]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
Extract comments only when needed #95 (Closes #94)
[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_requests: Number of seconds to sleep between requests
328 during extraction
329 sleep_interval: Number of seconds to sleep before each download when
330 used alone or a lower bound of a range for randomized
331 sleep before each download (minimum possible number
332 of seconds to sleep) when used along with
333 max_sleep_interval.
334 max_sleep_interval:Upper bound of a range for randomized sleep before each
335 download (maximum possible number of seconds to sleep).
336 Must only be used along with sleep_interval.
337 Actual sleep time will be a random float from range
338 [sleep_interval; max_sleep_interval].
339 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
340 listformats: Print an overview of available video formats and exit.
341 list_thumbnails: Print a table of all thumbnails and exit.
342 match_filter: A function that gets called with the info_dict of
343 every video.
344 If it returns a message, the video is ignored.
345 If it returns None, the video is downloaded.
346 match_filter_func in utils.py is one example for this.
347 no_color: Do not emit color codes in output.
348 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
349 HTTP header
350 geo_bypass_country:
351 Two-letter ISO 3166-2 country code that will be used for
352 explicit geographic restriction bypassing via faking
353 X-Forwarded-For HTTP header
354 geo_bypass_ip_block:
355 IP range in CIDR notation that will be used similarly to
356 geo_bypass_country
357
358 The following options determine which downloader is picked:
359 external_downloader: Executable of the external downloader to call.
360 None or unset for standard (built-in) downloader.
361 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
362 if True, otherwise use ffmpeg/avconv if False, otherwise
363 use downloader suggested by extractor if None.
364
365 The following parameters are not used by YoutubeDL itself, they are used by
366 the downloader (see yt_dlp/downloader/common.py):
367 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
368 noresizebuffer, retries, continuedl, noprogress, consoletitle,
369 xattr_set_filesize, external_downloader_args, hls_use_mpegts,
370 http_chunk_size.
371
372 The following options are used by the post processors:
373 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
374 otherwise prefer ffmpeg. (avconv support is deprecated)
375 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
376 to the binary or its containing directory.
377 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
378 and a list of additional command-line arguments for the
379 postprocessor/executable. The dict can also have "PP+EXE" keys
380 which are used when the given exe is used by the given PP.
381 Use 'default' as the name for arguments to passed to all PP
382
383 The following options are used by the extractors:
384 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
385 hls_split_discontinuity: Split HLS playlists to different formats at
386 discontinuities such as ad breaks (default: False)
387 youtube_include_dash_manifest: If True (default), DASH manifests and related
388 data will be downloaded and processed by extractor.
389 You can reduce network I/O by disabling it if you don't
390 care about DASH. (only for youtube)
391 youtube_include_hls_manifest: If True (default), HLS manifests and related
392 data will be downloaded and processed by extractor.
393 You can reduce network I/O by disabling it if you don't
394 care about HLS. (only for youtube)
395 """
396
397 _NUMERIC_FIELDS = set((
398 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
399 'timestamp', 'upload_year', 'upload_month', 'upload_day',
400 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
401 'average_rating', 'comment_count', 'age_limit',
402 'start_time', 'end_time',
403 'chapter_number', 'season_number', 'episode_number',
404 'track_number', 'disc_number', 'release_year',
405 'playlist_index',
406 ))
407
408 params = None
409 _ies = []
410 _pps = {'beforedl': [], 'aftermove': [], 'normal': []}
411 __prepare_filename_warned = False
412 _first_webpage_request = True
413 _download_retcode = None
414 _num_downloads = None
415 _playlist_level = 0
416 _playlist_urls = set()
417 _screen_file = None
418
419 def __init__(self, params=None, auto_init=True):
420 """Create a FileDownloader object with the given options."""
421 if params is None:
422 params = {}
423 self._ies = []
424 self._ies_instances = {}
425 self._pps = {'beforedl': [], 'aftermove': [], 'normal': []}
426 self.__prepare_filename_warned = False
427 self._first_webpage_request = True
428 self._post_hooks = []
429 self._progress_hooks = []
430 self._download_retcode = 0
431 self._num_downloads = 0
432 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
433 self._err_file = sys.stderr
434 self.params = {
435 # Default parameters
436 'nocheckcertificate': False,
437 }
438 self.params.update(params)
439 self.cache = Cache(self)
440 self.archive = set()
441
442 """Preload the archive, if any is specified"""
443 def preload_download_archive(self):
444 fn = self.params.get('download_archive')
445 if fn is None:
446 return False
447 try:
448 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
449 for line in archive_file:
450 self.archive.add(line.strip())
451 except IOError as ioe:
452 if ioe.errno != errno.ENOENT:
453 raise
454 return False
455 return True
456
457 def check_deprecated(param, option, suggestion):
458 if self.params.get(param) is not None:
459 self.report_warning(
460 '%s is deprecated. Use %s instead.' % (option, suggestion))
461 return True
462 return False
463
464 if self.params.get('verbose'):
465 self.to_stdout('[debug] Loading archive file %r' % self.params.get('download_archive'))
466
467 preload_download_archive(self)
468
469 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
470 if self.params.get('geo_verification_proxy') is None:
471 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
472
473 if self.params.get('final_ext'):
474 if self.params.get('merge_output_format'):
475 self.report_warning('--merge-output-format will be ignored since --remux-video or --recode-video is given')
476 self.params['merge_output_format'] = self.params['final_ext']
477
478 if 'overwrites' in self.params and self.params['overwrites'] is None:
479 del self.params['overwrites']
480
481 check_deprecated('autonumber_size', '--autonumber-size', 'output template with %(autonumber)0Nd, where N in the number of digits')
482 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
483 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
484
485 if params.get('bidi_workaround', False):
486 try:
487 import pty
488 master, slave = pty.openpty()
489 width = compat_get_terminal_size().columns
490 if width is None:
491 width_args = []
492 else:
493 width_args = ['-w', str(width)]
494 sp_kwargs = dict(
495 stdin=subprocess.PIPE,
496 stdout=slave,
497 stderr=self._err_file)
498 try:
499 self._output_process = subprocess.Popen(
500 ['bidiv'] + width_args, **sp_kwargs
501 )
502 except OSError:
503 self._output_process = subprocess.Popen(
504 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
505 self._output_channel = os.fdopen(master, 'rb')
506 except OSError as ose:
507 if ose.errno == errno.ENOENT:
508 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 $PATH.')
509 else:
510 raise
511
512 if (sys.platform != 'win32'
513 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
514 and not params.get('restrictfilenames', False)):
515 # Unicode filesystem API will throw errors (#1474, #13027)
516 self.report_warning(
517 'Assuming --restrict-filenames since file system encoding '
518 'cannot encode all characters. '
519 'Set the LC_ALL environment variable to fix this.')
520 self.params['restrictfilenames'] = True
521
522 self.outtmpl_dict = self.parse_outtmpl()
523
524 self._setup_opener()
525
526 if auto_init:
527 self.print_debug_header()
528 self.add_default_info_extractors()
529
530 for pp_def_raw in self.params.get('postprocessors', []):
531 pp_class = get_postprocessor(pp_def_raw['key'])
532 pp_def = dict(pp_def_raw)
533 del pp_def['key']
534 if 'when' in pp_def:
535 when = pp_def['when']
536 del pp_def['when']
537 else:
538 when = 'normal'
539 pp = pp_class(self, **compat_kwargs(pp_def))
540 self.add_post_processor(pp, when=when)
541
542 for ph in self.params.get('post_hooks', []):
543 self.add_post_hook(ph)
544
545 for ph in self.params.get('progress_hooks', []):
546 self.add_progress_hook(ph)
547
548 register_socks_protocols()
549
550 def warn_if_short_id(self, argv):
551 # short YouTube ID starting with dash?
552 idxs = [
553 i for i, a in enumerate(argv)
554 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
555 if idxs:
556 correct_argv = (
557 ['yt-dlp']
558 + [a for i, a in enumerate(argv) if i not in idxs]
559 + ['--'] + [argv[i] for i in idxs]
560 )
561 self.report_warning(
562 'Long argument string detected. '
563 'Use -- to separate parameters and URLs, like this:\n%s\n' %
564 args_to_str(correct_argv))
565
566 def add_info_extractor(self, ie):
567 """Add an InfoExtractor object to the end of the list."""
568 self._ies.append(ie)
569 if not isinstance(ie, type):
570 self._ies_instances[ie.ie_key()] = ie
571 ie.set_downloader(self)
572
573 def get_info_extractor(self, ie_key):
574 """
575 Get an instance of an IE with name ie_key, it will try to get one from
576 the _ies list, if there's no instance it will create a new one and add
577 it to the extractor list.
578 """
579 ie = self._ies_instances.get(ie_key)
580 if ie is None:
581 ie = get_info_extractor(ie_key)()
582 self.add_info_extractor(ie)
583 return ie
584
585 def add_default_info_extractors(self):
586 """
587 Add the InfoExtractors returned by gen_extractors to the end of the list
588 """
589 for ie in gen_extractor_classes():
590 self.add_info_extractor(ie)
591
592 def add_post_processor(self, pp, when='normal'):
593 """Add a PostProcessor object to the end of the chain."""
594 self._pps[when].append(pp)
595 pp.set_downloader(self)
596
597 def add_post_hook(self, ph):
598 """Add the post hook"""
599 self._post_hooks.append(ph)
600
601 def add_progress_hook(self, ph):
602 """Add the progress hook (currently only for the file downloader)"""
603 self._progress_hooks.append(ph)
604
605 def _bidi_workaround(self, message):
606 if not hasattr(self, '_output_channel'):
607 return message
608
609 assert hasattr(self, '_output_process')
610 assert isinstance(message, compat_str)
611 line_count = message.count('\n') + 1
612 self._output_process.stdin.write((message + '\n').encode('utf-8'))
613 self._output_process.stdin.flush()
614 res = ''.join(self._output_channel.readline().decode('utf-8')
615 for _ in range(line_count))
616 return res[:-len('\n')]
617
618 def to_screen(self, message, skip_eol=False):
619 """Print message to stdout if not in quiet mode."""
620 return self.to_stdout(message, skip_eol, check_quiet=True)
621
622 def _write_string(self, s, out=None):
623 write_string(s, out=out, encoding=self.params.get('encoding'))
624
625 def to_stdout(self, message, skip_eol=False, check_quiet=False):
626 """Print message to stdout if not in quiet mode."""
627 if self.params.get('logger'):
628 self.params['logger'].debug(message)
629 elif not check_quiet or not self.params.get('quiet', False):
630 message = self._bidi_workaround(message)
631 terminator = ['\n', ''][skip_eol]
632 output = message + terminator
633
634 self._write_string(output, self._screen_file)
635
636 def to_stderr(self, message):
637 """Print message to stderr."""
638 assert isinstance(message, compat_str)
639 if self.params.get('logger'):
640 self.params['logger'].error(message)
641 else:
642 message = self._bidi_workaround(message)
643 output = message + '\n'
644 self._write_string(output, self._err_file)
645
646 def to_console_title(self, message):
647 if not self.params.get('consoletitle', False):
648 return
649 if compat_os_name == 'nt':
650 if ctypes.windll.kernel32.GetConsoleWindow():
651 # c_wchar_p() might not be necessary if `message` is
652 # already of type unicode()
653 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
654 elif 'TERM' in os.environ:
655 self._write_string('\033]0;%s\007' % message, self._screen_file)
656
657 def save_console_title(self):
658 if not self.params.get('consoletitle', False):
659 return
660 if self.params.get('simulate', False):
661 return
662 if compat_os_name != 'nt' and 'TERM' in os.environ:
663 # Save the title on stack
664 self._write_string('\033[22;0t', self._screen_file)
665
666 def restore_console_title(self):
667 if not self.params.get('consoletitle', False):
668 return
669 if self.params.get('simulate', False):
670 return
671 if compat_os_name != 'nt' and 'TERM' in os.environ:
672 # Restore the title from stack
673 self._write_string('\033[23;0t', self._screen_file)
674
675 def __enter__(self):
676 self.save_console_title()
677 return self
678
679 def __exit__(self, *args):
680 self.restore_console_title()
681
682 if self.params.get('cookiefile') is not None:
683 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
684
685 def trouble(self, message=None, tb=None):
686 """Determine action to take when a download problem appears.
