]> jfr.im git - yt-dlp.git/blob - youtube_dl/YoutubeDL.py
[YoutubeDL] Propagate overridden metadata to IE results of type url (closes #11163)
[yt-dlp.git] / youtube_dl / 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 .compat import (
30 compat_basestring,
31 compat_cookiejar,
32 compat_get_terminal_size,
33 compat_http_client,
34 compat_kwargs,
35 compat_numeric_types,
36 compat_os_name,
37 compat_str,
38 compat_tokenize_tokenize,
39 compat_urllib_error,
40 compat_urllib_request,
41 compat_urllib_request_DataHandler,
42 )
43 from .utils import (
44 age_restricted,
45 args_to_str,
46 ContentTooShortError,
47 date_from_str,
48 DateRange,
49 DEFAULT_OUTTMPL,
50 determine_ext,
51 determine_protocol,
52 DownloadError,
53 encode_compat_str,
54 encodeFilename,
55 error_to_compat_str,
56 expand_path,
57 ExtractorError,
58 format_bytes,
59 formatSeconds,
60 GeoRestrictedError,
61 ISO3166Utils,
62 locked_file,
63 make_HTTPS_handler,
64 MaxDownloadsReached,
65 PagedList,
66 parse_filesize,
67 PerRequestProxyHandler,
68 platform_name,
69 PostProcessingError,
70 preferredencoding,
71 prepend_extension,
72 register_socks_protocols,
73 render_table,
74 replace_extension,
75 SameFileError,
76 sanitize_filename,
77 sanitize_path,
78 sanitize_url,
79 sanitized_Request,
80 std_headers,
81 subtitles_filename,
82 UnavailableVideoError,
83 url_basename,
84 version_tuple,
85 write_json_file,
86 write_string,
87 YoutubeDLCookieProcessor,
88 YoutubeDLHandler,
89 )
90 from .cache import Cache
91 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
92 from .downloader import get_suitable_downloader
93 from .downloader.rtmp import rtmpdump_version
94 from .postprocessor import (
95 FFmpegFixupM3u8PP,
96 FFmpegFixupM4aPP,
97 FFmpegFixupStretchedPP,
98 FFmpegMergerPP,
99 FFmpegPostProcessor,
100 get_postprocessor,
101 )
102 from .version import __version__
103
104 if compat_os_name == 'nt':
105 import ctypes
106
107
108 class YoutubeDL(object):
109 """YoutubeDL class.
110
111 YoutubeDL objects are the ones responsible of downloading the
112 actual video file and writing it to disk if the user has requested
113 it, among some other tasks. In most cases there should be one per
114 program. As, given a video URL, the downloader doesn't know how to
115 extract all the needed information, task that InfoExtractors do, it
116 has to pass the URL to one of them.
117
118 For this, YoutubeDL objects have a method that allows
119 InfoExtractors to be registered in a given order. When it is passed
120 a URL, the YoutubeDL object handles it to the first InfoExtractor it
121 finds that reports being able to handle it. The InfoExtractor extracts
122 all the information about the video or videos the URL refers to, and
123 YoutubeDL process the extracted information, possibly using a File
124 Downloader to download the video.
125
126 YoutubeDL objects accept a lot of parameters. In order not to saturate
127 the object constructor with arguments, it receives a dictionary of
128 options instead. These options are available through the params
129 attribute for the InfoExtractors to use. The YoutubeDL also
130 registers itself as the downloader in charge for the InfoExtractors
131 that are added to it, so this is a "mutual registration".
132
133 Available options:
134
135 username: Username for authentication purposes.
136 password: Password for authentication purposes.
137 videopassword: Password for accessing a video.
138 ap_mso: Adobe Pass multiple-system operator identifier.
139 ap_username: Multiple-system operator account username.
140 ap_password: Multiple-system operator account password.
141 usenetrc: Use netrc for authentication instead.
142 verbose: Print additional info to stdout.
143 quiet: Do not print messages to stdout.
144 no_warnings: Do not print out anything for warnings.
145 forceurl: Force printing final URL.
