path, force=self.params.get('windowsfilenames'))
917
918 def _match_entry(self, info_dict, incomplete):
919 """ Returns None if the file should be downloaded """
920
921 def check_filter():
922 video_title = info_dict.get('title', info_dict.get('id', 'video'))
923 if 'title' in info_dict:
924 # This can happen when we're just evaluating the playlist
925 title = info_dict['title']
926 matchtitle = self.params.get('matchtitle', False)
927 if matchtitle:
928 if not re.search(matchtitle, title, re.IGNORECASE):
929 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
930 rejecttitle = self.params.get('rejecttitle', False)
931 if rejecttitle:
932 if re.search(rejecttitle, title, re.IGNORECASE):
933 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
934 date = info_dict.get('upload_date')
935 if date is not None:
936 dateRange = self.params.get('daterange', DateRange())
937 if date not in dateRange:
938 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
939 view_count = info_dict.get('view_count')
940 if view_count is not None:
941 min_views = self.params.get('min_views')
942 if min_views is not None and view_count < min_views:
943 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
944 max_views = self.params.get('max_views')
945 if max_views is not None and view_count > max_views:
946 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
947 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
948 return 'Skipping "%s" because it is age restricted' % video_title
949 if self.in_download_archive(info_dict):
950 return '%s has already been recorded in archive' % video_title
951
952 if not incomplete:
953 match_filter = self.params.get('match_filter')
954 if match_filter is not None:
955 ret = match_filter(info_dict)
956 if ret is not None:
957 return ret
958 return None
959
960 reason = check_filter()
961 if reason is not None:
962 self.to_screen('[download] ' + reason)
963 if reason.endswith('has already been recorded in the archive') and self.params.get('break_on_existing', False):
964 raise ExistingVideoReached()
965 elif self.params.get('break_on_reject', False):
966 raise RejectedVideoReached()
967 return reason
968
969 @staticmethod
970 def add_extra_info(info_dict, extra_info):
971 '''Set the keys from extra_info in info dict if they are missing'''
972 for key, value in extra_info.items():
973 info_dict.setdefault(key, value)
974
975 def extract_info(self, url, download=True, ie_key=None, info_dict=None, extra_info={},
976 process=True, force_generic_extractor=False):
977 '''
978 Returns a list with a dictionary for each video we find.
979 If 'download', also downloads the videos.
980 extra_info is a dict containing the extra values to add to each result
981 '''
982
983 if not ie_key and force_generic_extractor:
984 ie_key = 'Generic'
985
986 if ie_key:
987 ies = [self.get_info_extractor(ie_key)]
988 else:
989 ies = self._ies
990
991 for ie in ies:
992 if not ie.suitable(url):
993 continue
994
995 ie_key = ie.ie_key()
996 ie = self.get_info_extractor(ie_key)
997 if not ie.working():
998 self.report_warning('The program functionality for this site has been marked as broken, '
999 'and will probably not work.')
1000
1001 try:
1002 temp_id = str_or_none(
1003 ie.extract_id(url) if callable(getattr(ie, 'extract_id', None))
1004 else ie._match_id(url))
1005 except (AssertionError, IndexError, AttributeError):
1006 temp_id = None
1007 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1008 self.to_screen("[%s] %s: has already been recorded in archive" % (
1009 ie_key, temp_id))
1010 break
1011 return self.__extract_info(url, ie, download, extra_info, process, info_dict)
1012 else:
1013 self.report_error('no suitable InfoExtractor for URL %s' % url)
1014
1015 def __handle_extraction_exceptions(func):
1016 def wrapper(self, *args, **kwargs):
1017 try:
1018 return func(self, *args, **kwargs)
1019 except GeoRestrictedError as e:
1020 msg = e.msg
1021 if e.countries:
1022 msg += '\nThis video is available in %s.' % ', '.join(
1023 map(ISO3166Utils.short2full, e.countries))
1024 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1025 self.report_error(msg)
1026 except ExtractorError as e: # An error we somewhat expected
1027 self.report_error(compat_str(e), e.format_traceback())
1028 except (MaxDownloadsReached, ExistingVideoReached, RejectedVideoReached):
1029 raise
1030 except Exception as e:
1031 if self.params.get('ignoreerrors', False):
1032 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
1033 else:
1034 raise
1035 return wrapper
1036
1037 @__handle_extraction_exceptions
1038 def __extract_info(self, url, ie, download, extra_info, process, info_dict):
1039 ie_result = ie.extract(url)
1040 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1041 return
1042 if isinstance(ie_result, list):
1043 # Backwards compatibility: old IE result format
1044 ie_result = {
1045 '_type': 'compat_list',
1046 'entries': ie_result,
1047 }
1048 if info_dict:
1049 if info_dict.get('id'):
1050 ie_result['id'] = info_dict['id']
1051 if info_dict.get('title'):
1052 ie_result['title'] = info_dict['title']
1053 self.add_default_extra_info(ie_result, ie, url)
1054 if process:
1055 return self.process_ie_result(ie_result, download, extra_info)
1056 else:
1057 return ie_result
1058
1059 def add_default_extra_info(self, ie_result, ie, url):
1060 self.add_extra_info(ie_result, {
1061 'extractor': ie.IE_NAME,
1062 'webpage_url': url,
1063 'webpage_url_basename': url_basename(url),
1064 'extractor_key': ie.ie_key(),
1065 })
1066
1067 def process_ie_result(self, ie_result, download=True, extra_info={}):
1068 """
1069 Take the result of the ie(may be modified) and resolve all unresolved
1070 references (URLs, playlist items).
1071
1072 It will also download the videos if 'download'.
1073 Returns the resolved ie_result.
1074 """
1075 result_type = ie_result.get('_type', 'video')
1076
1077 if result_type in ('url', 'url_transparent'):
1078 ie_result['url'] = sanitize_url(ie_result['url'])
1079 extract_flat = self.params.get('extract_flat', False)
1080 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1081 or extract_flat is True):
1082 self.__forced_printings(ie_result, self.prepare_filename(ie_result), incomplete=True)
1083 return ie_result
1084
1085 if result_type == 'video':
1086 self.add_extra_info(ie_result, extra_info)
1087 return self.process_video_result(ie_result, download=download)
1088 elif result_type == 'url':
1089 # We have to add extra_info to the results because it may be
1090 # contained in a playlist
1091 return self.extract_info(ie_result['url'],
1092 download, info_dict=ie_result,
1093 ie_key=ie_result.get('ie_key'),
1094 extra_info=extra_info)
1095 elif result_type == 'url_transparent':
1096 # Use the information from the embedding page
1097 info = self.extract_info(
1098 ie_result['url'], ie_key=ie_result.get('ie_key'),
1099 extra_info=extra_info, download=False, process=False)
1100
1101 # extract_info may return None when ignoreerrors is enabled and
1102 # extraction failed with an error, don't crash and return early
1103 # in this case
1104 if not info:
1105 return info
1106
1107 force_properties = dict(
1108 (k, v) for k, v in ie_result.items() if v is not None)
1109 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1110 if f in force_properties:
1111 del force_properties[f]
1112 new_result = info.copy()
1113 new_result.update(force_properties)
1114
1115 # Extracted info may not be a video result (i.e.
1116 # info.get('_type', 'video') != video) but rather an url or
1117 # url_transparent. In such cases outer metadata (from ie_result)
1118 # should be propagated to inner one (info). For this to happen
1119 # _type of info should be overridden with url_transparent. This
1120 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1121 if new_result.get('_type') == 'url':
1122 new_result['_type'] = 'url_transparent'
1123
1124 return self.process_ie_result(
1125 new_result, download=download, extra_info=extra_info)
1126 elif result_type in ('playlist', 'multi_video'):
1127 # Protect from infinite recursion due to recursively nested playlists
1128 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1129 webpage_url = ie_result['webpage_url']
1130 if webpage_url in self._playlist_urls:
1131 self.to_screen(
1132 '[download] Skipping already downloaded playlist: %s'
1133 % ie_result.get('title') or ie_result.get('id'))
1134 return
1135
1136 self._playlist_level += 1
1137 self._playlist_urls.add(webpage_url)
1138 try:
1139 return self.__process_playlist(ie_result, download)
1140 finally:
1141 self._playlist_level -= 1
1142 if not self._playlist_level:
1143 self._playlist_urls.clear()
1144 elif result_type == 'compat_list':
1145 self.report_warning(
1146 'Extractor %s returned a compat_list result. '
1147 'It needs to be updated.' % ie_result.get('extractor'))
1148
1149 def _fixup(r):
1150 self.add_extra_info(
1151 r,
1152 {
1153 'extractor': ie_result['extractor'],
1154 'webpage_url': ie_result['webpage_url'],
1155 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1156 'extractor_key': ie_result['extractor_key'],
1157 }
1158 )
1159 return r
1160 ie_result['entries'] = [
1161 self.process_ie_result(_fixup(r), download, extra_info)
1162 for r in ie_result['entries']
1163 ]
1164 return ie_result
1165 else:
1166 raise Exception('Invalid result type: %s' % result_type)
1167
1168 def __process_playlist(self, ie_result, download):
1169 # We process each entry in the playlist
1170 playlist = ie_result.get('title') or ie_result.get('id')
1171 self.to_screen('[download] Downloading playlist: %s' % playlist)
1172
1173 if self.params.get('allow_playlist_files', True):
1174 ie_copy = {
1175 'playlist': playlist,
1176 'playlist_id': ie_result.get('id'),
1177 'playlist_title': ie_result.get('title'),
1178 'playlist_uploader': ie_result.get('uploader'),
1179 'playlist_uploader_id': ie_result.get('uploader_id'),
1180 'playlist_index': 0
1181 }
1182 ie_copy.update(dict(ie_result))
1183
1184 def ensure_dir_exists(path):
1185 return make_dir(path, self.report_error)
1186
1187 if self.params.get('writeinfojson', False):
1188 infofn = self.prepare_filename(ie_copy, 'pl_infojson')
1189 if not ensure_dir_exists(encodeFilename(infofn)):
1190 return
1191 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
1192 self.to_screen('[info] Playlist metadata is already present')
1193 else:
1194 playlist_info = dict(ie_result)
1195 # playlist_info['entries'] = list(playlist_info['entries']) # Entries is a generator which shouldnot be resolved here
1196 del playlist_info['entries']
1197 self.to_screen('[info] Writing playlist metadata as JSON to: ' + infofn)
1198 try:
1199 write_json_file(self.filter_requested_info(playlist_info), infofn)
1200 except (OSError, IOError):
1201 self.report_error('Cannot write playlist metadata to JSON file ' + infofn)
1202
1203 if self.params.get('writedescription', False):
1204 descfn = self.prepare_filename(ie_copy, 'pl_description')
1205 if not ensure_dir_exists(encodeFilename(descfn)):
1206 return
1207 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
1208 self.to_screen('[info] Playlist description is already present')
1209 elif ie_result.get('description') is None:
1210 self.report_warning('There\'s no playlist description to write.')