687
688 Depending on if the downloader has been configured to ignore
689 download errors or not, this method may throw an exception or
690 not when errors are found, after printing the message.
691
692 tb, if given, is additional traceback information.
693 """
694 if message is not None:
695 self.to_stderr(message)
696 if self.params.get('verbose'):
697 if tb is None:
698 if sys.exc_info()[0]: # if .trouble has been called from an except block
699 tb = ''
700 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
701 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
702 tb += encode_compat_str(traceback.format_exc())
703 else:
704 tb_data = traceback.format_list(traceback.extract_stack())
705 tb = ''.join(tb_data)
706 self.to_stderr(tb)
707 if not self.params.get('ignoreerrors', False):
708 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
709 exc_info = sys.exc_info()[1].exc_info
710 else:
711 exc_info = sys.exc_info()
712 raise DownloadError(message, exc_info)
713 self._download_retcode = 1
714
715 def report_warning(self, message):
716 '''
717 Print the message to stderr, it will be prefixed with 'WARNING:'
718 If stderr is a tty file the 'WARNING:' will be colored
719 '''
720 if self.params.get('logger') is not None:
721 self.params['logger'].warning(message)
722 else:
723 if self.params.get('no_warnings'):
724 return
725 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
726 _msg_header = '\033[0;33mWARNING:\033[0m'
727 else:
728 _msg_header = 'WARNING:'
729 warning_message = '%s %s' % (_msg_header, message)
730 self.to_stderr(warning_message)
731
732 def report_error(self, message, tb=None):
733 '''
734 Do the same as trouble, but prefixes the message with 'ERROR:', colored
735 in red if stderr is a tty file.
736 '''
737 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
738 _msg_header = '\033[0;31mERROR:\033[0m'
739 else:
740 _msg_header = 'ERROR:'
741 error_message = '%s %s' % (_msg_header, message)
742 self.trouble(error_message, tb)
743
744 def report_file_already_downloaded(self, file_name):
745 """Report file has already been fully downloaded."""
746 try:
747 self.to_screen('[download] %s has already been downloaded' % file_name)
748 except UnicodeEncodeError:
749 self.to_screen('[download] The file has already been downloaded')
750
751 def report_file_delete(self, file_name):
752 """Report that existing file will be deleted."""
753 try:
754 self.to_screen('Deleting existing file %s' % file_name)
755 except UnicodeEncodeError:
756 self.to_screen('Deleting existing file')
757
758 def parse_outtmpl(self):
759 outtmpl_dict = self.params.get('outtmpl', {})
760 if not isinstance(outtmpl_dict, dict):
761 outtmpl_dict = {'default': outtmpl_dict}
762 outtmpl_dict.update({
763 k: v for k, v in DEFAULT_OUTTMPL.items()
764 if not outtmpl_dict.get(k)})
765 for key, val in outtmpl_dict.items():
766 if isinstance(val, bytes):
767 self.report_warning(
768 'Parameter outtmpl is bytes, but should be a unicode string. '
769 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
770 return outtmpl_dict
771
772 def _prepare_filename(self, info_dict, tmpl_type='default'):
773 try:
774 template_dict = dict(info_dict)
775
776 template_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
777 formatSeconds(info_dict['duration'], '-')
778 if info_dict.get('duration', None) is not None
779 else None)
780
781 template_dict['epoch'] = int(time.time())
782 autonumber_size = self.params.get('autonumber_size')
783 if autonumber_size is None:
784 autonumber_size = 5
785 template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
786 if template_dict.get('resolution') is None:
787 if template_dict.get('width') and template_dict.get('height'):
788 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
789 elif template_dict.get('height'):
790 template_dict['resolution'] = '%sp' % template_dict['height']
791 elif template_dict.get('width'):
792 template_dict['resolution'] = '%dx?' % template_dict['width']
793
794 sanitize = lambda k, v: sanitize_filename(
795 compat_str(v),
796 restricted=self.params.get('restrictfilenames'),
797 is_id=(k == 'id' or k.endswith('_id')))
798 template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
799 for k, v in template_dict.items()
800 if v is not None and not isinstance(v, (list, tuple, dict)))
801 na = self.params.get('outtmpl_na_placeholder', 'NA')
802 template_dict = collections.defaultdict(lambda: na, template_dict)
803
804 outtmpl = self.outtmpl_dict.get(tmpl_type, self.outtmpl_dict['default'])
805 force_ext = OUTTMPL_TYPES.get(tmpl_type)
806
807 # For fields playlist_index and autonumber convert all occurrences
808 # of %(field)s to %(field)0Nd for backward compatibility
809 field_size_compat_map = {
810 'playlist_index': len(str(template_dict['n_entries'])),
811 'autonumber': autonumber_size,
812 }
813 FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
814 mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
815 if mobj:
816 outtmpl = re.sub(
817 FIELD_SIZE_COMPAT_RE,
818 r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
819 outtmpl)
820
821 # As of [1] format syntax is:
822 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
823 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
824 FORMAT_RE = r'''(?x)
825 (?<!%)
826 %
827 \({0}\) # mapping key
828 (?:[#0\-+ ]+)? # conversion flags (optional)
829 (?:\d+)? # minimum field width (optional)
830 (?:\.\d+)? # precision (optional)
831 [hlL]? # length modifier (optional)
832 (?P<type>[diouxXeEfFgGcrs%]) # conversion type
833 '''
834
835 numeric_fields = list(self._NUMERIC_FIELDS)
836
837 # Format date
838 FORMAT_DATE_RE = FORMAT_RE.format(r'(?P<key>(?P<field>\w+)>(?P<format>.+?))')
839 for mobj in re.finditer(FORMAT_DATE_RE, outtmpl):
840 conv_type, field, frmt, key = mobj.group('type', 'field', 'format', 'key')
841 if key in template_dict:
842 continue
843 value = strftime_or_none(template_dict.get(field), frmt, na)
844 if conv_type in 'crs': # string
845 value = sanitize(field, value)
846 else: # number
847 numeric_fields.append(key)
848 value = float_or_none(value, default=None)
849 if value is not None:
850 template_dict[key] = value
851
852 # Missing numeric fields used together with integer presentation types
853 # in format specification will break the argument substitution since
854 # string NA placeholder is returned for missing fields. We will patch
855 # output template for missing fields to meet string presentation type.
856 for numeric_field in numeric_fields:
857 if numeric_field not in template_dict:
858 outtmpl = re.sub(
859 FORMAT_RE.format(re.escape(numeric_field)),
860 r'%({0})s'.format(numeric_field), outtmpl)
861
862 # expand_path translates '%%' into '%' and '$$' into '$'
863 # correspondingly that is not what we want since we need to keep
864 # '%%' intact for template dict substitution step. Working around
865 # with boundary-alike separator hack.
866 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
867 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
868
869 # outtmpl should be expand_path'ed before template dict substitution
870 # because meta fields may contain env variables we don't want to
871 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
872 # title "Hello $PATH", we don't want `$PATH` to be expanded.
873 filename = expand_path(outtmpl).replace(sep, '') % template_dict
874
875 if force_ext is not None:
876 filename = replace_extension(filename, force_ext, template_dict.get('ext'))
877
878 # https://github.com/blackjack4494/youtube-dlc/issues/85
879 trim_file_name = self.params.get('trim_file_name', False)
880 if trim_file_name:
881 fn_groups = filename.rsplit('.')
882 ext = fn_groups[-1]
883 sub_ext = ''
884 if len(fn_groups) > 2:
885 sub_ext = fn_groups[-2]
886 filename = '.'.join(filter(None, [fn_groups[0][:trim_file_name], sub_ext, ext]))
887
888 return filename
889 except ValueError as err:
890 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
891 return None
892
893 def prepare_filename(self, info_dict, dir_type='', warn=False):
894 """Generate the output filename."""
895 paths = self.params.get('paths', {})
896 assert isinstance(paths, dict)
897 filename = self._prepare_filename(info_dict, dir_type or 'default')
898
899 if warn and not self.__prepare_filename_warned:
900 if not paths:
901 pass
902 elif filename == '-':
903 self.report_warning('--paths is ignored when an outputting to stdout')
904 elif os.path.isabs(filename):
905 self.report_warning('--paths is ignored since an absolute path is given in output template')
906 self.__prepare_filename_warned = True
907 if filename == '-' or not filename:
908 return filename
909
910 homepath = expand_path(paths.get('home', '').strip())
911 assert isinstance(homepath, compat_str)
912 subdir = expand_path(paths.get(dir_type, '').strip()) if dir_type else ''
913 assert isinstance(subdir, compat_str)
914 path = os.path.join(homepath, subdir, filename)
915
916 # Temporary fix for #4787
917 # 'Treat' all problem characters by passing filename through preferredencoding
918 # to workaround encoding issues with subprocess on python2 @ Windows
919 if sys.version_info < (3, 0) and sys.platform == 'win32':
920 path = encodeFilename(path, True).decode(preferredencoding())
921 return sanitize_path(path, force=self.params.get('windowsfilenames'))
922
923 def _match_entry(self, info_dict, incomplete):
924 """ Returns None if the file should be downloaded """
925
926 def check_filter():
927 video_title = info_dict.get('title', info_dict.get('id', 'video'))
928 if 'title' in info_dict:
929 # This can happen when we're just evaluating the playlist
930 title = info_dict['title']
931 matchtitle = self.params.get('matchtitle', False)
932 if matchtitle:
933 if not re.search(matchtitle, title, re.IGNORECASE):
934 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
935 rejecttitle = self.params.get('rejecttitle', False)
936 if rejecttitle:
937 if re.search(rejecttitle, title, re.IGNORECASE):
938 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
939 date = info_dict.get('upload_date')
940 if date is not None:
941 dateRange = self.params.get('daterange', DateRange())
942 if date not in dateRange:
943 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
944 view_count = info_dict.get('view_count')
945 if view_count is not None:
946 min_views = self.params.get('min_views')
947 if min_views is not None and view_count < min_views:
948 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
949 max_views = self.params.get('max_views')
950 if max_views is not None and view_count > max_views:
951 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
952 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
953 return 'Skipping "%s" because it is age restricted' % video_title
954 if self.in_download_archive(info_dict):
955 return '%s has already been recorded in archive' % video_title
956
957 if not incomplete:
958 match_filter = self.params.get('match_filter')
959 if match_filter is not None:
960 ret = match_filter(info_dict)
961 if ret is not None:
962 return ret
963 return None
964
965 reason = check_filter()
966 if reason is not None:
967 self.to_screen('[download] ' + reason)
968 if reason.endswith('has already been recorded in the archive') and self.params.get('break_on_existing', False):
969 raise ExistingVideoReached()
970 elif self.params.get('break_on_reject', False):
971 raise RejectedVideoReached()
972 return reason
973
974 @staticmethod
975 def add_extra_info(info_dict, extra_info):
976 '''Set the keys from extra_info in info dict if they are missing'''
977 for key, value in extra_info.items():
978 info_dict.setdefault(key, value)
979
980 def extract_info(self, url, download=True, ie_key=None, info_dict=None, extra_info={},
981 process=True, force_generic_extractor=False):
982 '''
983 Returns a list with a dictionary for each video we find.
984 If 'download', also downloads the videos.
985 extra_info is a dict containing the extra values to add to each result
986 '''
987
988 if not ie_key and force_generic_extractor:
989 ie_key = 'Generic'
990
991 if ie_key:
992 ies = [self.get_info_extractor(ie_key)]
993 else:
994 ies = self._ies
995
996 for ie in ies:
997 if not ie.suitable(url):
998 continue
999
1000 ie_key = ie.ie_key()
1001 ie = self.get_info_extractor(ie_key)
1002 if not ie.working():
1003 self.report_warning('The program functionality for this site has been marked as broken, '
1004 'and will probably not work.')