146 forcetitle: Force printing title.
147 forceid: Force printing ID.
148 forcethumbnail: Force printing thumbnail URL.
149 forcedescription: Force printing description.
150 forcefilename: Force printing final filename.
151 forceduration: Force printing duration.
152 forcejson: Force printing info_dict as JSON.
153 dump_single_json: Force printing the info_dict of the whole playlist
154 (or video) as a single JSON line.
155 simulate: Do not download the video files.
156 format: Video format code. See options.py for more information.
157 outtmpl: Template for output names.
158 restrictfilenames: Do not allow "&" and spaces in file names
159 ignoreerrors: Do not stop on download errors.
160 force_generic_extractor: Force downloader to use the generic extractor
161 nooverwrites: Prevent overwriting files.
162 playliststart: Playlist item to start at.
163 playlistend: Playlist item to end at.
164 playlist_items: Specific indices of playlist to download.
165 playlistreverse: Download playlist items in reverse order.
166 playlistrandom: Download playlist items in random order.
167 matchtitle: Download only matching titles.
168 rejecttitle: Reject downloads for matching titles.
169 logger: Log messages to a logging.Logger instance.
170 logtostderr: Log messages to stderr instead of stdout.
171 writedescription: Write the video description to a .description file
172 writeinfojson: Write the video description to a .info.json file
173 writeannotations: Write the video annotations to a .annotations.xml file
174 writethumbnail: Write the thumbnail image to a file
175 write_all_thumbnails: Write all thumbnail formats to files
176 writesubtitles: Write the video subtitles to a file
177 writeautomaticsub: Write the automatically generated subtitles to a file
178 allsubtitles: Downloads all the subtitles of the video
179 (requires writesubtitles or writeautomaticsub)
180 listsubtitles: Lists all available subtitles for the video
181 subtitlesformat: The format code for subtitles
182 subtitleslangs: List of languages of the subtitles to download
183 keepvideo: Keep the video file after post-processing
184 daterange: A DateRange object, download only if the upload_date is in the range.
185 skip_download: Skip the actual download of the video file
186 cachedir: Location of the cache files in the filesystem.
187 False to disable filesystem cache.
188 noplaylist: Download single video instead of a playlist if in doubt.
189 age_limit: An integer representing the user's age in years.
190 Unsuitable videos for the given age are skipped.
191 min_views: An integer representing the minimum view count the video
192 must have in order to not be skipped.
193 Videos without view count information are always
194 downloaded. None for no limit.
195 max_views: An integer representing the maximum view count.
196 Videos that are more popular than that are not
197 downloaded.
198 Videos without view count information are always
199 downloaded. None for no limit.
200 download_archive: File name of a file where all downloads are recorded.
201 Videos already present in the file are not downloaded
202 again.
203 cookiefile: File name where cookies should be read from and dumped to.
204 nocheckcertificate:Do not verify SSL certificates
205 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
206 At the moment, this is only supported by YouTube.
207 proxy: URL of the proxy server to use
208 geo_verification_proxy: URL of the proxy to use for IP address verification
209 on geo-restricted sites. (Experimental)
210 socket_timeout: Time to wait for unresponsive hosts, in seconds
211 bidi_workaround: Work around buggy terminals without bidirectional text
212 support, using fridibi
213 debug_printtraffic:Print out sent and received HTTP traffic
214 include_ads: Download ads as well
215 default_search: Prepend this string if an input url is not valid.
216 'auto' for elaborate guessing
217 encoding: Use this encoding instead of the system-specified.
218 extract_flat: Do not resolve URLs, return the immediate result.
219 Pass in 'in_playlist' to only show this behavior for
220 playlist items.
221 postprocessors: A list of dictionaries, each with an entry
222 * key: The name of the postprocessor. See
223 youtube_dl/postprocessor/__init__.py for a list.
224 as well as any further keyword arguments for the
225 postprocessor.
226 progress_hooks: A list of functions that get called on download
227 progress, with a dictionary with the entries
228 * status: One of "downloading", "error", or "finished".