1211 else:
1212 try:
1213 self.to_screen('[info] Writing playlist description to: ' + descfn)
1214 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1215 descfile.write(ie_result['description'])
1216 except (OSError, IOError):
1217 self.report_error('Cannot write playlist description file ' + descfn)
1218 return
1219
1220 playlist_results = []
1221
1222 playliststart = self.params.get('playliststart', 1) - 1
1223 playlistend = self.params.get('playlistend')
1224 # For backwards compatibility, interpret -1 as whole list
1225 if playlistend == -1:
1226 playlistend = None
1227
1228 playlistitems_str = self.params.get('playlist_items')
1229 playlistitems = None
1230 if playlistitems_str is not None:
1231 def iter_playlistitems(format):
1232 for string_segment in format.split(','):
1233 if '-' in string_segment:
1234 start, end = string_segment.split('-')
1235 for item in range(int(start), int(end) + 1):
1236 yield int(item)
1237 else:
1238 yield int(string_segment)
1239 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1240
1241 ie_entries = ie_result['entries']
1242
1243 def make_playlistitems_entries(list_ie_entries):
1244 num_entries = len(list_ie_entries)
1245 return [
1246 list_ie_entries[i - 1] for i in playlistitems
1247 if -num_entries <= i - 1 < num_entries]
1248
1249 def report_download(num_entries):
1250 self.to_screen(
1251 '[%s] playlist %s: Downloading %d videos' %
1252 (ie_result['extractor'], playlist, num_entries))
1253
1254 if isinstance(ie_entries, list):
1255 n_all_entries = len(ie_entries)
1256 if playlistitems:
1257 entries = make_playlistitems_entries(ie_entries)
1258 else:
1259 entries = ie_entries[playliststart:playlistend]
1260 n_entries = len(entries)
1261 self.to_screen(
1262 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
1263 (ie_result['extractor'], playlist, n_all_entries, n_entries))
1264 elif isinstance(ie_entries, PagedList):
1265 if playlistitems:
1266 entries = []
1267 for item in playlistitems:
1268 entries.extend(ie_entries.getslice(
1269 item - 1, item
1270 ))
1271 else:
1272 entries = ie_entries.getslice(
1273 playliststart, playlistend)
1274 n_entries = len(entries)
1275 report_download(n_entries)
1276 else: # iterable
1277 if playlistitems:
1278 entries = make_playlistitems_entries(list(itertools.islice(
1279 ie_entries, 0, max(playlistitems))))
1280 else:
1281 entries = list(itertools.islice(
1282 ie_entries, playliststart, playlistend))
1283 n_entries = len(entries)
1284 report_download(n_entries)
1285
1286 if self.params.get('playlistreverse', False):
1287 entries = entries[::-1]
1288
1289 if self.params.get('playlistrandom', False):
1290 random.shuffle(entries)
1291
1292 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1293
1294 for i, entry in enumerate(entries, 1):
1295 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1296 # This __x_forwarded_for_ip thing is a bit ugly but requires
1297 # minimal changes
1298 if x_forwarded_for:
1299 entry['__x_forwarded_for_ip'] = x_forwarded_for
1300 extra = {
1301 'n_entries': n_entries,
1302 'playlist': playlist,
1303 'playlist_id': ie_result.get('id'),
1304 'playlist_title': ie_result.get('title'),
1305 'playlist_uploader': ie_result.get('uploader'),
1306 'playlist_uploader_id': ie_result.get('uploader_id'),
1307 'playlist_index': playlistitems[i - 1] if playlistitems else i + playliststart,
1308 'extractor': ie_result['extractor'],
1309 'webpage_url': ie_result['webpage_url'],
1310 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1311 'extractor_key': ie_result['extractor_key'],
1312 }
1313
1314 if self._match_entry(entry, incomplete=True) is not None:
1315 continue
1316
1317 entry_result = self.__process_iterable_entry(entry, download, extra)
1318 # TODO: skip failed (empty) entries?
1319 playlist_results.append(entry_result)
1320 ie_result['entries'] = playlist_results
1321 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1322 return ie_result
1323
1324 @__handle_extraction_exceptions
1325 def __process_iterable_entry(self, entry, download, extra_info):
1326 return self.process_ie_result(
1327 entry, download=download, extra_info=extra_info)
1328
1329 def _build_format_filter(self, filter_spec):
1330 " Returns a function to filter the formats according to the filter_spec "
1331
1332 OPERATORS = {
1333 '<': operator.lt,
1334 '<=': operator.le,
1335 '>': operator.gt,
1336 '>=': operator.ge,
1337 '=': operator.eq,
1338 '!=': operator.ne,
1339 }
1340 operator_rex = re.compile(r'''(?x)\s*
1341 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)
1342 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1343 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
1344 $
1345 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1346 m = operator_rex.search(filter_spec)
1347 if m:
1348 try:
1349 comparison_value = int(m.group('value'))
1350 except ValueError:
1351 comparison_value = parse_filesize(m.group('value'))
1352 if comparison_value is None:
1353 comparison_value = parse_filesize(m.group('value') + 'B')
1354 if comparison_value is None:
1355 raise ValueError(
1356 'Invalid value %r in format specification %r' % (
1357 m.group('value'), filter_spec))
1358 op = OPERATORS[m.group('op')]
1359
1360 if not m:
1361 STR_OPERATORS = {
1362 '=': operator.eq,
1363 '^=': lambda attr, value: attr.startswith(value),
1364 '$=': lambda attr, value: attr.endswith(value),
1365 '*=': lambda attr, value: value in attr,
1366 }
1367 str_operator_rex = re.compile(r'''(?x)
1368 \s*(?P<key>[a-zA-Z0-9._-]+)
1369 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?