1005
1006 try:
1007 temp_id = str_or_none(
1008 ie.extract_id(url) if callable(getattr(ie, 'extract_id', None))
1009 else ie._match_id(url))
1010 except (AssertionError, IndexError, AttributeError):
1011 temp_id = None
1012 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1013 self.to_screen("[%s] %s: has already been recorded in archive" % (
1014 ie_key, temp_id))
1015 break
1016 return self.__extract_info(url, ie, download, extra_info, process, info_dict)
1017 else:
1018 self.report_error('no suitable InfoExtractor for URL %s' % url)
1019
1020 def __handle_extraction_exceptions(func):
1021 def wrapper(self, *args, **kwargs):
1022 try:
1023 return func(self, *args, **kwargs)
1024 except GeoRestrictedError as e:
1025 msg = e.msg
1026 if e.countries:
1027 msg += '\nThis video is available in %s.' % ', '.join(
1028 map(ISO3166Utils.short2full, e.countries))
1029 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1030 self.report_error(msg)
1031 except ExtractorError as e: # An error we somewhat expected
1032 self.report_error(compat_str(e), e.format_traceback())
1033 except (MaxDownloadsReached, ExistingVideoReached, RejectedVideoReached):
1034 raise
1035 except Exception as e:
1036 if self.params.get('ignoreerrors', False):
1037 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
1038 else:
1039 raise
1040 return wrapper
1041
1042 @__handle_extraction_exceptions
1043 def __extract_info(self, url, ie, download, extra_info, process, info_dict):
1044 ie_result = ie.extract(url)
1045 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1046 return
1047 if isinstance(ie_result, list):
1048 # Backwards compatibility: old IE result format
1049 ie_result = {
1050 '_type': 'compat_list',
1051 'entries': ie_result,
1052 }
1053 if info_dict:
1054 if info_dict.get('id'):
1055 ie_result['id'] = info_dict['id']
1056 if info_dict.get('title'):
1057 ie_result['title'] = info_dict['title']
1058 self.add_default_extra_info(ie_result, ie, url)
1059 if process:
1060 return self.process_ie_result(ie_result, download, extra_info)
1061 else:
1062 return ie_result
1063
1064 def add_default_extra_info(self, ie_result, ie, url):
1065 self.add_extra_info(ie_result, {
1066 'extractor': ie.IE_NAME,
1067 'webpage_url': url,
1068 'webpage_url_basename': url_basename(url),
1069 'extractor_key': ie.ie_key(),
1070 })
1071
1072 def process_ie_result(self, ie_result, download=True, extra_info={}):
1073 """
1074 Take the result of the ie(may be modified) and resolve all unresolved
1075 references (URLs, playlist items).
1076
1077 It will also download the videos if 'download'.
1078 Returns the resolved ie_result.
1079 """
1080 result_type = ie_result.get('_type', 'video')
1081
1082 if result_type in ('url', 'url_transparent'):
1083 ie_result['url'] = sanitize_url(ie_result['url'])
1084 extract_flat = self.params.get('extract_flat', False)
1085 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1086 or extract_flat is True):
1087 self.__forced_printings(ie_result, self.prepare_filename(ie_result), incomplete=True)
1088 return ie_result
1089
1090 if result_type == 'video':
1091 self.add_extra_info(ie_result, extra_info)
1092 return self.process_video_result(ie_result, download=download)
1093 elif result_type == 'url':
1094 # We have to add extra_info to the results because it may be
1095 # contained in a playlist
1096 return self.extract_info(ie_result['url'],
1097 download, info_dict=ie_result,
1098 ie_key=ie_result.get('ie_key'),
1099 extra_info=extra_info)
1100 elif result_type == 'url_transparent':
1101 # Use the information from the embedding page
1102 info = self.extract_info(
1103 ie_result['url'], ie_key=ie_result.get('ie_key'),
1104 extra_info=extra_info, download=False, process=False)
1105
1106 # extract_info may return None when ignoreerrors is enabled and
1107 # extraction failed with an error, don't crash and return early
1108 # in this case
1109 if not info:
1110 return info
1111
1112 force_properties = dict(
1113 (k, v) for k, v in ie_result.items() if v is not None)
1114 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1115 if f in force_properties:
1116 del force_properties[f]
1117 new_result = info.copy()
1118 new_result.update(force_properties)
1119
1120 # Extracted info may not be a video result (i.e.
1121 # info.get('_type', 'video') != video) but rather an url or
1122 # url_transparent. In such cases outer metadata (from ie_result)
1123 # should be propagated to inner one (info). For this to happen
1124 # _type of info should be overridden with url_transparent. This
1125 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1126 if new_result.get('_type') == 'url':
1127 new_result['_type'] = 'url_transparent'
1128
1129 return self.process_ie_result(
1130 new_result, download=download, extra_info=extra_info)
1131 elif result_type in ('playlist', 'multi_video'):
1132 # Protect from infinite recursion due to recursively nested playlists
1133 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1134 webpage_url = ie_result['webpage_url']
1135 if webpage_url in self._playlist_urls:
1136 self.to_screen(
1137 '[download] Skipping already downloaded playlist: %s'
1138 % ie_result.get('title') or ie_result.get('id'))
1139 return
1140
1141 self._playlist_level += 1
1142 self._playlist_urls.add(webpage_url)
1143 try:
1144 return self.__process_playlist(ie_result, download)
1145 finally:
1146 self._playlist_level -= 1
1147 if not self._playlist_level:
1148 self._playlist_urls.clear()
1149 elif result_type == 'compat_list':
1150 self.report_warning(
1151 'Extractor %s returned a compat_list result. '
1152 'It needs to be updated.' % ie_result.get('extractor'))
1153
1154 def _fixup(r):
1155 self.add_extra_info(
1156 r,
1157 {
1158 'extractor': ie_result['extractor'],
1159 'webpage_url': ie_result['webpage_url'],
1160 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1161 'extractor_key': ie_result['extractor_key'],
1162 }
1163 )
1164 return r
1165 ie_result['entries'] = [
1166 self.process_ie_result(_fixup(r), download, extra_info)
1167 for r in ie_result['entries']
1168 ]
1169 return ie_result
1170 else:
1171 raise Exception('Invalid result type: %s' % result_type)
1172
1173 def __process_playlist(self, ie_result, download):
1174 # We process each entry in the playlist
1175 playlist = ie_result.get('title') or ie_result.get('id')
1176 self.to_screen('[download] Downloading playlist: %s' % playlist)
1177
1178 if self.params.get('allow_playlist_files', True):
1179 ie_copy = {
1180 'playlist': playlist,
1181 'playlist_id': ie_result.get('id'),
1182 'playlist_title': ie_result.get('title'),
1183 'playlist_uploader': ie_result.get('uploader'),
1184 'playlist_uploader_id': ie_result.get('uploader_id'),
1185 'playlist_index': 0
1186 }
1187 ie_copy.update(dict(ie_result))
1188
1189 def ensure_dir_exists(path):
1190 return make_dir(path, self.report_error)
1191
1192 if self.params.get('writeinfojson', False):
1193 infofn = self.prepare_filename(ie_copy, 'pl_infojson')
1194 if not ensure_dir_exists(encodeFilename(infofn)):
1195 return
1196 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
1197 self.to_screen('[info] Playlist metadata is already present')
1198 else:
1199 playlist_info = dict(ie_result)
1200 # playlist_info['entries'] = list(playlist_info['entries']) # Entries is a generator which shouldnot be resolved here
1201 del playlist_info['entries']
1202 self.to_screen('[info] Writing playlist metadata as JSON to: ' + infofn)
1203 try:
1204 write_json_file(self.filter_requested_info(playlist_info), infofn)
1205 except (OSError, IOError):
1206 self.report_error('Cannot write playlist metadata to JSON file ' + infofn)
1207
1208 if self.params.get('writedescription', False):
1209 descfn = self.prepare_filename(ie_copy, 'pl_description')
1210 if not ensure_dir_exists(encodeFilename(descfn)):
1211 return
1212 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
1213 self.to_screen('[info] Playlist description is already present')
1214 elif ie_result.get('description') is None:
1215 self.report_warning('There\'s no playlist description to write.')
1216 else:
1217 try:
1218 self.to_screen('[info] Writing playlist description to: ' + descfn)
1219 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1220 descfile.write(ie_result['description'])
1221 except (OSError, IOError):
1222 self.report_error('Cannot write playlist description file ' + descfn)
1223 return
1224
1225 playlist_results = []
1226
1227 playliststart = self.params.get('playliststart', 1) - 1
1228 playlistend = self.params.get('playlistend')
1229 # For backwards compatibility, interpret -1 as whole list
1230 if playlistend == -1:
1231 playlistend = None
1232
1233 playlistitems_str = self.params.get('playlist_items')
1234 playlistitems = None
1235 if playlistitems_str is not None:
1236 def iter_playlistitems(format):
1237 for string_segment in format.split(','):
1238 if '-' in string_segment:
1239 start, end = string_segment.split('-')
1240 for item in range(int(start), int(end) + 1):
1241 yield int(item)
1242 else:
1243 yield int(string_segment)
1244 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1245
1246 ie_entries = ie_result['entries']
1247
1248 def make_playlistitems_entries(list_ie_entries):
1249 num_entries = len(list_ie_entries)
1250 return [
1251 list_ie_entries[i - 1] for i in playlistitems
1252 if -num_entries <= i - 1 < num_entries]
1253
1254 def report_download(num_entries):
1255 self.to_screen(
1256 '[%s] playlist %s: Downloading %d videos' %
1257 (ie_result['extractor'], playlist, num_entries))
1258
1259 if isinstance(ie_entries, list):
1260 n_all_entries = len(ie_entries)
1261 if playlistitems:
1262 entries = make_playlistitems_entries(ie_entries)
1263 else:
1264 entries = ie_entries[playliststart:playlistend]
1265 n_entries = len(entries)
1266 self.to_screen(
1267 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
1268 (ie_result['extractor'], playlist, n_all_entries, n_entries))
1269 elif isinstance(ie_entries, PagedList):
1270 if playlistitems:
1271 entries = []
1272 for item in playlistitems:
1273 entries.extend(ie_entries.getslice(
1274 item - 1, item
1275 ))
1276 else:
1277 entries = ie_entries.getslice(
1278 playliststart, playlistend)
1279 n_entries = len(entries)
1280 report_download(n_entries)
1281 else: # iterable
1282 if playlistitems:
1283 entries = make_playlistitems_entries(list(itertools.islice(
1284 ie_entries, 0, max(playlistitems))))
1285 else:
1286 entries = list(itertools.islice(
1287 ie_entries, playliststart, playlistend))
1288 n_entries = len(entries)
1289 report_download(n_entries)
1290
1291 if self.params.get('playlistreverse', False):
1292 entries = entries[::-1]
1293
1294 if self.params.get('playlistrandom', False):
1295 random.shuffle(entries)
1296
1297 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1298
1299 for i, entry in enumerate(entries, 1):
1300 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1301 # This __x_forwarded_for_ip thing is a bit ugly but requires
1302 # minimal changes
1303 if x_forwarded_for:
1304 entry['__x_forwarded_for_ip'] = x_forwarded_for
1305 extra = {
1306 'n_entries': n_entries,
1307 'playlist': playlist,
1308 'playlist_id': ie_result.get('id'),
1309 'playlist_title': ie_result.get('title'),
1310 'playlist_uploader': ie_result.get('uploader'),
1311 'playlist_uploader_id': ie_result.get('uploader_id'),
1312 'playlist_index': playlistitems[i - 1] if playlistitems else i + playliststart,
1313 'extractor': ie_result['extractor'],
1314 'webpage_url': ie_result['webpage_url'],
1315 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1316 'extractor_key': ie_result['extractor_key'],
1317 }
1318
1319 if self._match_entry(entry, incomplete=True) is not None:
1320 continue
1321
1322 entry_result = self.__process_iterable_entry(entry, download, extra)
1323 # TODO: skip failed (empty) entries?