229 Check this first and ignore unknown values.
230
231 If status is one of "downloading", or "finished", the
232 following properties may also be present:
233 * filename: The final filename (always present)
234 * tmpfilename: The filename we're currently writing to
235 * downloaded_bytes: Bytes on disk
236 * total_bytes: Size of the whole file, None if unknown
237 * total_bytes_estimate: Guess of the eventual file size,
238 None if unavailable.
239 * elapsed: The number of seconds since download started.
240 * eta: The estimated time in seconds, None if unknown
241 * speed: The download speed in bytes/second, None if
242 unknown
243 * fragment_index: The counter of the currently
244 downloaded video fragment.
245 * fragment_count: The number of fragments (= individual
246 files that will be merged)
247
248 Progress hooks are guaranteed to be called at least once
249 (with status "finished") if the download is successful.
250 merge_output_format: Extension to use when merging formats.
251 fixup: Automatically correct known faults of the file.
252 One of:
253 - "never": do nothing
254 - "warn": only emit a warning
255 - "detect_or_warn": check whether we can do anything
256 about it, warn otherwise (default)
257 source_address: (Experimental) Client-side IP address to bind to.
258 call_home: Boolean, true iff we are allowed to contact the
259 youtube-dl servers for debugging.
260 sleep_interval: Number of seconds to sleep before each download when
261 used alone or a lower bound of a range for randomized
262 sleep before each download (minimum possible number
263 of seconds to sleep) when used along with
264 max_sleep_interval.
265 max_sleep_interval:Upper bound of a range for randomized sleep before each
266 download (maximum possible number of seconds to sleep).
267 Must only be used along with sleep_interval.
268 Actual sleep time will be a random float from range
269 [sleep_interval; max_sleep_interval].
270 listformats: Print an overview of available video formats and exit.
271 list_thumbnails: Print a table of all thumbnails and exit.
272 match_filter: A function that gets called with the info_dict of
273 every video.
274 If it returns a message, the video is ignored.
275 If it returns None, the video is downloaded.
276 match_filter_func in utils.py is one example for this.
277 no_color: Do not emit color codes in output.
278 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
279 HTTP header (experimental)
280 geo_bypass_country:
281 Two-letter ISO 3166-2 country code that will be used for
282 explicit geographic restriction bypassing via faking
283 X-Forwarded-For HTTP header (experimental)
284
285 The following options determine which downloader is picked:
286 external_downloader: Executable of the external downloader to call.
287 None or unset for standard (built-in) downloader.
288 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
289 if True, otherwise use ffmpeg/avconv if False, otherwise
290 use downloader suggested by extractor if None.
291
292 The following parameters are not used by YoutubeDL itself, they are used by
293 the downloader (see youtube_dl/downloader/common.py):
294 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
295 noresizebuffer, retries, continuedl, noprogress, consoletitle,
296 xattr_set_filesize, external_downloader_args, hls_use_mpegts.
297
298 The following options are used by the post processors:
299 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
300 otherwise prefer avconv.
301 postprocessor_args: A list of additional command-line arguments for the
302 postprocessor.
303 """
304
305 params = None
306 _ies = []
307 _pps = []
308 _download_retcode = None
309 _num_downloads = None
310 _screen_file = None
311
312 def __init__(self, params=None, auto_init=True):
313 """Create a FileDownloader object with the given options."""