1370 \s*(?P<value>[a-zA-Z0-9._-]+)
1371 \s*$
1372 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1373 m = str_operator_rex.search(filter_spec)
1374 if m:
1375 comparison_value = m.group('value')
1376 str_op = STR_OPERATORS[m.group('op')]
1377 if m.group('negation'):
1378 op = lambda attr, value: not str_op(attr, value)
1379 else:
1380 op = str_op
1381
1382 if not m:
1383 raise ValueError('Invalid filter specification %r' % filter_spec)
1384
1385 def _filter(f):
1386 actual_value = f.get(m.group('key'))
1387 if actual_value is None:
1388 return m.group('none_inclusive')
1389 return op(actual_value, comparison_value)
1390 return _filter
1391
1392 def _default_format_spec(self, info_dict, download=True):
1393
1394 def can_merge():
1395 merger = FFmpegMergerPP(self)
1396 return merger.available and merger.can_merge()
1397
1398 prefer_best = (
1399 not self.params.get('simulate', False)
1400 and download
1401 and (
1402 not can_merge()
1403 or info_dict.get('is_live', False)
1404 or self.outtmpl_dict['default'] == '-'))
1405
1406 return (
1407 'best/bestvideo+bestaudio'
1408 if prefer_best
1409 else 'bestvideo*+bestaudio/best'
1410 if not self.params.get('allow_multiple_audio_streams', False)
1411 else 'bestvideo+bestaudio/best')
1412
1413 def build_format_selector(self, format_spec):
1414 def syntax_error(note, start):
1415 message = (
1416 'Invalid format specification: '
1417 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1418 return SyntaxError(message)
1419
1420 PICKFIRST = 'PICKFIRST'
1421 MERGE = 'MERGE'
1422 SINGLE = 'SINGLE'
1423 GROUP = 'GROUP'
1424 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1425
1426 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1427 'video': self.params.get('allow_multiple_video_streams', False)}
1428
1429 def _parse_filter(tokens):
1430 filter_parts = []
1431 for type, string, start, _, _ in tokens:
1432 if type == tokenize.OP and string == ']':
1433 return ''.join(filter_parts)
1434 else:
1435 filter_parts.append(string)
1436
1437 def _remove_unused_ops(tokens):
1438 # Remove operators that we don't use and join them with the surrounding strings
1439 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1440 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1441 last_string, last_start, last_end, last_line = None, None, None, None
1442 for type, string, start, end, line in tokens:
1443 if type == tokenize.OP and string == '[':
1444 if last_string:
1445 yield tokenize.NAME, last_string, last_start, last_end, last_line
1446 last_string = None
1447 yield type, string, start, end, line
1448 # everything inside brackets will be handled by _parse_filter
1449 for type, string, start, end, line in tokens:
1450 yield type, string, start, end, line
1451 if type == tokenize.OP and string == ']':
1452 break
1453 elif type == tokenize.OP and string in ALLOWED_OPS:
1454 if last_string:
1455 yield tokenize.NAME, last_string, last_start, last_end, last_line
1456 last_string = None
1457 yield type, string, start, end, line
1458 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1459 if not last_string:
1460 last_string = string
1461 last_start = start
1462 last_end = end
1463 else:
1464 last_string += string
1465 if last_string:
1466 yield tokenize.NAME, last_string, last_start, last_end, last_line
1467
1468 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1469 selectors = []
1470 current_selector = None
1471 for type, string, start, _, _ in tokens:
1472 # ENCODING is only defined in python 3.x
1473 if type == getattr(tokenize, 'ENCODING', None):
1474 continue
1475 elif type in [tokenize.NAME, tokenize.NUMBER]:
1476 current_selector = FormatSelector(SINGLE, string, [])
1477 elif type == tokenize.OP:
1478 if string == ')':
1479 if not inside_group:
1480 # ')' will be handled by the parentheses group
1481 tokens.restore_last_token()
1482 break
1483 elif inside_merge and string in ['/', ',']:
1484 tokens.restore_last_token()
1485 break
1486 elif inside_choice and string == ',':
1487 tokens.restore_last_token()
1488 break
1489 elif string == ',':
1490 if not current_selector:
1491 raise syntax_error('"," must follow a format selector', start)
1492 selectors.append(current_selector)
1493 current_selector = None
1494 elif string == '/':
1495 if not current_selector:
1496 raise syntax_error('"/" must follow a format selector', start)
1497 first_choice = current_selector
1498 second_choice = _parse_format_selection(tokens, inside_choice=True)
1499 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1500 elif string == '[':
1501 if not current_selector:
1502 current_selector = FormatSelector(SINGLE, 'best', [])
1503 format_filter = _parse_filter(tokens)
1504 current_selector.filters.append(format_filter)
1505 elif string == '(':
1506 if current_selector:
1507 raise syntax_error('Unexpected "("', start)
1508 group = _parse_format_selection(tokens, inside_group=True)
1509 current_selector = FormatSelector(GROUP, group, [])
1510 elif string == '+':
1511 if not current_selector:
1512 raise syntax_error('Unexpected "+"', start)
1513 selector_1 = current_selector
1514 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1515 if not selector_2:
1516 raise syntax_error('Expected a selector', start)
1517 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1518 else:
1519 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1520 elif type == tokenize.ENDMARKER:
1521 break
1522 if current_selector:
1523 selectors.append(current_selector)
1524 return selectors
1525
1526 def _build_selector_function(selector):
1527 if isinstance(selector, list): # ,
1528 fs = [_build_selector_function(s) for s in selector]
1529
1530 def selector_function(ctx):
1531 for f in fs:
1532 for format in f(ctx):
1533 yield format
1534 return selector_function
1535
1536 elif selector.type == GROUP: # ()
1537 selector_function = _build_selector_function(selector.selector)
1538
1539 elif selector.type == PICKFIRST: # /
1540 fs = [_build_selector_function(s) for s in selector.selector]
1541
1542 def selector_function(ctx):
1543 for f in fs:
1544 picked_formats = list(f(ctx))
1545 if picked_formats:
1546 return picked_formats
1547 return []
1548
1549 elif selector.type == SINGLE: # atom
1550 format_spec = selector.selector if selector.selector is not None else 'best'
1551
1552 if format_spec == 'all':
1553 def selector_function(ctx):
1554 formats = list(ctx['formats'])
1555 if formats:
1556 for f in formats:
1557 yield f
1558
1559 else:
1560 format_fallback = False
1561 format_spec_obj = re.match(r'(best|worst|b|w)(video|audio|v|a)?(\*)?$', format_spec)
1562 if format_spec_obj is not None:
1563 format_idx = 0 if format_spec_obj.group(1)[0] == 'w' else -1
1564 format_type = format_spec_obj.group(2)[0] if format_spec_obj.group(2) else False
1565 not_format_type = 'v' if format_type == 'a' else 'a'
1566 format_modified = format_spec_obj.group(3) is not None
1567
1568 format_fallback = not format_type and not format_modified # for b, w
1569 filter_f = ((lambda f: f.get(format_type + 'codec') != 'none')
1570 if format_type and format_modified # bv*, ba*, wv*, wa*
1571 else (lambda f: f.get(not_format_type + 'codec') == 'none')
1572 if format_type # bv, ba, wv, wa
1573 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
1574 if not format_modified # b, w
1575 else None) # b*, w*
1576 else:
1577 format_idx = -1
1578 filter_f = ((lambda f: f.get('ext') == format_spec)
1579 if format_spec in ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] # extension
1580 else (lambda f: f.get('format_id') == format_spec)) # id
1581
1582 def selector_function(ctx):
1583 formats = list(ctx['formats'])
1584 if not formats:
1585 return
1586 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
1587 if matches:
1588 yield matches[format_idx]
1589 elif format_fallback == 'force' or (format_fallback and ctx['incomplete_formats']):
1590 # for extractors with incomplete formats (audio only (soundcloud)
1591 # or video only (imgur)) best/worst will fallback to
1592 # best/worst {video,audio}-only format
1593 yield formats[format_idx]
1594
1595 elif selector.type == MERGE: # +
1596 def _merge(formats_pair):
1597 format_1, format_2 = formats_pair
1598
1599 formats_info = []
1600 formats_info.extend(format_1.get('requested_formats', (format_1,)))
1601 formats_info.extend(format_2.get('requested_formats', (format_2,)))
1602
1603 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
1604 get_no_more = {"video": False, "audio": False}
1605 for (i, fmt_info) in enumerate(formats_info):
1606 for aud_vid in ["audio", "video"]:
1607 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
1608 if get_no_more[aud_vid]:
1609 formats_info.pop(i)
1610 get_no_more[aud_vid] = True
1611
1612 if len(formats_info) == 1:
1613 return formats_info[0]
1614
1615 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
1616 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
1617
1618 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
1619 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
1620
1621 output_ext = self.params.get('merge_output_format')
1622 if not output_ext:
1623 if the_only_video:
1624 output_ext = the_only_video['ext']
1625 elif the_only_audio and not video_fmts:
1626 output_ext = the_only_audio['ext']
1627 else:
1628 output_ext = 'mkv'
1629
1630 new_dict = {
1631 'requested_formats': formats_info,
1632 'format': '+'.join(fmt_info.get('format') for fmt_info in formats_info),
1633 'format_id': '+'.join(fmt_info.get('format_id') for fmt_info in formats_info),
1634 'ext': output_ext,
1635 }
1636
1637 if the_only_video:
1638 new_dict.update({
1639 'width': the_only_video.get('width'),
1640 'height': the_only_video.get('height'),
1641 'resolution': the_only_video.get('resolution'),
1642 'fps': the_only_video.get('fps'),
1643 'vcodec': the_only_video.get('vcodec'),
1644 'vbr': the_only_video.get('vbr'),
1645 'stretched_ratio': the_only_video.get('stretched_ratio'),
1646 })
1647
1648 if the_only_audio:
1649 new_dict.update({
1650 'acodec': the_only_audio.get('acodec'),
1651 'abr': the_only_audio.get('abr'),
1652 })
1653
1654 return new_dict
1655
1656 selector_1, selector_2 = map(_build_selector_function, selector.selector)
1657
1658 def selector_function(ctx):
1659 for pair in itertools.product(