1324 playlist_results.append(entry_result)
1325 ie_result['entries'] = playlist_results
1326 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1327 return ie_result
1328
1329 @__handle_extraction_exceptions
1330 def __process_iterable_entry(self, entry, download, extra_info):
1331 return self.process_ie_result(
1332 entry, download=download, extra_info=extra_info)
1333
1334 def _build_format_filter(self, filter_spec):
1335 " Returns a function to filter the formats according to the filter_spec "
1336
1337 OPERATORS = {
1338 '<': operator.lt,
1339 '<=': operator.le,
1340 '>': operator.gt,
1341 '>=': operator.ge,
1342 '=': operator.eq,
1343 '!=': operator.ne,
1344 }
1345 operator_rex = re.compile(r'''(?x)\s*
1346 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)
1347 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1348 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
1349 $
1350 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1351 m = operator_rex.search(filter_spec)
1352 if m:
1353 try:
1354 comparison_value = int(m.group('value'))
1355 except ValueError:
1356 comparison_value = parse_filesize(m.group('value'))
1357 if comparison_value is None:
1358 comparison_value = parse_filesize(m.group('value') + 'B')
1359 if comparison_value is None:
1360 raise ValueError(
1361 'Invalid value %r in format specification %r' % (
1362 m.group('value'), filter_spec))
1363 op = OPERATORS[m.group('op')]
1364
1365 if not m:
1366 STR_OPERATORS = {
1367 '=': operator.eq,
1368 '^=': lambda attr, value: attr.startswith(value),
1369 '$=': lambda attr, value: attr.endswith(value),
1370 '*=': lambda attr, value: value in attr,
1371 }
1372 str_operator_rex = re.compile(r'''(?x)
1373 \s*(?P<key>[a-zA-Z0-9._-]+)
1374 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?
1375 \s*(?P<value>[a-zA-Z0-9._-]+)
1376 \s*$
1377 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1378 m = str_operator_rex.search(filter_spec)
1379 if m:
1380 comparison_value = m.group('value')
1381 str_op = STR_OPERATORS[m.group('op')]
1382 if m.group('negation'):
1383 op = lambda attr, value: not str_op(attr, value)
1384 else:
1385 op = str_op
1386
1387 if not m:
1388 raise ValueError('Invalid filter specification %r' % filter_spec)
1389
1390 def _filter(f):
1391 actual_value = f.get(m.group('key'))
1392 if actual_value is None:
1393 return m.group('none_inclusive')
1394 return op(actual_value, comparison_value)
1395 return _filter
1396
1397 def _default_format_spec(self, info_dict, download=True):
1398
1399 def can_merge():
1400 merger = FFmpegMergerPP(self)
1401 return merger.available and merger.can_merge()
1402
1403 prefer_best = (
1404 not self.params.get('simulate', False)
1405 and download
1406 and (
1407 not can_merge()
1408 or info_dict.get('is_live', False)
1409 or self.outtmpl_dict['default'] == '-'))
1410
1411 return (
1412 'best/bestvideo+bestaudio'
1413 if prefer_best
1414 else 'bestvideo*+bestaudio/best'
1415 if not self.params.get('allow_multiple_audio_streams', False)
1416 else 'bestvideo+bestaudio/best')
1417
1418 def build_format_selector(self, format_spec):
1419 def syntax_error(note, start):
1420 message = (
1421 'Invalid format specification: '
1422 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1423 return SyntaxError(message)
1424
1425 PICKFIRST = 'PICKFIRST'
1426 MERGE = 'MERGE'
1427 SINGLE = 'SINGLE'
1428 GROUP = 'GROUP'
1429 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1430
1431 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1432 'video': self.params.get('allow_multiple_video_streams', False)}
1433
1434 def _parse_filter(tokens):
1435 filter_parts = []
1436 for type, string, start, _, _ in tokens:
1437 if type == tokenize.OP and string == ']':
1438 return ''.join(filter_parts)
1439 else:
1440 filter_parts.append(string)
1441
1442 def _remove_unused_ops(tokens):
1443 # Remove operators that we don't use and join them with the surrounding strings
1444 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1445 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1446 last_string, last_start, last_end, last_line = None, None, None, None
1447 for type, string, start, end, line in tokens:
1448 if type == tokenize.OP and string == '[':
1449 if last_string:
1450 yield tokenize.NAME, last_string, last_start, last_end, last_line
1451 last_string = None
1452 yield type, string, start, end, line
1453 # everything inside brackets will be handled by _parse_filter
1454 for type, string, start, end, line in tokens:
1455 yield type, string, start, end, line
1456 if type == tokenize.OP and string == ']':
1457 break
1458 elif type == tokenize.OP and string in ALLOWED_OPS:
1459 if last_string:
1460 yield tokenize.NAME, last_string, last_start, last_end, last_line
1461 last_string = None
1462 yield type, string, start, end, line
1463 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1464 if not last_string:
1465 last_string = string
1466 last_start = start
1467 last_end = end
1468 else:
1469 last_string += string
1470 if last_string:
1471 yield tokenize.NAME, last_string, last_start, last_end, last_line
1472
1473 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1474 selectors = []
1475 current_selector = None
1476 for type, string, start, _, _ in tokens:
1477 # ENCODING is only defined in python 3.x
1478 if type == getattr(tokenize, 'ENCODING', None):
1479 continue
1480 elif type in [tokenize.NAME, tokenize.NUMBER]:
1481 current_selector = FormatSelector(SINGLE, string, [])
1482 elif type == tokenize.OP:
1483 if string == ')':
1484 if not inside_group:
1485 # ')' will be handled by the parentheses group
1486 tokens.restore_last_token()
1487 break
1488 elif inside_merge and string in ['/', ',']:
1489 tokens.restore_last_token()
1490 break
1491 elif inside_choice and string == ',':
1492 tokens.restore_last_token()
1493 break
1494 elif string == ',':
1495 if not current_selector:
1496 raise syntax_error('"," must follow a format selector', start)
1497 selectors.append(current_selector)
1498 current_selector = None
1499 elif string == '/':
1500 if not current_selector:
1501 raise syntax_error('"/" must follow a format selector', start)
1502 first_choice = current_selector
1503 second_choice = _parse_format_selection(tokens, inside_choice=True)
1504 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1505 elif string == '[':
1506 if not current_selector:
1507 current_selector = FormatSelector(SINGLE, 'best', [])
1508 format_filter = _parse_filter(tokens)
1509 current_selector.filters.append(format_filter)
1510 elif string == '(':
1511 if current_selector:
1512 raise syntax_error('Unexpected "("', start)
1513 group = _parse_format_selection(tokens, inside_group=True)
1514 current_selector = FormatSelector(GROUP, group, [])
1515 elif string == '+':
1516 if not current_selector:
1517 raise syntax_error('Unexpected "+"', start)
1518 selector_1 = current_selector
1519 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1520 if not selector_2:
1521 raise syntax_error('Expected a selector', start)
1522 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1523 else:
1524 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1525 elif type == tokenize.ENDMARKER:
1526 break
1527 if current_selector:
1528 selectors.append(current_selector)
1529 return selectors
1530
1531 def _build_selector_function(selector):
1532 if isinstance(selector, list): # ,
1533 fs = [_build_selector_function(s) for s in selector]
1534
1535 def selector_function(ctx):
1536 for f in fs:
1537 for format in f(ctx):
1538 yield format
1539 return selector_function
1540
1541 elif selector.type == GROUP: # ()
1542 selector_function = _build_selector_function(selector.selector)
1543
1544 elif selector.type == PICKFIRST: # /
1545 fs = [_build_selector_function(s) for s in selector.selector]
1546
1547 def selector_function(ctx):
1548 for f in fs:
1549 picked_formats = list(f(ctx))
1550 if picked_formats:
1551 return picked_formats
1552 return []
1553
1554 elif selector.type == SINGLE: # atom
1555 format_spec = selector.selector if selector.selector is not None else 'best'
1556
1557 if format_spec == 'all':
1558 def selector_function(ctx):
1559 formats = list(ctx['formats'])
1560 if formats:
1561 for f in formats:
1562 yield f
1563
1564 else:
1565 format_fallback = False
1566 format_spec_obj = re.match(r'(best|worst|b|w)(video|audio|v|a)?(\*)?$', format_spec)
1567 if format_spec_obj is not None:
1568 format_idx = 0 if format_spec_obj.group(1)[0] == 'w' else -1
1569 format_type = format_spec_obj.group(2)[0] if format_spec_obj.group(2) else False
1570 not_format_type = 'v' if format_type == 'a' else 'a'
1571 format_modified = format_spec_obj.group(3) is not None
1572
1573 format_fallback = not format_type and not format_modified # for b, w
1574 filter_f = ((lambda f: f.get(format_type + 'codec') != 'none')
1575 if format_type and format_modified # bv*, ba*, wv*, wa*
1576 else (lambda f: f.get(not_format_type + 'codec') == 'none')
1577 if format_type # bv, ba, wv, wa
1578 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
1579 if not format_modified # b, w
1580 else None) # b*, w*
1581 else:
1582 format_idx = -1
1583 filter_f = ((lambda f: f.get('ext') == format_spec)
1584 if format_spec in ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] # extension
1585 else (lambda f: f.get('format_id') == format_spec)) # id
1586
1587 def selector_function(ctx):
1588 formats = list(ctx['formats'])
1589 if not formats:
1590 return
1591 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
1592 if matches:
1593 yield matches[format_idx]
1594 elif format_fallback == 'force' or (format_fallback and ctx['incomplete_formats']):
1595 # for extractors with incomplete formats (audio only (soundcloud)
1596 # or video only (imgur)) best/worst will fallback to
1597 # best/worst {video,audio}-only format
1598 yield formats[format_idx]
1599
1600 elif selector.type == MERGE: # +
1601 def _merge(formats_pair):
1602 format_1, format_2 = formats_pair
1603
1604 formats_info = []
1605 formats_info.extend(format_1.get('requested_formats', (format_1,)))
1606 formats_info.extend(format_2.get('requested_formats', (format_2,)))
1607
1608 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
1609 get_no_more = {"video": False, "audio": False}
1610 for (i, fmt_info) in enumerate(formats_info):
1611 for aud_vid in ["audio", "video"]:
1612 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
1613 if get_no_more[aud_vid]:
1614 formats_info.pop(i)
1615 get_no_more[aud_vid] = True
1616
1617 if len(formats_info) == 1:
1618 return formats_info[0]
1619
1620 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
1621 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
1622
1623 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
1624 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
1625
1626 output_ext = self.params.get('merge_output_format')
1627 if not output_ext:
1628 if the_only_video:
1629 output_ext = the_only_video['ext']
1630 elif the_only_audio and not video_fmts:
1631 output_ext = the_only_audio['ext']
1632 else:
1633 output_ext = 'mkv'
1634
1635 new_dict = {
1636 'requested_formats': formats_info,
1637 'format': '+'.join(fmt_info.get('format') for fmt_info in formats_info),
1638 'format_id': '+'.join(fmt_info.get('format_id') for fmt_info in formats_info),
1639 'ext': output_ext,
1640 }
1641
1642 if the_only_video:
1643 new_dict.update({
1644 'width': the_only_video.get('width'),
1645 'height': the_only_video.get('height'),
1646 'resolution': the_only_video.get('resolution'),
1647 'fps': the_only_video.get('fps'),
1648 'vcodec': the_only_video.get('vcodec'),
1649 'vbr': the_only_video.get('vbr'),
1650 'stretched_ratio': the_only_video.get('stretched_ratio'),
1651 })
1652
1653 if the_only_audio:
1654 new_dict.update({
1655 'acodec': the_only_audio.get('acodec'),
1656 'abr': the_only_audio.get('abr'),
1657 })
1658
1659 return new_dict
1660
1661 selector_1, selector_2 = map(_build_selector_function, selector.selector)
1662
1663 def selector_function(ctx):
1664 for pair in itertools.product(
1665 selector_1(copy.deepcopy(ctx)), selector_2(copy.deepcopy(ctx))):
1666 yield _merge(pair)
1667
1668 filters = [self._build_format_filter(f) for f in selector.filters]
1669
1670 def final_selector(ctx):
1671 ctx_copy = copy.deepcopy(ctx)
1672 for _filter in filters:
1673 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1674 return selector_function(ctx_copy)
1675 return final_selector
1676
1677 stream = io.BytesIO(format_spec.encode('utf-8'))
1678 try:
1679 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1680 except tokenize.TokenError:
1681 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1682
1683 class TokenIterator(object):
1684 def __init__(self, tokens):
1685 self.tokens = tokens
1686 self.counter = 0
1687
1688 def __iter__(self):
1689 return self
1690
1691 def __next__(self):
1692 if self.counter >= len(self.tokens):
1693 raise StopIteration()
1694 value = self.tokens[self.counter]
1695 self.counter += 1
1696 return value
1697
1698 next = __next__
1699
1700 def restore_last_token(self):
1701 self.counter -= 1
1702
1703 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1704 return _build_selector_function(parsed_selector)