314 if params is None:
315 params = {}
316 self._ies = []
317 self._ies_instances = {}
318 self._pps = []
319 self._progress_hooks = []
320 self._download_retcode = 0
321 self._num_downloads = 0
322 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
323 self._err_file = sys.stderr
324 self.params = {
325 # Default parameters
326 'nocheckcertificate': False,
327 }
328 self.params.update(params)
329 self.cache = Cache(self)
330
331 def check_deprecated(param, option, suggestion):
332 if self.params.get(param) is not None:
333 self.report_warning(
334 '%s is deprecated. Use %s instead.' % (option, suggestion))
335 return True
336 return False
337
338 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
339 if self.params.get('geo_verification_proxy') is None:
340 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
341
342 check_deprecated('autonumber_size', '--autonumber-size', 'output template with %(autonumber)0Nd, where N in the number of digits')
343 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
344 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
345
346 if params.get('bidi_workaround', False):
347 try:
348 import pty
349 master, slave = pty.openpty()
350 width = compat_get_terminal_size().columns
351 if width is None:
352 width_args = []
353 else:
354 width_args = ['-w', str(width)]
355 sp_kwargs = dict(
356 stdin=subprocess.PIPE,
357 stdout=slave,
358 stderr=self._err_file)
359 try:
360 self._output_process = subprocess.Popen(
361 ['bidiv'] + width_args, **sp_kwargs
362 )
363 except OSError:
364 self._output_process = subprocess.Popen(
365 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
366 self._output_channel = os.fdopen(master, 'rb')
367 except OSError as ose:
368 if ose.errno == errno.ENOENT:
369 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 $PATH.')
370 else:
371 raise
372
373 if (sys.version_info >= (3,) and sys.platform != 'win32' and
374 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
375 not params.get('restrictfilenames', False)):
376 # On Python 3, the Unicode filesystem API will throw errors (#1474)
377 self.report_warning(
378 'Assuming --restrict-filenames since file system encoding '
379 'cannot encode all characters. '
380 'Set the LC_ALL environment variable to fix this.')
381 self.params['restrictfilenames'] = True
382
383 if isinstance(params.get('outtmpl'), bytes):
384 self.report_warning(
385 'Parameter outtmpl is bytes, but should be a unicode string. '
386 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
387
388 self._setup_opener()
389
390 if auto_init:
391 self.print_debug_header()
392 self.add_default_info_extractors()
393
394 for pp_def_raw in self.params.get('postprocessors', []):
395 pp_class = get_postprocessor(pp_def_raw['key'])
396 pp_def = dict(pp_def_raw)
397 del pp_def['key']
398 pp = pp_class(self, **compat_kwargs(pp_def))
399 self.add_post_processor(pp)
400
401 for ph in self.params.get('progress_hooks', []):
402 self.add_progress_hook(ph)
403
404 register_socks_protocols()
405
406 def warn_if_short_id(self, argv):
407 # short YouTube ID starting with dash?
408 idxs = [
409 i for i, a in enumerate(argv)
410 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
411 if idxs:
412 correct_argv = (
413 ['youtube-dl'] +
414 [a for i, a in enumerate(argv) if i not in idxs] +
415 ['--'] + [argv[i] for i in idxs]
416 )
417 self.report_warning(
418 'Long argument string detected. '
419 'Use -- to separate parameters and URLs, like this:\n%s\n' %
420 args_to_str(correct_argv))
421
422 def add_info_extractor(self, ie):
423 """Add an InfoExtractor object to the end of the list."""
424 self._ies.append(ie)
425 if not isinstance(ie, type):
426 self._ies_instances[ie.ie_key()] = ie
427 ie.set_downloader(self)
428
429 def get_info_extractor(self, ie_key):
430 """
431 Get an instance of an IE with name ie_key, it will try to get one from
432 the _ies list, if there's no instance it will create a new one and add
433 it to the extractor list.
434 """
435 ie = self._ies_instances.get(ie_key)
436 if ie is None:
437 ie = get_info_extractor(ie_key)()
438 self.add_info_extractor(ie)
439 return ie
440
441 def add_default_info_extractors(self):
442 """
443 Add the InfoExtractors returned by gen_extractors to the end of the list
444 """
445 for ie in gen_extractor_classes():
446 self.add_info_extractor(ie)
447
448 def add_post_processor(self, pp):
449 """Add a PostProcessor object to the end of the chain."""