1660 selector_1(copy.deepcopy(ctx)), selector_2(copy.deepcopy(ctx))):
1661 yield _merge(pair)
1662
1663 filters = [self._build_format_filter(f) for f in selector.filters]
1664
1665 def final_selector(ctx):
1666 ctx_copy = copy.deepcopy(ctx)
1667 for _filter in filters:
1668 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1669 return selector_function(ctx_copy)
1670 return final_selector
1671
1672 stream = io.BytesIO(format_spec.encode('utf-8'))
1673 try:
1674 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1675 except tokenize.TokenError:
1676 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1677
1678 class TokenIterator(object):
1679 def __init__(self, tokens):
1680 self.tokens = tokens
1681 self.counter = 0
1682
1683 def __iter__(self):
1684 return self
1685
1686 def __next__(self):
1687 if self.counter >= len(self.tokens):
1688 raise StopIteration()
1689 value = self.tokens[self.counter]
1690 self.counter += 1
1691 return value
1692
1693 next = __next__
1694
1695 def restore_last_token(self):
1696 self.counter -= 1
1697
1698 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1699 return _build_selector_function(parsed_selector)
1700
1701 def _calc_headers(self, info_dict):
1702 res = std_headers.copy()
1703
1704 add_headers = info_dict.get('http_headers')
1705 if add_headers:
1706 res.update(add_headers)
1707
1708 cookies = self._calc_cookies(info_dict)
1709 if cookies:
1710 res['Cookie'] = cookies
1711
1712 if 'X-Forwarded-For' not in res:
1713 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1714 if x_forwarded_for_ip:
1715 res['X-Forwarded-For'] = x_forwarded_for_ip
1716
1717 return res
1718
1719 def _calc_cookies(self, info_dict):
1720 pr = sanitized_Request(info_dict['url'])
1721 self.cookiejar.add_cookie_header(pr)
1722 return pr.get_header('Cookie')
1723
1724 def process_video_result(self, info_dict, download=True):
1725 assert info_dict.get('_type', 'video') == 'video'
1726
1727 if 'id' not in info_dict:
1728 raise ExtractorError('Missing "id" field in extractor result')
1729 if 'title' not in info_dict:
1730 raise ExtractorError('Missing "title" field in extractor result')
1731
1732 def report_force_conversion(field, field_not, conversion):
1733 self.report_warning(
1734 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
1735 % (field, field_not, conversion))
1736
1737 def sanitize_string_field(info, string_field):
1738 field = info.get(string_field)
1739 if field is None or isinstance(field, compat_str):
1740 return
1741 report_force_conversion(string_field, 'a string', 'string')
1742 info[string_field] = compat_str(field)
1743
1744 def sanitize_numeric_fields(info):
1745 for numeric_field in self._NUMERIC_FIELDS:
1746 field = info.get(numeric_field)
1747 if field is None or isinstance(field, compat_numeric_types):
1748 continue
1749 report_force_conversion(numeric_field, 'numeric', 'int')
1750 info[numeric_field] = int_or_none(field)
1751
1752 sanitize_string_field(info_dict, 'id')
1753 sanitize_numeric_fields(info_dict)
1754
1755 if 'playlist' not in info_dict:
1756 # It isn't part of a playlist
1757 info_dict['playlist'] = None
1758 info_dict['playlist_index'] = None
1759
1760 thumbnails = info_dict.get('thumbnails')
1761 if thumbnails is None:
1762 thumbnail = info_dict.get('thumbnail')
1763 if thumbnail:
1764 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1765 if thumbnails:
1766 thumbnails.sort(key=lambda t: (
1767 t.get('preference') if t.get('preference') is not None else -1,
1768 t.get('width') if t.get('width') is not None else -1,
1769 t.get('height') if t.get('height') is not None else -1,
1770 t.get('id') if t.get('id') is not None else '', t.get('url')))
1771 for i, t in enumerate(thumbnails):
1772 t['url'] = sanitize_url(t['url'])
1773 if t.get('width') and t.get('height'):
1774 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1775 if t.get('id') is None:
1776 t['id'] = '%d' % i
1777
1778 if self.params.get('list_thumbnails'):
1779 self.list_thumbnails(info_dict)
1780 return
1781
1782 thumbnail = info_dict.get('thumbnail')
1783 if thumbnail:
1784 info_dict['thumbnail'] = sanitize_url(thumbnail)
1785 elif thumbnails:
1786 info_dict['thumbnail'] = thumbnails[-1]['url']
1787
1788 if 'display_id' not in info_dict and 'id' in info_dict:
1789 info_dict['display_id'] = info_dict['id']
1790
1791 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1792 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1793 # see http://bugs.python.org/issue1646728)
1794 try:
1795 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1796 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1797 except (ValueError, OverflowError, OSError):
1798 pass
1799
1800 # Auto generate title fields corresponding to the *_number fields when missing
1801 # in order to always have clean titles. This is very common for TV series.
1802 for field in ('chapter', 'season', 'episode'):
1803 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1804 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1805
1806 for cc_kind in ('subtitles', 'automatic_captions'):
1807 cc = info_dict.get(cc_kind)
1808 if cc:
1809 for _, subtitle in cc.items():
1810 for subtitle_format in subtitle:
1811 if subtitle_format.get('url'):
1812 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1813 if subtitle_format.get('ext') is None:
1814 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1815
1816 automatic_captions = info_dict.get('automatic_captions')
1817 subtitles = info_dict.get('subtitles')
1818
1819 if self.params.get('listsubtitles', False):
1820 if 'automatic_captions' in info_dict:
1821 self.list_subtitles(
1822 info_dict['id'], automatic_captions, 'automatic captions')
1823 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1824 return
1825
1826 info_dict['requested_subtitles'] = self.process_subtitles(
1827 info_dict['id'], subtitles, automatic_captions)
1828
1829 # We now pick which formats have to be downloaded
1830 if info_dict.get('formats') is None:
1831 # There's only one format available
1832 formats = [info_dict]
1833 else:
1834 formats = info_dict['formats']
1835
1836 if not formats:
1837 raise ExtractorError('No video formats found!')
1838
1839 def is_wellformed(f):
1840 url = f.get('url')
1841 if not url:
1842 self.report_warning(
1843 '"url" field is missing or empty - skipping format, '
1844 'there is an error in extractor')
1845 return False
1846 if isinstance(url, bytes):
1847 sanitize_string_field(f, 'url')
1848 return True
1849
1850 # Filter out malformed formats for better extraction robustness
1851 formats = list(filter(is_wellformed, formats))
1852
1853 formats_dict = {}
1854
1855 # We check that all the formats have the format and format_id fields
1856 for i, format in enumerate(formats):
1857 sanitize_string_field(format, 'format_id')
1858 sanitize_numeric_fields(format)
1859 format['url'] = sanitize_url(format['url'])
1860 if not format.get('format_id'):
1861 format['format_id'] = compat_str(i)
1862 else:
1863 # Sanitize format_id from characters used in format selector expression
1864 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
1865 format_id = format['format_id']
1866 if format_id not in formats_dict:
1867 formats_dict[format_id] = []
1868 formats_dict[format_id].append(format)
1869
1870 # Make sure all formats have unique format_id
1871 for format_id, ambiguous_formats in formats_dict.items():
1872 if len(ambiguous_formats) > 1:
1873 for i, format in enumerate(ambiguous_formats):
1874 format['format_id'] = '%s-%d' % (format_id, i)
1875
1876 for i, format in enumerate(formats):
1877 if format.get('format') is None:
1878 format['format'] = '{id} - {res}{note}'.format(
1879 id=format['format_id'],
1880 res=self.format_resolution(format),
1881 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1882 )
1883 # Automatically determine file extension if missing
1884 if format.get('ext') is None:
1885 format['ext'] = determine_ext(format['url']).lower()
1886 # Automatically determine protocol if missing (useful for format
1887 # selection purposes)
1888 if format.get('protocol') is None:
1889 format['protocol'] = determine_protocol(format)
1890 # Add HTTP headers, so that external programs can use them from the
1891 # json output
1892 full_format_info = info_dict.copy()
1893 full_format_info.update(format)
1894 format['http_headers'] = self._calc_headers(full_format_info)
1895 # Remove private housekeeping stuff
1896 if '__x_forwarded_for_ip' in info_dict:
1897 del info_dict['__x_forwarded_for_ip']
1898
1899 # TODO Central sorting goes here
1900
1901 if formats[0] is not info_dict:
1902 # only set the 'formats' fields if the original info_dict list them
1903 # otherwise we end up with a circular reference, the first (and unique)
1904 # element in the 'formats' field in info_dict is info_dict itself,
1905 # which can't be exported to json
1906 info_dict['formats'] = formats
1907 if self.params.get('listformats'):
1908 self.list_formats(info_dict)
1909 return
1910
1911 req_format = self.params.get('format')
1912 if req_format is None:
1913 req_format = self._default_format_spec(info_dict, download=download)
1914 if self.params.get('verbose'):
1915 self.to_screen('[debug] Default format spec: %s' % req_format)
1916
1917 format_selector = self.build_format_selector(req_format)
1918
1919 # While in format selection we may need to have an access to the original
1920 # format set in order to calculate some metrics or do some processing.
1921 # For now we need to be able to guess whether original formats provided
1922 # by extractor are incomplete or not (i.e. whether extractor provides only
1923 # video-only or audio-only formats) for proper formats selection for
1924 # extractors with such incomplete formats (see
1925 # https://github.com/ytdl-org/youtube-dl/pull/5556).
1926 # Since formats may be filtered during format selection and may not match
1927 # the original formats the results may be incorrect. Thus original formats
1928 # or pre-calculated metrics should be passed to format selection routines
1929 # as well.
1930 # We will pass a context object containing all necessary additional data
1931 # instead of just formats.
1932 # This fixes incorrect format selection issue (see
1933 # https://github.com/ytdl-org/youtube-dl/issues/10083).