1705
1706 def _calc_headers(self, info_dict):
1707 res = std_headers.copy()
1708
1709 add_headers = info_dict.get('http_headers')
1710 if add_headers:
1711 res.update(add_headers)
1712
1713 cookies = self._calc_cookies(info_dict)
1714 if cookies:
1715 res['Cookie'] = cookies
1716
1717 if 'X-Forwarded-For' not in res:
1718 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1719 if x_forwarded_for_ip:
1720 res['X-Forwarded-For'] = x_forwarded_for_ip
1721
1722 return res
1723
1724 def _calc_cookies(self, info_dict):
1725 pr = sanitized_Request(info_dict['url'])
1726 self.cookiejar.add_cookie_header(pr)
1727 return pr.get_header('Cookie')
1728
1729 def process_video_result(self, info_dict, download=True):
1730 assert info_dict.get('_type', 'video') == 'video'
1731
1732 if 'id' not in info_dict:
1733 raise ExtractorError('Missing "id" field in extractor result')
1734 if 'title' not in info_dict:
1735 raise ExtractorError('Missing "title" field in extractor result')
1736
1737 def report_force_conversion(field, field_not, conversion):
1738 self.report_warning(
1739 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
1740 % (field, field_not, conversion))
1741
1742 def sanitize_string_field(info, string_field):
1743 field = info.get(string_field)
1744 if field is None or isinstance(field, compat_str):
1745 return
1746 report_force_conversion(string_field, 'a string', 'string')
1747 info[string_field] = compat_str(field)
1748
1749 def sanitize_numeric_fields(info):
1750 for numeric_field in self._NUMERIC_FIELDS:
1751 field = info.get(numeric_field)
1752 if field is None or isinstance(field, compat_numeric_types):
1753 continue
1754 report_force_conversion(numeric_field, 'numeric', 'int')
1755 info[numeric_field] = int_or_none(field)
1756
1757 sanitize_string_field(info_dict, 'id')
1758 sanitize_numeric_fields(info_dict)
1759
1760 if 'playlist' not in info_dict:
1761 # It isn't part of a playlist
1762 info_dict['playlist'] = None
1763 info_dict['playlist_index'] = None
1764
1765 thumbnails = info_dict.get('thumbnails')
1766 if thumbnails is None:
1767 thumbnail = info_dict.get('thumbnail')
1768 if thumbnail:
1769 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1770 if thumbnails:
1771 thumbnails.sort(key=lambda t: (
1772 t.get('preference') if t.get('preference') is not None else -1,
1773 t.get('width') if t.get('width') is not None else -1,
1774 t.get('height') if t.get('height') is not None else -1,
1775 t.get('id') if t.get('id') is not None else '', t.get('url')))
1776 for i, t in enumerate(thumbnails):
1777 t['url'] = sanitize_url(t['url'])
1778 if t.get('width') and t.get('height'):
1779 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1780 if t.get('id') is None:
1781 t['id'] = '%d' % i
1782
1783 if self.params.get('list_thumbnails'):
1784 self.list_thumbnails(info_dict)
1785 return
1786
1787 thumbnail = info_dict.get('thumbnail')
1788 if thumbnail:
1789 info_dict['thumbnail'] = sanitize_url(thumbnail)
1790 elif thumbnails:
1791 info_dict['thumbnail'] = thumbnails[-1]['url']
1792
1793 if 'display_id' not in info_dict and 'id' in info_dict:
1794 info_dict['display_id'] = info_dict['id']
1795
1796 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1797 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1798 # see http://bugs.python.org/issue1646728)
1799 try:
1800 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1801 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1802 except (ValueError, OverflowError, OSError):
1803 pass
1804
1805 # Auto generate title fields corresponding to the *_number fields when missing
1806 # in order to always have clean titles. This is very common for TV series.
1807 for field in ('chapter', 'season', 'episode'):
1808 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1809 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1810
1811 for cc_kind in ('subtitles', 'automatic_captions'):
1812 cc = info_dict.get(cc_kind)
1813 if cc:
1814 for _, subtitle in cc.items():
1815 for subtitle_format in subtitle:
1816 if subtitle_format.get('url'):
1817 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1818 if subtitle_format.get('ext') is None:
1819 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1820
1821 automatic_captions = info_dict.get('automatic_captions')
1822 subtitles = info_dict.get('subtitles')
1823
1824 if self.params.get('listsubtitles', False):
1825 if 'automatic_captions' in info_dict:
1826 self.list_subtitles(
1827 info_dict['id'], automatic_captions, 'automatic captions')
1828 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1829 return
1830
1831 info_dict['requested_subtitles'] = self.process_subtitles(
1832 info_dict['id'], subtitles, automatic_captions)
1833
1834 # We now pick which formats have to be downloaded
1835 if info_dict.get('formats') is None:
1836 # There's only one format available
1837 formats = [info_dict]
1838 else:
1839 formats = info_dict['formats']
1840
1841 if not formats:
1842 raise ExtractorError('No video formats found!')
1843
1844 def is_wellformed(f):
1845 url = f.get('url')
1846 if not url:
1847 self.report_warning(
1848 '"url" field is missing or empty - skipping format, '
1849 'there is an error in extractor')
1850 return False
1851 if isinstance(url, bytes):
1852 sanitize_string_field(f, 'url')
1853 return True
1854
1855 # Filter out malformed formats for better extraction robustness
1856 formats = list(filter(is_wellformed, formats))
1857
1858 formats_dict = {}
1859
1860 # We check that all the formats have the format and format_id fields
1861 for i, format in enumerate(formats):
1862 sanitize_string_field(format, 'format_id')
1863 sanitize_numeric_fields(format)
1864 format['url'] = sanitize_url(format['url'])
1865 if not format.get('format_id'):
1866 format['format_id'] = compat_str(i)
1867 else:
1868 # Sanitize format_id from characters used in format selector expression
1869 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
1870 format_id = format['format_id']
1871 if format_id not in formats_dict:
1872 formats_dict[format_id] = []
1873 formats_dict[format_id].append(format)
1874
1875 # Make sure all formats have unique format_id
1876 for format_id, ambiguous_formats in formats_dict.items():
1877 if len(ambiguous_formats) > 1:
1878 for i, format in enumerate(ambiguous_formats):
1879 format['format_id'] = '%s-%d' % (format_id, i)
1880
1881 for i, format in enumerate(formats):
1882 if format.get('format') is None:
1883 format['format'] = '{id} - {res}{note}'.format(
1884 id=format['format_id'],
1885 res=self.format_resolution(format),
1886 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1887 )
1888 # Automatically determine file extension if missing
1889 if format.get('ext') is None:
1890 format['ext'] = determine_ext(format['url']).lower()
1891 # Automatically determine protocol if missing (useful for format
1892 # selection purposes)
1893 if format.get('protocol') is None:
1894 format['protocol'] = determine_protocol(format)
1895 # Add HTTP headers, so that external programs can use them from the
1896 # json output
1897 full_format_info = info_dict.copy()
1898 full_format_info.update(format)
1899 format['http_headers'] = self._calc_headers(full_format_info)
1900 # Remove private housekeeping stuff
1901 if '__x_forwarded_for_ip' in info_dict:
1902 del info_dict['__x_forwarded_for_ip']
1903
1904 # TODO Central sorting goes here
1905
1906 if formats[0] is not info_dict:
1907 # only set the 'formats' fields if the original info_dict list them
1908 # otherwise we end up with a circular reference, the first (and unique)
1909 # element in the 'formats' field in info_dict is info_dict itself,
1910 # which can't be exported to json
1911 info_dict['formats'] = formats
1912 if self.params.get('listformats'):
1913 self.list_formats(info_dict)
1914 return
1915
1916 req_format = self.params.get('format')
1917 if req_format is None:
1918 req_format = self._default_format_spec(info_dict, download=download)
1919 if self.params.get('verbose'):
1920 self.to_screen('[debug] Default format spec: %s' % req_format)
1921
1922 format_selector = self.build_format_selector(req_format)
1923
1924 # While in format selection we may need to have an access to the original
1925 # format set in order to calculate some metrics or do some processing.
1926 # For now we need to be able to guess whether original formats provided
1927 # by extractor are incomplete or not (i.e. whether extractor provides only
1928 # video-only or audio-only formats) for proper formats selection for
1929 # extractors with such incomplete formats (see
1930 # https://github.com/ytdl-org/youtube-dl/pull/5556).
1931 # Since formats may be filtered during format selection and may not match
1932 # the original formats the results may be incorrect. Thus original formats
1933 # or pre-calculated metrics should be passed to format selection routines
1934 # as well.
1935 # We will pass a context object containing all necessary additional data
1936 # instead of just formats.
1937 # This fixes incorrect format selection issue (see
1938 # https://github.com/ytdl-org/youtube-dl/issues/10083).
1939 incomplete_formats = (
1940 # All formats are video-only or
1941 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
1942 # all formats are audio-only
1943 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1944
1945 ctx = {
1946 'formats': formats,
1947 'incomplete_formats': incomplete_formats,
1948 }
1949
1950 formats_to_download = list(format_selector(ctx))
1951 if not formats_to_download:
1952 raise ExtractorError('requested format not available',
1953 expected=True)
1954
1955 if download:
1956 self.to_screen('[info] Downloading format(s) %s' % ", ".join([f['format_id'] for f in formats_to_download]))
1957 if len(formats_to_download) > 1:
1958 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1959 for format in formats_to_download:
1960 new_info = dict(info_dict)
1961 new_info.update(format)
1962 self.process_info(new_info)
1963 # We update the info dict with the best quality format (backwards compatibility)
1964 info_dict.update(formats_to_download[-1])
1965 return info_dict
1966
1967 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1968 """Select the requested subtitles and their format"""
1969 available_subs = {}
1970 if normal_subtitles and self.params.get('writesubtitles'):
1971 available_subs.update(normal_subtitles)
1972 if automatic_captions and self.params.get('writeautomaticsub'):
1973 for lang, cap_info in automatic_captions.items():
1974 if lang not in available_subs:
1975 available_subs[lang] = cap_info
1976
1977 if (not self.params.get('writesubtitles') and not
1978 self.params.get('writeautomaticsub') or not
1979 available_subs):
1980 return None
1981
1982 if self.params.get('allsubtitles', False):
1983 requested_langs = available_subs.keys()
1984 else:
1985 if self.params.get('subtitleslangs', False):
1986 requested_langs = self.params.get('subtitleslangs')
1987 elif 'en' in available_subs:
1988 requested_langs = ['en']
1989 else:
1990 requested_langs = [list(available_subs.keys())[0]]
1991
1992 formats_query = self.params.get('subtitlesformat', 'best')
1993 formats_preference = formats_query.split('/') if formats_query else []
1994 subs = {}
1995 for lang in requested_langs:
1996 formats = available_subs.get(lang)
1997 if formats is None:
1998 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1999 continue
2000 for ext in formats_preference:
2001 if ext == 'best':
2002 f = formats[-1]
2003 break
2004 matches = list(filter(lambda f: f['ext'] == ext, formats))
2005 if matches:
2006 f = matches[-1]
2007 break
2008 else:
2009 f = formats[-1]
2010 self.report_warning(
2011 'No subtitle format found matching "%s" for language %s, '
2012 'using %s' % (formats_query, lang, f['ext']))
2013 subs[lang] = f
2014 return subs
2015
2016 def __forced_printings(self, info_dict, filename, incomplete):
2017 def print_mandatory(field):
2018 if (self.params.get('force%s' % field, False)
2019 and (not incomplete or info_dict.get(field) is not None)):
2020 self.to_stdout(info_dict[field])
2021
2022 def print_optional(field):
2023 if (self.params.get('force%s' % field, False)
2024 and info_dict.get(field) is not None):
2025 self.to_stdout(info_dict[field])
2026
2027 print_mandatory('title')
2028 print_mandatory('id')
2029 if self.params.get('forceurl', False) and not incomplete:
2030 if info_dict.get('requested_formats') is not None:
2031 for f in info_dict['requested_formats']:
2032 self.to_stdout(f['url'] + f.get('play_path', ''))
2033 else:
2034 # For RTMP URLs, also include the playpath
2035 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
2036 print_optional('thumbnail')
2037 print_optional('description')
2038 if self.params.get('forcefilename', False) and filename is not None:
2039 self.to_stdout(filename)
2040 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
2041 self.to_stdout(formatSeconds(info_dict['duration']))
2042 print_mandatory('format')
2043 if self.params.get('forcejson', False):
2044 self.post_extract(info_dict)
2045 self.to_stdout(json.dumps(info_dict))
2046
2047 def process_info(self, info_dict):
2048 """Process a single resolved IE result."""