450 self._pps.append(pp)
451 pp.set_downloader(self)
452
453 def add_progress_hook(self, ph):
454 """Add the progress hook (currently only for the file downloader)"""
455 self._progress_hooks.append(ph)
456
457 def _bidi_workaround(self, message):
458 if not hasattr(self, '_output_channel'):
459 return message
460
461 assert hasattr(self, '_output_process')
462 assert isinstance(message, compat_str)
463 line_count = message.count('\n') + 1
464 self._output_process.stdin.write((message + '\n').encode('utf-8'))
465 self._output_process.stdin.flush()
466 res = ''.join(self._output_channel.readline().decode('utf-8')
467 for _ in range(line_count))
468 return res[:-len('\n')]
469
470 def to_screen(self, message, skip_eol=False):
471 """Print message to stdout if not in quiet mode."""
472 return self.to_stdout(message, skip_eol, check_quiet=True)
473
474 def _write_string(self, s, out=None):
475 write_string(s, out=out, encoding=self.params.get('encoding'))
476
477 def to_stdout(self, message, skip_eol=False, check_quiet=False):
478 """Print message to stdout if not in quiet mode."""
479 if self.params.get('logger'):
480 self.params['logger'].debug(message)
481 elif not check_quiet or not self.params.get('quiet', False):
482 message = self._bidi_workaround(message)
483 terminator = ['\n', ''][skip_eol]
484 output = message + terminator
485
486 self._write_string(output, self._screen_file)
487
488 def to_stderr(self, message):
489 """Print message to stderr."""
490 assert isinstance(message, compat_str)
491 if self.params.get('logger'):
492 self.params['logger'].error(message)
493 else:
494 message = self._bidi_workaround(message)
495 output = message + '\n'
496 self._write_string(output, self._err_file)
497
498 def to_console_title(self, message):
499 if not self.params.get('consoletitle', False):
500 return
501 if compat_os_name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
502 # c_wchar_p() might not be necessary if `message` is
503 # already of type unicode()
504 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
505 elif 'TERM' in os.environ:
506 self._write_string('\033]0;%s\007' % message, self._screen_file)
507
508 def save_console_title(self):
509 if not self.params.get('consoletitle', False):
510 return
511 if 'TERM' in os.environ:
512 # Save the title on stack
513 self._write_string('\033[22;0t', self._screen_file)
514
515 def restore_console_title(self):
516 if not self.params.get('consoletitle', False):
517 return
518 if 'TERM' in os.environ:
519 # Restore the title from stack
520 self._write_string('\033[23;0t', self._screen_file)
521
522 def __enter__(self):
523 self.save_console_title()
524 return self
525
526 def __exit__(self, *args):
527 self.restore_console_title()
528
529 if self.params.get('cookiefile') is not None:
530 self.cookiejar.save()
531
532 def trouble(self, message=None, tb=None):
533 """Determine action to take when a download problem appears.
534
535 Depending on if the downloader has been configured to ignore
536 download errors or not, this method may throw an exception or
537 not when errors are found, after printing the message.
538
539 tb, if given, is additional traceback information.
540 """
541 if message is not None:
542 self.to_stderr(message)
543 if self.params.get('verbose'):
544 if tb is None:
545 if sys.exc_info()[0]: # if .trouble has been called from an except block
546 tb = ''
547 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
548 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
549 tb += encode_compat_str(traceback.format_exc())
550 else:
551 tb_data = traceback.format_list(traceback.extract_stack())
552 tb = ''.join(tb_data)
553 self.to_stderr(tb)
554 if not self.params.get('ignoreerrors', False):
555 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
556 exc_info = sys.exc_info()[1].exc_info
557 else:
558 exc_info = sys.exc_info()
559 raise DownloadError(message, exc_info)
560 self._download_retcode = 1
561
562 def report_warning(self, message):
563 '''
564 Print the message to stderr, it will be prefixed with 'WARNING:'
565 If stderr is a tty file the 'WARNING:' will be colored
566 '''
567 if self.params.get('logger') is not None:
568 self.params['logger'].warning(message)
569 else:
570 if self.params.get('no_warnings'):
571 return
572 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
573 _msg_header = '\033[0;33mWARNING:\033[0m'
574 else:
575 _msg_header = 'WARNING:'
576 warning_message = '%s %s' % (_msg_header, message)
577 self.to_stderr(warning_message)
578
579 def report_error(self, message, tb=None):
580 '''
581 Do the same as trouble, but prefixes the message with 'ERROR:', colored
582 in red if stderr is a tty file.