1934 incomplete_formats = (
1935 # All formats are video-only or
1936 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
1937 # all formats are audio-only
1938 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1939
1940 ctx = {
1941 'formats': formats,
1942 'incomplete_formats': incomplete_formats,
1943 }
1944
1945 formats_to_download = list(format_selector(ctx))
1946 if not formats_to_download:
1947 raise ExtractorError('requested format not available',
1948 expected=True)
1949
1950 if download:
1951 self.to_screen('[info] Downloading format(s) %s' % ", ".join([f['format_id'] for f in formats_to_download]))
1952 if len(formats_to_download) > 1:
1953 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1954 for format in formats_to_download:
1955 new_info = dict(info_dict)
1956 new_info.update(format)
1957 self.process_info(new_info)
1958 # We update the info dict with the best quality format (backwards compatibility)
1959 info_dict.update(formats_to_download[-1])
1960 return info_dict
1961
1962 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1963 """Select the requested subtitles and their format"""
1964 available_subs = {}
1965 if normal_subtitles and self.params.get('writesubtitles'):
1966 available_subs.update(normal_subtitles)
1967 if automatic_captions and self.params.get('writeautomaticsub'):
1968 for lang, cap_info in automatic_captions.items():
1969 if lang not in available_subs:
1970 available_subs[lang] = cap_info
1971
1972 if (not self.params.get('writesubtitles') and not
1973 self.params.get('writeautomaticsub') or not
1974 available_subs):
1975 return None
1976
1977 if self.params.get('allsubtitles', False):
1978 requested_langs = available_subs.keys()
1979 else:
1980 if self.params.get('subtitleslangs', False):
1981 requested_langs = self.params.get('subtitleslangs')
1982 elif 'en' in available_subs:
1983 requested_langs = ['en']
1984 else:
1985 requested_langs = [list(available_subs.keys())[0]]
1986
1987 formats_query = self.params.get('subtitlesformat', 'best')
1988 formats_preference = formats_query.split('/') if formats_query else []
1989 subs = {}
1990 for lang in requested_langs:
1991 formats = available_subs.get(lang)
1992 if formats is None:
1993 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1994 continue
1995 for ext in formats_preference:
1996 if ext == 'best':
1997 f = formats[-1]
1998 break
1999 matches = list(filter(lambda f: f['ext'] == ext, formats))
2000 if matches:
2001 f = matches[-1]
2002 break
2003 else:
2004 f = formats[-1]
2005 self.report_warning(
2006 'No subtitle format found matching "%s" for language %s, '
2007 'using %s' % (formats_query, lang, f['ext']))
2008 subs[lang] = f
2009 return subs
2010
2011 def __forced_printings(self, info_dict, filename, incomplete):
2012 def print_mandatory(field):
2013 if (self.params.get('force%s' % field, False)
2014 and (not incomplete or info_dict.get(field) is not None)):
2015 self.to_stdout(info_dict[field])
2016
2017 def print_optional(field):
2018 if (self.params.get('force%s' % field, False)
2019 and info_dict.get(field) is not None):
2020 self.to_stdout(info_dict[field])
2021
2022 print_mandatory('title')
2023 print_mandatory('id')
2024 if self.params.get('forceurl', False) and not incomplete:
2025 if info_dict.get('requested_formats') is not None:
2026 for f in info_dict['requested_formats']:
2027 self.to_stdout(f['url'] + f.get('play_path', ''))
2028 else:
2029 # For RTMP URLs, also include the playpath
2030 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
2031 print_optional('thumbnail')
2032 print_optional('description')
2033 if self.params.get('forcefilename', False) and filename is not None:
2034 self.to_stdout(filename)
2035 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
2036 self.to_stdout(formatSeconds(info_dict['duration']))
2037 print_mandatory('format')
2038 if self.params.get('forcejson', False):
2039 self.to_stdout(json.dumps(info_dict))
2040
2041 def process_info(self, info_dict):
2042 """Process a single resolved IE result."""
2043
2044 assert info_dict.get('_type', 'video') == 'video'
2045
2046 info_dict.setdefault('__postprocessors', [])
2047
2048 max_downloads = self.params.get('max_downloads')
2049 if max_downloads is not None:
2050 if self._num_downloads >= int(max_downloads):
2051 raise MaxDownloadsReached()
2052
2053 # TODO: backward compatibility, to be removed
2054 info_dict['fulltitle'] = info_dict['title']
2055
2056 if 'format' not in info_dict:
2057 info_dict['format'] = info_dict['ext']
2058
2059 if self._match_entry(info_dict, incomplete=False) is not None:
2060 return
2061
2062 self._num_downloads += 1
2063
2064 info_dict = self.pre_process(info_dict)
2065
2066 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2067 temp_filename = self.prepare_filename(info_dict, 'temp')
2068 files_to_move = {}
2069 skip_dl = self.params.get('skip_download', False)
2070
2071 # Forced printings
2072 self.__forced_printings(info_dict, full_filename, incomplete=False)
2073
2074 if self.params.get('simulate', False):
2075 if self.params.get('force_write_download_archive', False):
2076 self.record_download_archive(info_dict)
2077
2078 # Do nothing else if in simulate mode
2079 return
2080
2081 if full_filename is None:
2082 return
2083
2084 def ensure_dir_exists(path):
2085 return make_dir(path, self.report_error)
2086
2087 if not ensure_dir_exists(encodeFilename(full_filename)):
2088 return
2089 if not ensure_dir_exists(encodeFilename(temp_filename)):
2090 return
2091
2092 if self.params.get('writedescription', False):
2093 descfn = self.prepare_filename(info_dict, 'description')
2094 if not ensure_dir_exists(encodeFilename(descfn)):
2095 return
2096 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
2097 self.to_screen('[info] Video description is already present')
2098 elif info_dict.get('description') is None:
2099 self.report_warning('There\'s no description to write.')
2100 else:
2101 try:
2102 self.to_screen('[info] Writing video description to: ' + descfn)
2103 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
2104 descfile.write(info_dict['description'])
2105 except (OSError, IOError):
2106 self.report_error('Cannot write description file ' + descfn)
2107 return
2108
2109 if self.params.get('writeannotations', False):
2110 annofn = self.prepare_filename(info_dict, 'annotation')
2111 if not ensure_dir_exists(encodeFilename(annofn)):
2112 return
2113 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2114 self.to_screen('[info] Video annotations are already present')
2115 elif not info_dict.get('annotations'):
2116 self.report_warning('There are no annotations to write.')
2117 else:
2118 try:
2119 self.to_screen('[info] Writing video annotations to: ' + annofn)
2120 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2121 annofile.write(info_dict['annotations'])
2122 except (KeyError, TypeError):
2123 self.report_warning('There are no annotations to write.')
2124 except (OSError, IOError):
2125 self.report_error('Cannot write annotations file: ' + annofn)
2126 return
2127
2128 def dl(name, info, subtitle=False):
2129 fd = get_suitable_downloader(info, self.params)(self, self.params)
2130 for ph in self._progress_hooks:
2131 fd.add_progress_hook(ph)
2132 if self.params.get('verbose'):
2133 self.to_screen('[debug] Invoking downloader on %r' % info.get('url'))
2134 return fd.download(name, info, subtitle)
2135
2136 subtitles_are_requested = any([self.params.get('writesubtitles', False),
2137 self.params.get('writeautomaticsub')])
2138
2139 if subtitles_are_requested and info_dict.get('requested_subtitles'):
2140 # subtitles download errors are already managed as troubles in relevant IE
2141 # that way it will silently go on when used with unsupporting IE
2142 subtitles = info_dict['requested_subtitles']
2143 # ie = self.get_info_extractor(info_dict['extractor_key'])
2144 for sub_lang, sub_info in subtitles.items():
2145 sub_format = sub_info['ext']
2146 sub_fn = self.prepare_filename(info_dict, 'subtitle')
2147 sub_filename = subtitles_filename(
2148 temp_filename if not skip_dl else sub_fn,
2149 sub_lang, sub_format, info_dict.get('ext'))
2150 sub_filename_final = subtitles_filename(sub_fn, sub_lang, sub_format, info_dict.get('ext'))
2151 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(sub_filename)):
2152 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
2153 files_to_move[sub_filename] = sub_filename_final
2154 else:
2155 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
2156 if sub_info.get('data') is not None:
2157 try:
2158 # Use newline='' to prevent conversion of newline characters
2159 # See https://github.com/ytdl-org/youtube-dl/issues/10268
2160 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
2161 subfile.write(sub_info['data'])
2162 files_to_move[sub_filename] = sub_filename_final
2163 except (OSError, IOError):
2164 self.report_error('Cannot write subtitles file ' + sub_filename)
2165 return
2166 else:
2167 try:
2168 dl(sub_filename, sub_info, subtitle=True)
2169 '''
2170 if self.params.get('sleep_interval_subtitles', False):
2171 dl(sub_filename, sub_info)
2172 else:
2173 sub_data = ie._request_webpage(
2174 sub_info['url'], info_dict['id'], note=False).read()
2175 with io.open(encodeFilename(sub_filename), 'wb') as subfile:
2176 subfile.write(sub_data)
2177 '''
2178 files_to_move[sub_filename] = sub_filename_final
2179 except (ExtractorError, IOError, OSError, ValueError, compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2180 self.report_warning('Unable to download subtitle for "%s": %s' %
2181 (sub_lang, error_to_compat_str(err)))
2182 continue
2183
2184 if skip_dl:
2185 if self.params.get('convertsubtitles', False):
2186 # subconv = FFmpegSubtitlesConvertorPP(self, format=self.params.get('convertsubtitles'))
2187 filename_real_ext = os.path.splitext(full_filename)[1][1:]
2188 filename_wo_ext = (
2189 os.path.splitext(full_filename)[0]
2190 if filename_real_ext == info_dict['ext']
2191 else full_filename)
2192 afilename = '%s.%s' % (filename_wo_ext, self.params.get('convertsubtitles'))
2193 # if subconv.available:
2194 # info_dict['__postprocessors'].append(subconv)
2195 if os.path.exists(encodeFilename(afilename)):
2196 self.to_screen(
2197 '[download] %s has already been downloaded and '
2198 'converted' % afilename)
2199 else:
2200 try:
2201 self.post_process(full_filename, info_dict, files_to_move)
2202 except PostProcessingError as err:
2203 self.report_error('Postprocessing: %s' % str(err))
2204 return
2205
2206 if self.params.get('writeinfojson', False):
2207 infofn = self.prepare_filename(info_dict, 'infojson')
2208 if not ensure_dir_exists(encodeFilename(infofn)):
2209 return
2210 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
2211 self.to_screen('[info] Video metadata is already present')
2212 else:
2213 self.to_screen('[info] Writing video metadata as JSON to: ' + infofn)
2214 try:
2215 write_json_file(self.filter_requested_info(info_dict), infofn)
2216 except (OSError, IOError):
2217 self.report_error('Cannot write video metadata to JSON file ' + infofn)
2218 return
2219 info_dict['__infojson_filename'] = infofn
2220
2221 thumbfn = self.prepare_filename(info_dict, 'thumbnail')
2222 thumb_fn_temp = temp_filename if not skip_dl else thumbfn
2223 for thumb_ext in self._write_thumbnails(info_dict, thumb_fn_temp):
2224 thumb_filename_temp = replace_extension(thumb_fn_temp, thumb_ext, info_dict.get('ext'))
2225 thumb_filename = replace_extension(thumbfn, thumb_ext, info_dict.get('ext'))
2226 files_to_move[thumb_filename_temp] = info_dict['__thumbnail_filename'] = thumb_filename
2227
2228 # Write internet shortcut files
2229 url_link = webloc_link = desktop_link = False
2230 if self.params.get('writelink', False):
2231 if sys.platform == "darwin": # macOS.