2049
2050 assert info_dict.get('_type', 'video') == 'video'
2051
2052 info_dict.setdefault('__postprocessors', [])
2053
2054 max_downloads = self.params.get('max_downloads')
2055 if max_downloads is not None:
2056 if self._num_downloads >= int(max_downloads):
2057 raise MaxDownloadsReached()
2058
2059 # TODO: backward compatibility, to be removed
2060 info_dict['fulltitle'] = info_dict['title']
2061
2062 if 'format' not in info_dict:
2063 info_dict['format'] = info_dict['ext']
2064
2065 if self._match_entry(info_dict, incomplete=False) is not None:
2066 return
2067
2068 self.post_extract(info_dict)
2069 self._num_downloads += 1
2070
2071 info_dict = self.pre_process(info_dict)
2072
2073 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2074 temp_filename = self.prepare_filename(info_dict, 'temp')
2075 files_to_move = {}
2076 skip_dl = self.params.get('skip_download', False)
2077
2078 # Forced printings
2079 self.__forced_printings(info_dict, full_filename, incomplete=False)
2080
2081 if self.params.get('simulate', False):
2082 if self.params.get('force_write_download_archive', False):
2083 self.record_download_archive(info_dict)
2084
2085 # Do nothing else if in simulate mode
2086 return
2087
2088 if full_filename is None:
2089 return
2090
2091 def ensure_dir_exists(path):
2092 return make_dir(path, self.report_error)
2093
2094 if not ensure_dir_exists(encodeFilename(full_filename)):
2095 return
2096 if not ensure_dir_exists(encodeFilename(temp_filename)):
2097 return
2098
2099 if self.params.get('writedescription', False):
2100 descfn = self.prepare_filename(info_dict, 'description')
2101 if not ensure_dir_exists(encodeFilename(descfn)):
2102 return
2103 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
2104 self.to_screen('[info] Video description is already present')
2105 elif info_dict.get('description') is None:
2106 self.report_warning('There\'s no description to write.')
2107 else:
2108 try:
2109 self.to_screen('[info] Writing video description to: ' + descfn)
2110 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
2111 descfile.write(info_dict['description'])
2112 except (OSError, IOError):
2113 self.report_error('Cannot write description file ' + descfn)
2114 return
2115
2116 if self.params.get('writeannotations', False):
2117 annofn = self.prepare_filename(info_dict, 'annotation')
2118 if not ensure_dir_exists(encodeFilename(annofn)):
2119 return
2120 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2121 self.to_screen('[info] Video annotations are already present')
2122 elif not info_dict.get('annotations'):
2123 self.report_warning('There are no annotations to write.')
2124 else:
2125 try:
2126 self.to_screen('[info] Writing video annotations to: ' + annofn)
2127 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2128 annofile.write(info_dict['annotations'])
2129 except (KeyError, TypeError):
2130 self.report_warning('There are no annotations to write.')
2131 except (OSError, IOError):
2132 self.report_error('Cannot write annotations file: ' + annofn)
2133 return
2134
2135 def dl(name, info, subtitle=False):
2136 fd = get_suitable_downloader(info, self.params)(self, self.params)
2137 for ph in self._progress_hooks:
2138 fd.add_progress_hook(ph)
2139 if self.params.get('verbose'):
2140 self.to_screen('[debug] Invoking downloader on %r' % info.get('url'))
2141 return fd.download(name, info, subtitle)
2142
2143 subtitles_are_requested = any([self.params.get('writesubtitles', False),
2144 self.params.get('writeautomaticsub')])
2145
2146 if subtitles_are_requested and info_dict.get('requested_subtitles'):
2147 # subtitles download errors are already managed as troubles in relevant IE
2148 # that way it will silently go on when used with unsupporting IE
2149 subtitles = info_dict['requested_subtitles']
2150 # ie = self.get_info_extractor(info_dict['extractor_key'])
2151 for sub_lang, sub_info in subtitles.items():
2152 sub_format = sub_info['ext']
2153 sub_fn = self.prepare_filename(info_dict, 'subtitle')
2154 sub_filename = subtitles_filename(
2155 temp_filename if not skip_dl else sub_fn,
2156 sub_lang, sub_format, info_dict.get('ext'))
2157 sub_filename_final = subtitles_filename(sub_fn, sub_lang, sub_format, info_dict.get('ext'))
2158 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(sub_filename)):
2159 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
2160 files_to_move[sub_filename] = sub_filename_final
2161 else:
2162 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
2163 if sub_info.get('data') is not None:
2164 try:
2165 # Use newline='' to prevent conversion of newline characters
2166 # See https://github.com/ytdl-org/youtube-dl/issues/10268
2167 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
2168 subfile.write(sub_info['data'])
2169 files_to_move[sub_filename] = sub_filename_final
2170 except (OSError, IOError):
2171 self.report_error('Cannot write subtitles file ' + sub_filename)
2172 return
2173 else:
2174 try:
2175 dl(sub_filename, sub_info, subtitle=True)
2176 files_to_move[sub_filename] = sub_filename_final
2177 except (ExtractorError, IOError, OSError, ValueError, compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2178 self.report_warning('Unable to download subtitle for "%s": %s' %
2179 (sub_lang, error_to_compat_str(err)))
2180 continue
2181
2182 if skip_dl:
2183 if self.params.get('convertsubtitles', False):
2184 # subconv = FFmpegSubtitlesConvertorPP(self, format=self.params.get('convertsubtitles'))
2185 filename_real_ext = os.path.splitext(full_filename)[1][1:]
2186 filename_wo_ext = (
2187 os.path.splitext(full_filename)[0]
2188 if filename_real_ext == info_dict['ext']
2189 else full_filename)
2190 afilename = '%s.%s' % (filename_wo_ext, self.params.get('convertsubtitles'))
2191 # if subconv.available:
2192 # info_dict['__postprocessors'].append(subconv)
2193 if os.path.exists(encodeFilename(afilename)):
2194 self.to_screen(
2195 '[download] %s has already been downloaded and '
2196 'converted' % afilename)
2197 else:
2198 try:
2199 self.post_process(full_filename, info_dict, files_to_move)
2200 except PostProcessingError as err:
2201 self.report_error('Postprocessing: %s' % str(err))
2202 return
2203
2204 if self.params.get('writeinfojson', False):
2205 infofn = self.prepare_filename(info_dict, 'infojson')
2206 if not ensure_dir_exists(encodeFilename(infofn)):
2207 return
2208 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
2209 self.to_screen('[info] Video metadata is already present')
2210 else:
2211 self.to_screen('[info] Writing video metadata as JSON to: ' + infofn)
2212 try:
2213 write_json_file(self.filter_requested_info(info_dict), infofn)
2214 except (OSError, IOError):
2215 self.report_error('Cannot write video metadata to JSON file ' + infofn)
2216 return
2217 info_dict['__infojson_filename'] = infofn
2218
2219 thumbfn = self.prepare_filename(info_dict, 'thumbnail')
2220 thumb_fn_temp = temp_filename if not skip_dl else thumbfn
2221 for thumb_ext in self._write_thumbnails(info_dict, thumb_fn_temp):
2222 thumb_filename_temp = replace_extension(thumb_fn_temp, thumb_ext, info_dict.get('ext'))
2223 thumb_filename = replace_extension(thumbfn, thumb_ext, info_dict.get('ext'))
2224 files_to_move[thumb_filename_temp] = info_dict['__thumbnail_filename'] = thumb_filename
2225
2226 # Write internet shortcut files
2227 url_link = webloc_link = desktop_link = False
2228 if self.params.get('writelink', False):
2229 if sys.platform == "darwin": # macOS.
2230 webloc_link = True
2231 elif sys.platform.startswith("linux"):
2232 desktop_link = True
2233 else: # if sys.platform in ['win32', 'cygwin']:
2234 url_link = True
2235 if self.params.get('writeurllink', False):
2236 url_link = True
2237 if self.params.get('writewebloclink', False):
2238 webloc_link = True
2239 if self.params.get('writedesktoplink', False):
2240 desktop_link = True
2241
2242 if url_link or webloc_link or desktop_link:
2243 if 'webpage_url' not in info_dict:
2244 self.report_error('Cannot write internet shortcut file because the "webpage_url" field is missing in the media information')
2245 return
2246 ascii_url = iri_to_uri(info_dict['webpage_url'])
2247
2248 def _write_link_file(extension, template, newline, embed_filename):
2249 linkfn = replace_extension(full_filename, extension, info_dict.get('ext'))
2250 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2251 self.to_screen('[info] Internet shortcut is already present')
2252 else:
2253 try:
2254 self.to_screen('[info] Writing internet shortcut to: ' + linkfn)
2255 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', newline=newline) as linkfile:
2256 template_vars = {'url': ascii_url}
2257 if embed_filename:
2258 template_vars['filename'] = linkfn[:-(len(extension) + 1)]
2259 linkfile.write(template % template_vars)
2260 except (OSError, IOError):
2261 self.report_error('Cannot write internet shortcut ' + linkfn)
2262 return False
2263 return True
2264
2265 if url_link:
2266 if not _write_link_file('url', DOT_URL_LINK_TEMPLATE, '\r\n', embed_filename=False):
2267 return
2268 if webloc_link:
2269 if not _write_link_file('webloc', DOT_WEBLOC_LINK_TEMPLATE, '\n', embed_filename=False):
2270 return
2271 if desktop_link:
2272 if not _write_link_file('desktop', DOT_DESKTOP_LINK_TEMPLATE, '\n', embed_filename=True):
2273 return
2274
2275 # Download
2276 must_record_download_archive = False
2277 if not skip_dl:
2278 try:
2279
2280 def existing_file(*filepaths):
2281 ext = info_dict.get('ext')
2282 final_ext = self.params.get('final_ext', ext)
2283 existing_files = []
2284 for file in orderedSet(filepaths):
2285 if final_ext != ext:
2286 converted = replace_extension(file, final_ext, ext)
2287 if os.path.exists(encodeFilename(converted)):
2288 existing_files.append(converted)
2289 if os.path.exists(encodeFilename(file)):
2290 existing_files.append(file)
2291
2292 if not existing_files or self.params.get('overwrites', False):
2293 for file in orderedSet(existing_files):
2294 self.report_file_delete(file)
2295 os.remove(encodeFilename(file))
2296 return None
2297
2298 self.report_file_already_downloaded(existing_files[0])
2299 info_dict['ext'] = os.path.splitext(existing_files[0])[1][1:]
2300 return existing_files[0]
2301
2302 success = True
2303 if info_dict.get('requested_formats') is not None:
2304 downloaded = []
2305 merger = FFmpegMergerPP(self)
2306 if self.params.get('allow_unplayable_formats'):
2307 self.report_warning(
2308 'You have requested merging of multiple formats '
2309 'while also allowing unplayable formats to be downloaded. '
2310 'The formats won\'t be merged to prevent data corruption.')
2311 elif not merger.available:
2312 self.report_warning(
2313 'You have requested merging of multiple formats but ffmpeg is not installed. '
2314 'The formats won\'t be merged.')