583 '''
584 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
585 _msg_header = '\033[0;31mERROR:\033[0m'
586 else:
587 _msg_header = 'ERROR:'
588 error_message = '%s %s' % (_msg_header, message)
589 self.trouble(error_message, tb)
590
591 def report_file_already_downloaded(self, file_name):
592 """Report file has already been fully downloaded."""
593 try:
594 self.to_screen('[download] %s has already been downloaded' % file_name)
595 except UnicodeEncodeError:
596 self.to_screen('[download] The file has already been downloaded')
597
598 def prepare_filename(self, info_dict):
599 """Generate the output filename."""
600 try:
601 template_dict = dict(info_dict)
602
603 template_dict['epoch'] = int(time.time())
604 autonumber_size = self.params.get('autonumber_size')
605 if autonumber_size is None:
606 autonumber_size = 5
607 template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
608 if template_dict.get('resolution') is None:
609 if template_dict.get('width') and template_dict.get('height'):
610 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
611 elif template_dict.get('height'):
612 template_dict['resolution'] = '%sp' % template_dict['height']
613 elif template_dict.get('width'):
614 template_dict['resolution'] = '%dx?' % template_dict['width']
615
616 sanitize = lambda k, v: sanitize_filename(
617 compat_str(v),
618 restricted=self.params.get('restrictfilenames'),
619 is_id=(k == 'id' or k.endswith('_id')))
620 template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
621 for k, v in template_dict.items()
622 if v is not None and not isinstance(v, (list, tuple, dict)))
623 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
624
625 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
626
627 # For fields playlist_index and autonumber convert all occurrences
628 # of %(field)s to %(field)0Nd for backward compatibility
629 field_size_compat_map = {
630 'playlist_index': len(str(template_dict['n_entries'])),
631 'autonumber': autonumber_size,
632 }
633 FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
634 mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
635 if mobj:
636 outtmpl = re.sub(
637 FIELD_SIZE_COMPAT_RE,
638 r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
639 outtmpl)
640
641 NUMERIC_FIELDS = set((
642 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
643 'upload_year', 'upload_month', 'upload_day',
644 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
645 'average_rating', 'comment_count', 'age_limit',
646 'start_time', 'end_time',
647 'chapter_number', 'season_number', 'episode_number',
648 'track_number', 'disc_number', 'release_year',
649 'playlist_index',
650 ))
651
652 # Missing numeric fields used together with integer presentation types
653 # in format specification will break the argument substitution since
654 # string 'NA' is returned for missing fields. We will patch output
655 # template for missing fields to meet string presentation type.
656 for numeric_field in NUMERIC_FIELDS:
657 if numeric_field not in template_dict:
658 # As of [1] format syntax is:
659 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
660 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
661 FORMAT_RE = r'''(?x)
662 (?<!%)
663 %
664 \({0}\) # mapping key
665 (?:[#0\-+ ]+)? # conversion flags (optional)
666 (?:\d+)? # minimum field width (optional)
667 (?:\.\d+)? # precision (optional)
668 [hlL]? # length modifier (optional)
669 [diouxXeEfFgGcrs%] # conversion type
670 '''
671 outtmpl = re.sub(
672 FORMAT_RE.format(numeric_field),
673 r'%({0})s'.format(numeric_field), outtmpl)
674
675 tmpl = expand_path(outtmpl)
676 filename = tmpl % template_dict
677 # Temporary fix for #4787
678 # 'Treat' all problem characters by passing filename through preferredencoding
679 # to workaround encoding issues with subprocess on python2 @ Windows
680 if sys.version_info < (3, 0) and sys.platform == 'win32':
681 filename = encodeFilename(filename, True).decode(preferredencoding())
682 return sanitize_path(filename)
683 except ValueError as err:
684 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
685 return None
686
687 def _match_entry(self, info_dict, incomplete):
688 """ Returns None iff the file should be downloaded """