2232 webloc_link = True
2233 elif sys.platform.startswith("linux"):
2234 desktop_link = True
2235 else: # if sys.platform in ['win32', 'cygwin']:
2236 url_link = True
2237 if self.params.get('writeurllink', False):
2238 url_link = True
2239 if self.params.get('writewebloclink', False):
2240 webloc_link = True
2241 if self.params.get('writedesktoplink', False):
2242 desktop_link = True
2243
2244 if url_link or webloc_link or desktop_link:
2245 if 'webpage_url' not in info_dict:
2246 self.report_error('Cannot write internet shortcut file because the "webpage_url" field is missing in the media information')
2247 return
2248 ascii_url = iri_to_uri(info_dict['webpage_url'])
2249
2250 def _write_link_file(extension, template, newline, embed_filename):
2251 linkfn = replace_extension(full_filename, extension, info_dict.get('ext'))
2252 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2253 self.to_screen('[info] Internet shortcut is already present')
2254 else:
2255 try:
2256 self.to_screen('[info] Writing internet shortcut to: ' + linkfn)
2257 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', newline=newline) as linkfile:
2258 template_vars = {'url': ascii_url}
2259 if embed_filename:
2260 template_vars['filename'] = linkfn[:-(len(extension) + 1)]
2261 linkfile.write(template % template_vars)
2262 except (OSError, IOError):
2263 self.report_error('Cannot write internet shortcut ' + linkfn)
2264 return False
2265 return True
2266
2267 if url_link:
2268 if not _write_link_file('url', DOT_URL_LINK_TEMPLATE, '\r\n', embed_filename=False):
2269 return
2270 if webloc_link:
2271 if not _write_link_file('webloc', DOT_WEBLOC_LINK_TEMPLATE, '\n', embed_filename=False):
2272 return
2273 if desktop_link:
2274 if not _write_link_file('desktop', DOT_DESKTOP_LINK_TEMPLATE, '\n', embed_filename=True):
2275 return
2276
2277 # Download
2278 must_record_download_archive = False
2279 if not skip_dl:
2280 try:
2281
2282 def existing_file(*filepaths):
2283 ext = info_dict.get('ext')
2284 final_ext = self.params.get('final_ext', ext)
2285 existing_files = []
2286 for file in orderedSet(filepaths):
2287 if final_ext != ext:
2288 converted = replace_extension(file, final_ext, ext)
2289 if os.path.exists(encodeFilename(converted)):
2290 existing_files.append(converted)
2291 if os.path.exists(encodeFilename(file)):
2292 existing_files.append(file)
2293
2294 if not existing_files or self.params.get('overwrites', False):
2295 for file in orderedSet(existing_files):
2296 self.report_file_delete(file)
2297 os.remove(encodeFilename(file))
2298 return None
2299
2300 self.report_file_already_downloaded(existing_files[0])
2301 info_dict['ext'] = os.path.splitext(existing_files[0])[1][1:]
2302 return existing_files[0]
2303
2304 success = True
2305 if info_dict.get('requested_formats') is not None:
2306 downloaded = []
2307 merger = FFmpegMergerPP(self)
2308 if self.params.get('allow_unplayable_formats'):
2309 self.report_warning(
2310 'You have requested merging of multiple formats '
2311 'while also allowing unplayable formats to be downloaded. '
2312 'The formats won\'t be merged to prevent data corruption.')
2313 elif not merger.available:
2314 self.report_warning(
2315 'You have requested merging of multiple formats but ffmpeg is not installed. '
2316 'The formats won\'t be merged.')
2317
2318 def compatible_formats(formats):
2319 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2320 video_formats = [format for format in formats if format.get('vcodec') != 'none']
2321 audio_formats = [format for format in formats if format.get('acodec') != 'none']
2322 if len(video_formats) > 2 or len(audio_formats) > 2:
2323 return False
2324
2325 # Check extension
2326 exts = set(format.get('ext') for format in formats)
2327 COMPATIBLE_EXTS = (
2328 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
2329 set(('webm',)),
2330 )
2331 for ext_sets in COMPATIBLE_EXTS:
2332 if ext_sets.issuperset(exts):
2333 return True
2334 # TODO: Check acodec/vcodec
2335 return False
2336
2337 requested_formats = info_dict['requested_formats']
2338 old_ext = info_dict['ext']
2339 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
2340 info_dict['ext'] = 'mkv'
2341 self.report_warning(
2342 'Requested formats are incompatible for merge and will be merged into mkv.')
2343
2344 def correct_ext(filename):
2345 filename_real_ext = os.path.splitext(filename)[1][1:]
2346 filename_wo_ext = (
2347 os.path.splitext(filename)[0]
2348 if filename_real_ext == old_ext
2349 else filename)
2350 return '%s.%s' % (filename_wo_ext, info_dict['ext'])
2351
2352 # Ensure filename always has a correct extension for successful merge
2353 full_filename = correct_ext(full_filename)
2354 temp_filename = correct_ext(temp_filename)
2355 dl_filename = existing_file(full_filename, temp_filename)
2356 info_dict['__real_download'] = False
2357 if dl_filename is None:
2358 for f in requested_formats:
2359 new_info = dict(info_dict)
2360 new_info.update(f)
2361 fname = prepend_extension(
2362 self.prepare_filename(new_info, 'temp'),
2363 'f%s' % f['format_id'], new_info['ext'])
2364 if not ensure_dir_exists(fname):
2365 return
2366 downloaded.append(fname)
2367 partial_success, real_download = dl(fname, new_info)
2368 info_dict['__real_download'] = info_dict['__real_download'] or real_download
2369 success = success and partial_success
2370 if merger.available and not self.params.get('allow_unplayable_formats'):
2371 info_dict['__postprocessors'].append(merger)
2372 info_dict['__files_to_merge'] = downloaded
2373 # Even if there were no downloads, it is being merged only now
2374 info_dict['__real_download'] = True
2375 else:
2376 for file in downloaded:
2377 files_to_move[file] = None
2378 else:
2379 # Just a single file
2380 dl_filename = existing_file(full_filename, temp_filename)
2381 if dl_filename is None:
2382 success, real_download = dl(temp_filename, info_dict)
2383 info_dict['__real_download'] = real_download
2384
2385 dl_filename = dl_filename or temp_filename
2386 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2387
2388 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2389 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
2390 return
2391 except (OSError, IOError) as err:
2392 raise UnavailableVideoError(err)
2393 except (ContentTooShortError, ) as err:
2394 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
2395 return
2396
2397 if success and full_filename != '-':
2398 # Fixup content
2399 fixup_policy = self.params.get('fixup')
2400 if fixup_policy is None:
2401 fixup_policy = 'detect_or_warn'
2402
2403 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg to fix this automatically.'
2404
2405 stretched_ratio = info_dict.get('stretched_ratio')
2406 if stretched_ratio is not None and stretched_ratio != 1:
2407 if fixup_policy == 'warn':
2408 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
2409 info_dict['id'], stretched_ratio))
2410 elif fixup_policy == 'detect_or_warn':
2411 stretched_pp = FFmpegFixupStretchedPP(self)
2412 if stretched_pp.available:
2413 info_dict['__postprocessors'].append(stretched_pp)
2414 else:
2415 self.report_warning(
2416 '%s: Non-uniform pixel ratio (%s). %s'
2417 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
2418 else:
2419 assert fixup_policy in ('ignore', 'never')
2420
2421 if (info_dict.get('requested_formats') is None
2422 and info_dict.get('container') == 'm4a_dash'
2423 and info_dict.get('ext') == 'm4a'):
2424 if fixup_policy == 'warn':
2425 self.report_warning(
2426 '%s: writing DASH m4a. '
2427 'Only some players support this container.'