2315
2316 def compatible_formats(formats):
2317 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2318 video_formats = [format for format in formats if format.get('vcodec') != 'none']
2319 audio_formats = [format for format in formats if format.get('acodec') != 'none']
2320 if len(video_formats) > 2 or len(audio_formats) > 2:
2321 return False
2322
2323 # Check extension
2324 exts = set(format.get('ext') for format in formats)
2325 COMPATIBLE_EXTS = (
2326 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
2327 set(('webm',)),
2328 )
2329 for ext_sets in COMPATIBLE_EXTS:
2330 if ext_sets.issuperset(exts):
2331 return True
2332 # TODO: Check acodec/vcodec
2333 return False
2334
2335 requested_formats = info_dict['requested_formats']
2336 old_ext = info_dict['ext']
2337 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
2338 info_dict['ext'] = 'mkv'
2339 self.report_warning(
2340 'Requested formats are incompatible for merge and will be merged into mkv.')
2341
2342 def correct_ext(filename):
2343 filename_real_ext = os.path.splitext(filename)[1][1:]
2344 filename_wo_ext = (
2345 os.path.splitext(filename)[0]
2346 if filename_real_ext == old_ext
2347 else filename)
2348 return '%s.%s' % (filename_wo_ext, info_dict['ext'])
2349
2350 # Ensure filename always has a correct extension for successful merge
2351 full_filename = correct_ext(full_filename)
2352 temp_filename = correct_ext(temp_filename)
2353 dl_filename = existing_file(full_filename, temp_filename)
2354 info_dict['__real_download'] = False
2355 if dl_filename is None:
2356 for f in requested_formats:
2357 new_info = dict(info_dict)
2358 new_info.update(f)
2359 fname = prepend_extension(
2360 self.prepare_filename(new_info, 'temp'),
2361 'f%s' % f['format_id'], new_info['ext'])
2362 if not ensure_dir_exists(fname):
2363 return
2364 downloaded.append(fname)
2365 partial_success, real_download = dl(fname, new_info)
2366 info_dict['__real_download'] = info_dict['__real_download'] or real_download
2367 success = success and partial_success
2368 if merger.available and not self.params.get('allow_unplayable_formats'):
2369 info_dict['__postprocessors'].append(merger)
2370 info_dict['__files_to_merge'] = downloaded
2371 # Even if there were no downloads, it is being merged only now
2372 info_dict['__real_download'] = True
2373 else:
2374 for file in downloaded:
2375 files_to_move[file] = None
2376 else:
2377 # Just a single file
2378 dl_filename = existing_file(full_filename, temp_filename)
2379 if dl_filename is None:
2380 success, real_download = dl(temp_filename, info_dict)
2381 info_dict['__real_download'] = real_download
2382
2383 dl_filename = dl_filename or temp_filename
2384 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2385
2386 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2387 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
2388 return
2389 except (OSError, IOError) as err:
2390 raise UnavailableVideoError(err)
2391 except (ContentTooShortError, ) as err:
2392 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
2393 return
2394
2395 if success and full_filename != '-':
2396 # Fixup content
2397 fixup_policy = self.params.get('fixup')
2398 if fixup_policy is None:
2399 fixup_policy = 'detect_or_warn'
2400
2401 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg to fix this automatically.'
2402
2403 stretched_ratio = info_dict.get('stretched_ratio')
2404 if stretched_ratio is not None and stretched_ratio != 1:
2405 if fixup_policy == 'warn':
2406 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
2407 info_dict['id'], stretched_ratio))
2408 elif fixup_policy == 'detect_or_warn':
2409 stretched_pp = FFmpegFixupStretchedPP(self)
2410 if stretched_pp.available:
2411 info_dict['__postprocessors'].append(stretched_pp)
2412 else:
2413 self.report_warning(
2414 '%s: Non-uniform pixel ratio (%s). %s'
2415 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
2416 else:
2417 assert fixup_policy in ('ignore', 'never')
2418
2419 if (info_dict.get('requested_formats') is None
2420 and info_dict.get('container') == 'm4a_dash'
2421 and info_dict.get('ext') == 'm4a'):
2422 if fixup_policy == 'warn':
2423 self.report_warning(
2424 '%s: writing DASH m4a. '
2425 'Only some players support this container.'
2426 % info_dict['id'])
2427 elif fixup_policy == 'detect_or_warn':
2428 fixup_pp = FFmpegFixupM4aPP(self)
2429 if fixup_pp.available:
2430 info_dict['__postprocessors'].append(fixup_pp)
2431 else:
2432 self.report_warning(
2433 '%s: writing DASH m4a. '
2434 'Only some players support this container. %s'
2435 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
2436 else:
2437 assert fixup_policy in ('ignore', 'never')
2438
2439 if (info_dict.get('protocol') == 'm3u8_native'
2440 or info_dict.get('protocol') == 'm3u8'
2441 and self.params.get('hls_prefer_native')):
2442 if fixup_policy == 'warn':
2443 self.report_warning('%s: malformed AAC bitstream detected.' % (
2444 info_dict['id']))
2445 elif fixup_policy == 'detect_or_warn':
2446 fixup_pp = FFmpegFixupM3u8PP(self)
2447 if fixup_pp.available:
2448 info_dict['__postprocessors'].append(fixup_pp)
2449 else:
2450 self.report_warning(
2451 '%s: malformed AAC bitstream detected. %s'
2452 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
2453 else:
2454 assert fixup_policy in ('ignore', 'never')
2455
2456 try:
2457 self.post_process(dl_filename, info_dict, files_to_move)
2458 except PostProcessingError as err:
2459 self.report_error('Postprocessing: %s' % str(err))
2460 return
2461 try:
2462 for ph in self._post_hooks:
2463 ph(full_filename)
2464 except Exception as err:
2465 self.report_error('post hooks: %s' % str(err))
2466 return
2467 must_record_download_archive = True
2468
2469 if must_record_download_archive or self.params.get('force_write_download_archive', False):
2470 self.record_download_archive(info_dict)
2471 max_downloads = self.params.get('max_downloads')
2472 if max_downloads is not None and self._num_downloads >= int(max_downloads):
2473 raise MaxDownloadsReached()
2474
2475 def download(self, url_list):
2476 """Download a given list of URLs."""
2477 outtmpl = self.outtmpl_dict['default']
2478 if (len(url_list) > 1
2479 and outtmpl != '-'
2480 and '%' not in outtmpl
2481 and self.params.get('max_downloads') != 1):
2482 raise SameFileError(outtmpl)
2483
2484 for url in url_list:
2485 try:
2486 # It also downloads the videos
2487 res = self.extract_info(
2488 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2489 except UnavailableVideoError:
2490 self.report_error('unable to download video')
2491 except MaxDownloadsReached:
2492 self.to_screen('[info] Maximum number of downloaded files reached')
2493 raise
2494 except ExistingVideoReached:
2495 self.to_screen('[info] Encountered a file that is already in the archive, stopping due to --break-on-existing')
2496 raise
2497 except RejectedVideoReached:
2498 self.to_screen('[info] Encountered a file that did not match filter, stopping due to --break-on-reject')
2499 raise
2500 else:
2501 if self.params.get('dump_single_json', False):
2502 self.post_extract(res)
2503 self.to_stdout(json.dumps(res))
2504
2505 return self._download_retcode
2506
2507 def download_with_info_file(self, info_filename):
2508 with contextlib.closing(fileinput.FileInput(
2509 [info_filename], mode='r',
2510 openhook=fileinput.hook_encoded('utf-8'))) as f:
2511 # FileInput doesn't have a read method, we can't call json.load
2512 info = self.filter_requested_info(json.loads('\n'.join(f)))
2513 try:
2514 self.process_ie_result(info, download=True)
2515 except DownloadError:
2516 webpage_url = info.get('webpage_url')
2517 if webpage_url is not None:
2518 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2519 return self.download([webpage_url])
2520 else:
2521 raise
2522 return self._download_retcode
2523
2524 @staticmethod
2525 def filter_requested_info(info_dict):
2526 fields_to_remove = ('requested_formats', 'requested_subtitles')
2527 return dict(
2528 (k, v) for k, v in info_dict.items()
2529 if (k[0] != '_' or k == '_type') and k not in fields_to_remove)
2530
2531 def run_pp(self, pp, infodict, files_to_move={}):
2532 files_to_delete = []
2533 files_to_delete, infodict = pp.run(infodict)
2534 if not files_to_delete:
2535 return files_to_move, infodict
2536
2537 if self.params.get('keepvideo', False):
2538 for f in files_to_delete:
2539 files_to_move.setdefault(f, '')
2540 else:
2541 for old_filename in set(files_to_delete):
2542 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2543 try:
2544 os.remove(encodeFilename(old_filename))
2545 except (IOError, OSError):
2546 self.report_warning('Unable to remove downloaded original file')
2547 if old_filename in files_to_move:
2548 del files_to_move[old_filename]
2549 return files_to_move, infodict
2550
2551 @staticmethod
2552 def post_extract(info_dict):
2553 def actual_post_extract(info_dict):
2554 if info_dict.get('_type') in ('playlist', 'multi_video'):
2555 for video_dict in info_dict.get('entries', {}):
2556 actual_post_extract(video_dict)
2557 return
2558
2559 if '__post_extractor' not in info_dict:
2560 return
2561 post_extractor = info_dict['__post_extractor']
2562 if post_extractor:
2563 info_dict.update(post_extractor().items())
2564 del info_dict['__post_extractor']
2565 return
2566
2567 actual_post_extract(info_dict)
2568
2569 def pre_process(self, ie_info):
2570 info = dict(ie_info)
2571 for pp in self._pps['beforedl']:
2572 info = self.run_pp(pp, info)[1]
2573 return info
2574
2575 def post_process(self, filename, ie_info, files_to_move={}):
2576 """Run all the postprocessors on the given file."""
2577 info = dict(ie_info)
2578 info['filepath'] = filename
2579 info['__files_to_move'] = {}
2580
2581 for pp in ie_info.get('__postprocessors', []) + self._pps['normal']:
2582 files_to_move, info = self.run_pp(pp, info, files_to_move)
2583 info = self.run_pp(MoveFilesAfterDownloadPP(self, files_to_move), info)[1]
2584 for pp in self._pps['aftermove']:
2585 info = self.run_pp(pp, info, {})[1]
2586
2587 def _make_archive_id(self, info_dict):
2588 video_id = info_dict.get('id')
2589 if not video_id:
2590 return
2591 # Future-proof against any change in case
2592 # and backwards compatibility with prior versions
2593 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
2594 if extractor is None:
2595 url = str_or_none(info_dict.get('url'))
2596 if not url:
2597 return
2598 # Try to find matching extractor for the URL and take its ie_key
2599 for ie in self._ies:
2600 if ie.suitable(url):
2601 extractor = ie.ie_key()
2602 break
2603 else:
2604 return
2605 return '%s %s' % (extractor.lower(), video_id)
2606
2607 def in_download_archive(self, info_dict):
2608 fn = self.params.get('download_archive')
2609 if fn is None:
2610 return False
2611
2612 vid_id = self._make_archive_id(info_dict)
2613 if not vid_id:
2614 return False # Incomplete video information
2615
2616 return vid_id in self.archive
2617
2618 def record_download_archive(self, info_dict):
2619 fn = self.params.get('download_archive')
2620 if fn is None:
2621 return
2622 vid_id = self._make_archive_id(info_dict)
2623 assert vid_id
2624 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2625 archive_file.write(vid_id + '\n')
2626 self.archive.add(vid_id)
2627
2628 @staticmethod
2629 def format_resolution(format, default='unknown'):
2630 if format.get('vcodec') == 'none':
2631 return 'audio only'
2632 if format.get('resolution') is not None:
2633 return format['resolution']
2634 if format.get('height') is not None:
2635 if format.get('width') is not None:
2636 res = '%sx%s' % (format['width'], format['height'])
2637 else:
2638 res = '%sp' % format['height']
2639 elif format.get('width') is not None:
2640 res = '%dx?' % format['width']
2641 else:
2642 res = default
2643 return res
2644
2645 def _format_note(self, fdict):
2646 res = ''
2647 if fdict.get('ext') in ['f4f', 'f4m']:
2648 res += '(unsupported) '
2649 if fdict.get('language'):
2650 if res:
2651 res += ' '
2652 res += '[%s] ' % fdict['language']
2653 if fdict.get('format_note') is not None:
2654 res += fdict['format_note'] + ' '
2655 if fdict.get('tbr') is not None:
2656 res += '%4dk ' % fdict['tbr']
2657 if fdict.get('container') is not None:
2658 if res:
2659 res += ', '
2660 res += '%s container' % fdict['container']
2661 if (fdict.get('vcodec') is not None
2662 and fdict.get('vcodec') != 'none'):
2663 if res:
2664 res += ', '
2665 res += fdict['vcodec']
2666 if fdict.get('vbr') is not None:
2667 res += '@'
2668 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2669 res += 'video@'
2670 if fdict.get('vbr') is not None:
2671 res += '%4dk' % fdict['vbr']
2672 if fdict.get('fps') is not None:
2673 if res:
2674 res += ', '
2675 res += '%sfps' % fdict['fps']
2676 if fdict.get('acodec') is not None:
2677 if res:
2678 res += ', '
2679 if fdict['acodec'] == 'none':
2680 res += 'video only'
2681 else:
2682 res += '%-5s' % fdict['acodec']
2683 elif fdict.get('abr') is not None:
2684 if res:
2685 res += ', '
2686 res += 'audio'
2687 if fdict.get('abr') is not None:
2688 res += '@%3dk' % fdict['abr']
2689 if fdict.get('asr') is not None:
2690 res += ' (%5dHz)' % fdict['asr']
2691 if fdict.get('filesize') is not None:
2692 if res:
2693 res += ', '
2694 res += format_bytes(fdict['filesize'])
2695 elif fdict.get('filesize_approx') is not None:
2696 if res:
2697 res += ', '
2698 res += '~' + format_bytes(fdict['filesize_approx'])
2699 return res
2700
2701 def _format_note_table(self, f):
2702 def join_fields(*vargs):
2703 return ', '.join((val for val in vargs if val != ''))
2704
2705 return join_fields(
2706 'UNSUPPORTED' if f.get('ext') in ('f4f', 'f4m') else '',
2707 format_field(f, 'language', '[%s]'),
2708 format_field(f, 'format_note'),
2709 format_field(f, 'container', ignore=(None, f.get('ext'))),
2710 format_field(f, 'asr', '%5dHz'))
2711
2712 def list_formats(self, info_dict):
2713 formats = info_dict.get('formats', [info_dict])
2714 new_format = self.params.get('listformats_table', False)
2715 if new_format:
2716 table = [
2717 [
2718 format_field(f, 'format_id'),
2719 format_field(f, 'ext'),
2720 self.format_resolution(f),
2721 format_field(f, 'fps', '%d'),
2722 '|',
2723 format_field(f, 'filesize', ' %s', func=format_bytes) + format_field(f, 'filesize_approx', '~%s', func=format_bytes),
2724 format_field(f, 'tbr', '%4dk'),
2725 f.get('protocol').replace('http_dash_segments', 'dash').replace("native", "n").replace('niconico_', ''),
2726 '|',
2727 format_field(f, 'vcodec', default='unknown').replace('none', ''),
2728 format_field(f, 'vbr', '%4dk'),
2729 format_field(f, 'acodec', default='unknown').replace('none', ''),
2730 format_field(f, 'abr', '%3dk'),
2731 format_field(f, 'asr', '%5dHz'),
2732 self._format_note_table(f)]
2733 for f in formats
2734 if f.get('preference') is None or f['preference'] >= -1000]
2735 header_line = ['ID', 'EXT', 'RESOLUTION', 'FPS', '|', ' FILESIZE', ' TBR', 'PROTO',
2736 '|', 'VCODEC', ' VBR', 'ACODEC', ' ABR', ' ASR', 'NOTE']
2737 else:
2738 table = [
2739 [
2740 format_field(f, 'format_id'),
2741 format_field(f, 'ext'),
2742 self.format_resolution(f),
2743 self._format_note(f)]
2744 for f in formats
2745 if f.get('preference') is None or f['preference'] >= -1000]
2746 header_line = ['format code', 'extension', 'resolution', 'note']
2747
2748 self.to_screen(
2749 '[info] Available formats for %s:\n%s' % (info_dict['id'], render_table(
2750 header_line,
2751 table,
2752 delim=new_format,
2753 extraGap=(0 if new_format else 1),
2754 hideEmpty=new_format)))
2755
2756 def list_thumbnails(self, info_dict):
2757 thumbnails = info_dict.get('thumbnails')
2758 if not thumbnails:
2759 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
2760 return
2761
2762 self.to_screen(
2763 '[info] Thumbnails for %s:' % info_dict['id'])
2764 self.to_screen(render_table(
2765 ['ID', 'width', 'height', 'URL'],
2766 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
2767
2768 def list_subtitles(self, video_id, subtitles, name='subtitles'):
2769 if not subtitles:
2770 self.to_screen('%s has no %s' % (video_id, name))
2771 return
2772 self.to_screen(
2773 'Available %s for %s:' % (name, video_id))
2774 self.to_screen(render_table(
2775 ['Language', 'formats'],
2776 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
2777 for lang, formats in subtitles.items()]))
2778
2779 def urlopen(self, req):
2780 """ Start an HTTP download """
2781 if isinstance(req, compat_basestring):
2782 req = sanitized_Request(req)
2783 return self._opener.open(req, timeout=self._socket_timeout)
2784
2785 def print_debug_header(self):
2786 if not self.params.get('verbose'):
2787 return
2788
2789 if type('') is not compat_str:
2790 # Python 2.6 on SLES11 SP1 (https://github.com/ytdl-org/youtube-dl/issues/3326)
2791 self.report_warning(
2792 'Your Python is broken! Update to a newer and supported version')
2793
2794 stdout_encoding = getattr(
2795 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2796 encoding_str = (
2797 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2798 locale.getpreferredencoding(),
2799 sys.getfilesystemencoding(),
2800 stdout_encoding,
2801 self.get_encoding()))
2802 write_string(encoding_str, encoding=None)
2803
2804 source = (
2805 '(exe)' if hasattr(sys, 'frozen')
2806 else '(zip)' if isinstance(globals().get('__loader__'), zipimporter)
2807 else '(source)' if os.path.basename(sys.argv[0]) == '__main__.py'
2808 else '')
2809 self._write_string('[debug] yt-dlp version %s %s\n' % (__version__, source))
2810 if _LAZY_LOADER:
2811 self._write_string('[debug] Lazy loading extractors enabled\n')
2812 if _PLUGIN_CLASSES:
2813 self._write_string(
2814 '[debug] Plugin Extractors: %s\n' % [ie.ie_key() for ie in _PLUGIN_CLASSES])
2815 try:
2816 sp = subprocess.Popen(
2817 ['git', 'rev-parse', '--short', 'HEAD'],
2818 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2819 cwd=os.path.dirname(os.path.abspath(__file__)))
2820 out, err = process_communicate_or_kill(sp)
2821 out = out.decode().strip()
2822 if re.match('[0-9a-f]+', out):
2823 self._write_string('[debug] Git HEAD: %s\n' % out)
2824 except Exception:
2825 try:
2826 sys.exc_clear()
2827 except Exception:
2828 pass
2829
2830 def python_implementation():
2831 impl_name = platform.python_implementation()
2832 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
2833 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
2834 return impl_name
2835
2836 self._write_string('[debug] Python version %s (%s %s) - %s\n' % (
2837 platform.python_version(),
2838 python_implementation(),
2839 platform.architecture()[0],
2840 platform_name()))
2841
2842 exe_versions = FFmpegPostProcessor.get_versions(self)
2843 exe_versions['rtmpdump'] = rtmpdump_version()
2844 exe_versions['phantomjs'] = PhantomJSwrapper._version()
2845 exe_str = ', '.join(
2846 '%s %s' % (exe, v)
2847 for exe, v in sorted(exe_versions.items())
2848 if v
2849 )
2850 if not exe_str:
2851 exe_str = 'none'
2852 self._write_string('[debug] exe versions: %s\n' % exe_str)
2853
2854 proxy_map = {}
2855 for handler in self._opener.handlers:
2856 if hasattr(handler, 'proxies'):
2857 proxy_map.update(handler.proxies)
2858 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2859
2860 if self.params.get('call_home', False):
2861 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2862 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2863 return
2864 latest_version = self.urlopen(
2865 'https://yt-dl.org/latest/version').read().decode('utf-8')
2866 if version_tuple(latest_version) > version_tuple(__version__):
2867 self.report_warning(
2868 'You are using an outdated version (newest version: %s)! '
2869 'See https://yt-dl.org/update if you need help updating.' %
2870 latest_version)
2871
2872 def _setup_opener(self):
2873 timeout_val = self.params.get('socket_timeout')
2874 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2875
2876 opts_cookiefile = self.params.get('cookiefile')
2877 opts_proxy = self.params.get('proxy')
2878
2879 if opts_cookiefile is None:
2880 self.cookiejar = compat_cookiejar.CookieJar()
2881 else:
2882 opts_cookiefile = expand_path(opts_cookiefile)
2883 self.cookiejar = YoutubeDLCookieJar(opts_cookiefile)
2884 if os.access(opts_cookiefile, os.R_OK):
2885 self.cookiejar.load(ignore_discard=True, ignore_expires=True)
2886
2887 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2888 if opts_proxy is not None:
2889 if opts_proxy == '':
2890 proxies = {}
2891 else:
2892 proxies = {'http': opts_proxy, 'https': opts_proxy}
2893 else:
2894 proxies = compat_urllib_request.getproxies()
2895 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
2896 if 'http' in proxies and 'https' not in proxies:
2897 proxies['https'] = proxies['http']
2898 proxy_handler = PerRequestProxyHandler(proxies)
2899
2900 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2901 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2902 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2903 redirect_handler = YoutubeDLRedirectHandler()
2904 data_handler = compat_urllib_request_DataHandler()
2905
2906 # When passing our own FileHandler instance, build_opener won't add the
2907 # default FileHandler and allows us to disable the file protocol, which
2908 # can be used for malicious purposes (see
2909 # https://github.com/ytdl-org/youtube-dl/issues/8227)
2910 file_handler = compat_urllib_request.FileHandler()
2911
2912 def file_open(*args, **kwargs):
2913 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
2914 file_handler.file_open = file_open
2915
2916 opener = compat_urllib_request.build_opener(
2917 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
2918
2919 # Delete the default user-agent header, which would otherwise apply in
2920 # cases where our custom HTTP handler doesn't come into play
2921 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
2922 opener.addheaders = []
2923 self._opener = opener
2924
2925 def encode(self, s):
2926 if isinstance(s, bytes):
2927 return s # Already encoded
2928
2929 try:
2930 return s.encode(self.get_encoding())
2931 except UnicodeEncodeError as err:
2932 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2933 raise
2934
2935 def get_encoding(self):
2936 encoding = self.params.get('encoding')
2937 if encoding is None:
2938 encoding = preferredencoding()
2939 return encoding
2940
2941 def _write_thumbnails(self, info_dict, filename): # return the extensions
2942 write_all = self.params.get('write_all_thumbnails', False)
2943 thumbnails = []
2944 if write_all or self.params.get('writethumbnail', False):
2945 thumbnails = info_dict.get('thumbnails') or []
2946 multiple = write_all and len(thumbnails) > 1
2947
2948 ret = []
2949 for t in thumbnails[::1 if write_all else -1]:
2950 thumb_ext = determine_ext(t['url'], 'jpg')
2951 suffix = '%s.' % t['id'] if multiple else ''
2952 thumb_display_id = '%s ' % t['id'] if multiple else ''
2953 t['filename'] = thumb_filename = replace_extension(filename, suffix + thumb_ext, info_dict.get('ext'))
2954
2955 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(thumb_filename)):
2956 ret.append(suffix + thumb_ext)
2957 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2958 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2959 else:
2960 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2961 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2962 try:
2963 uf = self.urlopen(t['url'])
2964 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2965 shutil.copyfileobj(uf, thumbf)
2966 ret.append(suffix + thumb_ext)
2967 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2968 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2969 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2970 self.report_warning('Unable to download thumbnail "%s": %s' %
2971 (t['url'], error_to_compat_str(err)))
2972 if ret and not write_all:
2973 break
2974 return ret