2428 % info_dict['id'])
2429 elif fixup_policy == 'detect_or_warn':
2430 fixup_pp = FFmpegFixupM4aPP(self)
2431 if fixup_pp.available:
2432 info_dict['__postprocessors'].append(fixup_pp)
2433 else:
2434 self.report_warning(
2435 '%s: writing DASH m4a. '
2436 'Only some players support this container. %s'
2437 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
2438 else:
2439 assert fixup_policy in ('ignore', 'never')
2440
2441 if (info_dict.get('protocol') == 'm3u8_native'
2442 or info_dict.get('protocol') == 'm3u8'
2443 and self.params.get('hls_prefer_native')):
2444 if fixup_policy == 'warn':
2445 self.report_warning('%s: malformed AAC bitstream detected.' % (
2446 info_dict['id']))
2447 elif fixup_policy == 'detect_or_warn':
2448 fixup_pp = FFmpegFixupM3u8PP(self)
2449 if fixup_pp.available:
2450 info_dict['__postprocessors'].append(fixup_pp)
2451 else:
2452 self.report_warning(
2453 '%s: malformed AAC bitstream detected. %s'
2454 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
2455 else:
2456 assert fixup_policy in ('ignore', 'never')
2457
2458 try:
2459 self.post_process(dl_filename, info_dict, files_to_move)
2460 except PostProcessingError as err:
2461 self.report_error('Postprocessing: %s' % str(err))
2462 return
2463 try:
2464 for ph in self._post_hooks:
2465 ph(full_filename)
2466 except Exception as err:
2467 self.report_error('post hooks: %s' % str(err))
2468 return
2469 must_record_download_archive = True
2470
2471 if must_record_download_archive or self.params.get('force_write_download_archive', False):
2472 self.record_download_archive(info_dict)
2473 max_downloads = self.params.get('max_downloads')
2474 if max_downloads is not None and self._num_downloads >= int(max_downloads):
2475 raise MaxDownloadsReached()
2476
2477 def download(self, url_list):
2478 """Download a given list of URLs."""
2479 outtmpl = self.outtmpl_dict['default']
2480 if (len(url_list) > 1
2481 and outtmpl != '-'
2482 and '%' not in outtmpl
2483 and self.params.get('max_downloads') != 1):
2484 raise SameFileError(outtmpl)
2485
2486 for url in url_list:
2487 try:
2488 # It also downloads the videos
2489 res = self.extract_info(
2490 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2491 except UnavailableVideoError:
2492 self.report_error('unable to download video')
2493 except MaxDownloadsReached:
2494 self.to_screen('[info] Maximum number of downloaded files reached')
2495 raise
2496 except ExistingVideoReached:
2497 self.to_screen('[info] Encountered a file that is already in the archive, stopping due to --break-on-existing')
2498 raise
2499 except RejectedVideoReached:
2500 self.to_screen('[info] Encountered a file that did not match filter, stopping due to --break-on-reject')
2501 raise
2502 else:
2503 if self.params.get('dump_single_json', False):
2504 self.to_stdout(json.dumps(res))
2505
2506 return self._download_retcode
2507
2508 def download_with_info_file(self, info_filename):
2509 with contextlib.closing(fileinput.FileInput(
2510 [info_filename], mode='r',
2511 openhook=fileinput.hook_encoded('utf-8'))) as f:
2512 # FileInput doesn't have a read method, we can't call json.load
2513 info = self.filter_requested_info(json.loads('\n'.join(f)))
2514 try:
2515 self.process_ie_result(info, download=True)
2516 except DownloadError:
2517 webpage_url = info.get('webpage_url')
2518 if webpage_url is not None:
2519 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2520 return self.download([webpage_url])
2521 else:
2522 raise
2523 return self._download_retcode
2524
2525 @staticmethod
2526 def filter_requested_info(info_dict):
2527 fields_to_remove = ('requested_formats', 'requested_subtitles')
2528 return dict(
2529 (k, v) for k, v in info_dict.items()
2530 if (k[0] != '_' or k == '_type') and k not in fields_to_remove)
2531
2532 def run_pp(self, pp, infodict, files_to_move={}):
2533 files_to_delete = []
2534 files_to_delete, infodict = pp.run(infodict)
2535 if not files_to_delete:
2536 return files_to_move, infodict
2537
2538 if self.params.get('keepvideo', False):
2539 for f in files_to_delete:
2540 files_to_move.setdefault(f, '')
2541 else:
2542 for old_filename in set(files_to_delete):
2543 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2544 try:
2545 os.remove(encodeFilename(old_filename))
2546 except (IOError, OSError):
2547 self.report_warning('Unable to remove downloaded original file')
2548 if old_filename in files_to_move:
2549 del files_to_move[old_filename]
2550 return files_to_move, infodict
2551
2552 def pre_process(self, ie_info):
2553 info = dict(ie_info)
2554 for pp in self._pps['beforedl']:
2555 info = self.run_pp(pp, info)[1]
2556 return info
2557
2558 def post_process(self, filename, ie_info, files_to_move={}):
2559 """Run all the postprocessors on the given file."""
2560 info = dict(ie_info)
2561 info['filepath'] = filename
2562 info['__files_to_move'] = {}
2563
2564 for pp in ie_info.get('__postprocessors', []) + self._pps['normal']:
2565 files_to_move, info = self.run_pp(pp, info, files_to_move)
2566 info = self.run_pp(MoveFilesAfterDownloadPP(self, files_to_move), info)[1]
2567 for pp in self._pps['aftermove']:
2568 info = self.run_pp(pp, info, {})[1]
2569
2570 def _make_archive_id(self, info_dict):
2571 video_id = info_dict.get('id')
2572 if not video_id:
2573 return
2574 # Future-proof against any change in case
2575 # and backwards compatibility with prior versions
2576 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
2577 if extractor is None:
2578 url = str_or_none(info_dict.get('url'))
2579 if not url:
2580 return
2581 # Try to find matching extractor for the URL and take its ie_key
2582 for ie in self._ies:
2583 if ie.suitable(url):
2584 extractor = ie.ie_key()
2585 break
2586 else:
2587 return
2588 return '%s %s' % (extractor.lower(), video_id)
2589
2590 def in_download_archive(self, info_dict):
2591 fn = self.params.get('download_archive')
2592 if fn is None:
2593 return False
2594
2595 vid_id = self._make_archive_id(info_dict)
2596 if not vid_id:
2597 return False # Incomplete video information
2598
2599 return vid_id in self.archive
2600
2601 def record_download_archive(self, info_dict):
2602 fn = self.params.get('download_archive')
2603 if fn is None:
2604 return
2605 vid_id = self._make_archive_id(info_dict)
2606 assert vid_id
2607 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2608 archive_file.write(vid_id + '\n')
2609 self.archive.add(vid_id)
2610
2611 @staticmethod
2612 def format_resolution(format, default='unknown'):
2613 if format.get('vcodec') == 'none':
2614 return 'audio only'
2615 if format.get('resolution') is not None:
2616 return format['resolution']
2617 if format.get('height') is not None:
2618 if format.get('width') is not None:
2619 res = '%sx%s' % (format['width'], format['height'])
2620 else:
2621 res = '%sp' % format['height']
2622 elif format.get('width') is not None:
2623 res = '%dx?' % format['width']
2624 else:
2625 res = default
2626 return res
2627
2628 def _format_note(self, fdict):
2629 res = ''
2630 if fdict.get('ext') in ['f4f', 'f4m']:
2631 res += '(unsupported) '
2632 if fdict.get('language'):
2633 if res:
2634 res += ' '
2635 res += '[%s] ' % fdict['language']
2636 if fdict.get('format_note') is not None:
2637 res += fdict['format_note'] + ' '
2638 if fdict.get('tbr') is not None:
2639 res += '%4dk ' % fdict['tbr']
2640 if fdict.get('container') is not None:
2641 if res:
2642 res += ', '
2643 res += '%s container' % fdict['container']
2644 if (fdict.get('vcodec') is not None
2645 and fdict.get('vcodec') != 'none'):
2646 if res:
2647 res += ', '
2648 res += fdict['vcodec']
2649 if fdict.get('vbr') is not None:
2650 res += '@'
2651 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2652 res += 'video@'
2653 if fdict.get('vbr') is not None:
2654 res += '%4dk' % fdict['vbr']
2655 if fdict.get('fps') is not None:
2656 if res:
2657 res += ', '
2658 res += '%sfps' % fdict['fps']
2659 if fdict.get('acodec') is not None:
2660 if res:
2661 res += ', '
2662 if fdict['acodec'] == 'none':
2663 res += 'video only'
2664 else:
2665 res += '%-5s' % fdict['acodec']
2666 elif fdict.get('abr') is not None:
2667 if res:
2668 res += ', '
2669 res += 'audio'
2670 if fdict.get('abr') is not None:
2671 res += '@%3dk' % fdict['abr']
2672 if fdict.get('asr') is not None:
2673 res += ' (%5dHz)' % fdict['asr']
2674 if fdict.get('filesize') is not None:
2675 if res:
2676 res += ', '
2677 res += format_bytes(fdict['filesize'])
2678 elif fdict.get('filesize_approx') is not None:
2679 if res:
2680 res += ', '
2681 res += '~' + format_bytes(fdict['filesize_approx'])
2682 return res
2683
2684 def _format_note_table(self, f):
2685 def join_fields(*vargs):
2686 return ', '.join((val for val in vargs if val != ''))
2687
2688 return join_fields(
2689 'UNSUPPORTED' if f.get('ext') in ('f4f', 'f4m') else '',
2690 format_field(f, 'language', '[%s]'),
2691 format_field(f, 'format_note'),
2692 format_field(f, 'container', ignore=(None, f.get('ext'))),
2693 format_field(f, 'asr', '%5dHz'))
2694
2695 def list_formats(self, info_dict):
2696 formats = info_dict.get('formats', [info_dict])
2697 new_format = self.params.get('listformats_table', False)
2698 if new_format:
2699 table = [
2700 [
2701 format_field(f, 'format_id'),
2702 format_field(f, 'ext'),
2703 self.format_resolution(f),
2704 format_field(f, 'fps', '%d'),
2705 '|',
2706 format_field(f, 'filesize', ' %s', func=format_bytes) + format_field(f, 'filesize_approx', '~%s', func=format_bytes),
2707 format_field(f, 'tbr', '%4dk'),
2708 f.get('protocol').replace('http_dash_segments', 'dash').replace("native", "n").replace('niconico_', ''),
2709 '|',
2710 format_field(f, 'vcodec', default='unknown').replace('none', ''),
2711 format_field(f, 'vbr', '%4dk'),
2712 format_field(f, 'acodec', default='unknown').replace('none', ''),
2713 format_field(f, 'abr', '%3dk'),
2714 format_field(f, 'asr', '%5dHz'),
2715 self._format_note_table(f)]
2716 for f in formats
2717 if f.get('preference') is None or f['preference'] >= -1000]
2718 header_line = ['ID', 'EXT', 'RESOLUTION', 'FPS', '|', ' FILESIZE', ' TBR', 'PROTO',
2719 '|', 'VCODEC', ' VBR', 'ACODEC', ' ABR', ' ASR', 'NOTE']
2720 else:
2721 table = [
2722 [
2723 format_field(f, 'format_id'),
2724 format_field(f, 'ext'),
2725 self.format_resolution(f),
2726 self._format_note(f)]
2727 for f in formats
2728 if f.get('preference') is None or f['preference'] >= -1000]
2729 header_line = ['format code', 'extension', 'resolution', 'note']
2730
2731 self.to_screen(
2732 '[info] Available formats for %s:\n%s' % (info_dict['id'], render_table(
2733 header_line,
2734 table,
2735 delim=new_format,
2736 extraGap=(0 if new_format else 1),
2737 hideEmpty=new_format)))
2738
2739 def list_thumbnails(self, info_dict):
2740 thumbnails = info_dict.get('thumbnails')
2741 if not thumbnails:
2742 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
2743 return
2744
2745 self.to_screen(
2746 '[info] Thumbnails for %s:' % info_dict['id'])
2747 self.to_screen(render_table(
2748 ['ID', 'width', 'height', 'URL'],
2749 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
2750
2751 def list_subtitles(self, video_id, subtitles, name='subtitles'):
2752 if not subtitles:
2753 self.to_screen('%s has no %s' % (video_id, name))
2754 return
2755 self.to_screen(
2756 'Available %s for %s:' % (name, video_id))
2757 self.to_screen(render_table(
2758 ['Language', 'formats'],
2759 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
2760 for lang, formats in subtitles.items()]))
2761
2762 def urlopen(self, req):
2763 """ Start an HTTP download """
2764 if isinstance(req, compat_basestring):
2765 req = sanitized_Request(req)
2766 return self._opener.open(req, timeout=self._socket_timeout)
2767
2768 def print_debug_header(self):
2769 if not self.params.get('verbose'):
2770 return
2771
2772 if type('') is not compat_str:
2773 # Python 2.6 on SLES11 SP1 (https://github.com/ytdl-org/youtube-dl/issues/3326)
2774 self.report_warning(
2775 'Your Python is broken! Update to a newer and supported version')
2776
2777 stdout_encoding = getattr(
2778 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2779 encoding_str = (
2780 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2781 locale.getpreferredencoding(),
2782 sys.getfilesystemencoding(),
2783 stdout_encoding,
2784 self.get_encoding()))
2785 write_string(encoding_str, encoding=None)
2786
2787 source = (
2788 '(exe)' if hasattr(sys, 'frozen')
2789 else '(zip)' if isinstance(globals().get('__loader__'), zipimporter)
2790 else '(source)' if os.path.basename(sys.argv[0]) == '__main__.py'
2791 else '')
2792 self._write_string('[debug] yt-dlp version %s %s\n' % (__version__, source))
2793 if _LAZY_LOADER:
2794 self._write_string('[debug] Lazy loading extractors enabled\n')
2795 if _PLUGIN_CLASSES:
2796 self._write_string(
2797 '[debug] Plugin Extractors: %s\n' % [ie.ie_key() for ie in _PLUGIN_CLASSES])
2798 try:
2799 sp = subprocess.Popen(
2800 ['git', 'rev-parse', '--short', 'HEAD'],
2801 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2802 cwd=os.path.dirname(os.path.abspath(__file__)))
2803 out, err = process_communicate_or_kill(sp)
2804 out = out.decode().strip()
2805 if re.match('[0-9a-f]+', out):
2806 self._write_string('[debug] Git HEAD: %s\n' % out)
2807 except Exception:
2808 try:
2809 sys.exc_clear()
2810 except Exception:
2811 pass
2812
2813 def python_implementation():
2814 impl_name = platform.python_implementation()
2815 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
2816 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
2817 return impl_name
2818
2819 self._write_string('[debug] Python version %s (%s %s) - %s\n' % (
2820 platform.python_version(),
2821 python_implementation(),
2822 platform.architecture()[0],
2823 platform_name()))
2824
2825 exe_versions = FFmpegPostProcessor.get_versions(self)
2826 exe_versions['rtmpdump'] = rtmpdump_version()
2827 exe_versions['phantomjs'] = PhantomJSwrapper._version()
2828 exe_str = ', '.join(
2829 '%s %s' % (exe, v)
2830 for exe, v in sorted(exe_versions.items())
2831 if v
2832 )
2833 if not exe_str:
2834 exe_str = 'none'
2835 self._write_string('[debug] exe versions: %s\n' % exe_str)
2836
2837 proxy_map = {}
2838 for handler in self._opener.handlers:
2839 if hasattr(handler, 'proxies'):
2840 proxy_map.update(handler.proxies)
2841 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2842
2843 if self.params.get('call_home', False):
2844 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2845 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2846 return
2847 latest_version = self.urlopen(
2848 'https://yt-dl.org/latest/version').read().decode('utf-8')
2849 if version_tuple(latest_version) > version_tuple(__version__):
2850 self.report_warning(
2851 'You are using an outdated version (newest version: %s)! '
2852 'See https://yt-dl.org/update if you need help updating.' %
2853 latest_version)
2854
2855 def _setup_opener(self):
2856 timeout_val = self.params.get('socket_timeout')
2857 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2858
2859 opts_cookiefile = self.params.get('cookiefile')
2860 opts_proxy = self.params.get('proxy')
2861
2862 if opts_cookiefile is None:
2863 self.cookiejar = compat_cookiejar.CookieJar()
2864 else:
2865 opts_cookiefile = expand_path(opts_cookiefile)
2866 self.cookiejar = YoutubeDLCookieJar(opts_cookiefile)
2867 if os.access(opts_cookiefile, os.R_OK):
2868 self.cookiejar.load(ignore_discard=True, ignore_expires=True)
2869
2870 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2871 if opts_proxy is not None:
2872 if opts_proxy == '':
2873 proxies = {}
2874 else:
2875 proxies = {'http': opts_proxy, 'https': opts_proxy}
2876 else:
2877 proxies = compat_urllib_request.getproxies()
2878 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
2879 if 'http' in proxies and 'https' not in proxies:
2880 proxies['https'] = proxies['http']
2881 proxy_handler = PerRequestProxyHandler(proxies)
2882
2883 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2884 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2885 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2886 redirect_handler = YoutubeDLRedirectHandler()
2887 data_handler = compat_urllib_request_DataHandler()
2888
2889 # When passing our own FileHandler instance, build_opener won't add the
2890 # default FileHandler and allows us to disable the file protocol, which
2891 # can be used for malicious purposes (see
2892 # https://github.com/ytdl-org/youtube-dl/issues/8227)
2893 file_handler = compat_urllib_request.FileHandler()
2894
2895 def file_open(*args, **kwargs):
2896 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
2897 file_handler.file_open = file_open
2898
2899 opener = compat_urllib_request.build_opener(
2900 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
2901
2902 # Delete the default user-agent header, which would otherwise apply in
2903 # cases where our custom HTTP handler doesn't come into play
2904 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
2905 opener.addheaders = []
2906 self._opener = opener
2907
2908 def encode(self, s):
2909 if isinstance(s, bytes):
2910 return s # Already encoded
2911
2912 try:
2913 return s.encode(self.get_encoding())
2914 except UnicodeEncodeError as err:
2915 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2916 raise
2917
2918 def get_encoding(self):
2919 encoding = self.params.get('encoding')
2920 if encoding is None:
2921 encoding = preferredencoding()
2922 return encoding
2923
2924 def _write_thumbnails(self, info_dict, filename): # return the extensions
2925 write_all = self.params.get('write_all_thumbnails', False)
2926 thumbnails = []
2927 if write_all or self.params.get('writethumbnail', False):
2928 thumbnails = info_dict.get('thumbnails') or []
2929 multiple = write_all and len(thumbnails) > 1
2930
2931 ret = []
2932 for t in thumbnails[::1 if write_all else -1]:
2933 thumb_ext = determine_ext(t['url'], 'jpg')
2934 suffix = '%s.' % t['id'] if multiple else ''
2935 thumb_display_id = '%s ' % t['id'] if multiple else ''
2936 t['filename'] = thumb_filename = replace_extension(filename, suffix + thumb_ext, info_dict.get('ext'))
2937
2938 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(thumb_filename)):
2939 ret.append(suffix + thumb_ext)
2940 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2941 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2942 else:
2943 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2944 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2945 try:
2946 uf = self.urlopen(t['url'])
2947 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2948 shutil.copyfileobj(uf, thumbf)
2949 ret.append(suffix + thumb_ext)
2950 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2951 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2952 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2953 self.report_warning('Unable to download thumbnail "%s": %s' %
2954 (t['url'], error_to_compat_str(err)))
2955 if ret and not write_all:
2956 break
2957 return ret