689
690 video_title = info_dict.get('title', info_dict.get('id', 'video'))
691 if 'title' in info_dict:
692 # This can happen when we're just evaluating the playlist
693 title = info_dict['title']
694 matchtitle = self.params.get('matchtitle', False)
695 if matchtitle:
696 if not re.search(matchtitle, title, re.IGNORECASE):
697 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
698 rejecttitle = self.params.get('rejecttitle', False)
699 if rejecttitle:
700 if re.search(rejecttitle, title, re.IGNORECASE):
701 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
702 date = info_dict.get('upload_date')
703 if date is not None:
704 dateRange = self.params.get('daterange', DateRange())
705 if date not in dateRange:
706 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
707 view_count = info_dict.get('view_count')
708 if view_count is not None:
709 min_views = self.params.get('min_views')
710 if min_views is not None and view_count < min_views:
711 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
712 max_views = self.params.get('max_views')
713 if max_views is not None and view_count > max_views:
714 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
715 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
716 return 'Skipping "%s" because it is age restricted' % video_title
717 if self.in_download_archive(info_dict):
718 return '%s has already been recorded in archive' % video_title
719
720 if not incomplete:
721 match_filter = self.params.get('match_filter')
722 if match_filter is not None:
723 ret = match_filter(info_dict)
724 if ret is not None:
725 return ret
726
727 return None
728
729 @staticmethod
730 def add_extra_info(info_dict, extra_info):
731 '''Set the keys from extra_info in info dict if they are missing'''
732 for key, value in extra_info.items():
733 info_dict.setdefault(key, value)
734
735 def extract_info(self, url, download=True, ie_key=None, extra_info={},
736 process=True, force_generic_extractor=False):
737 '''
738 Returns a list with a dictionary for each video we find.
739 If 'download', also downloads the videos.
740 extra_info is a dict containing the extra values to add to each result
741 '''
742
743 if not ie_key and force_generic_extractor:
744 ie_key = 'Generic'
745
746 if ie_key:
747 ies = [self.get_info_extractor(ie_key)]
748 else:
749 ies = self._ies
750
751 for ie in ies:
752 if not ie.suitable(url):
753 continue
754
755 ie = self.get_info_extractor(ie.ie_key())
756 if not ie.working():
757 self.report_warning('The program functionality for this site has been marked as broken, '
758 'and will probably not work.')
759
760 try:
761 ie_result = ie.extract(url)
762 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
763 break
764 if isinstance(ie_result, list):
765 # Backwards compatibility: old IE result format
766 ie_result = {
767 '_type': 'compat_list',
768 'entries': ie_result,
769 }
770 self.add_default_extra_info(ie_result, ie, url)
771 if process:
772 return self.process_ie_result(ie_result, download, extra_info)
773 else:
774 return ie_result
775 except GeoRestrictedError as e:
776 msg = e.msg
777 if e.countries:
778 msg += '\nThis video is available in %s.' % ', '.join(
779 map(ISO3166Utils.short2full, e.countries))
780 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
781 self.report_error(msg)
782 break
783 except ExtractorError as e: # An error we somewhat expected
784 self.report_error(compat_str(e), e.format_traceback())
785 break
786 except MaxDownloadsReached:
787 raise
788 except Exception as e:
789 if self.params.get('ignoreerrors', False):
790 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
791 break
792 else:
793 raise
794 else:
795 self.report_error('no suitable InfoExtractor for URL %s' % url)
796
797 def add_default_extra_info(self, ie_result, ie, url):
798 self.add_extra_info(ie_result, {
799 'extractor': ie.IE_NAME,
800 'webpage_url': url,
801 'webpage_url_basename': url_basename(url),
802 'extractor_key': ie.ie_key(),
803 })
804
805 def process_ie_result(self, ie_result, download=True, extra_info={}):
806 """
807 Take the result of the ie(may be modified) and resolve all unresolved
808 references (URLs, playlist items).
809
810 It will also download the videos if 'download'.
811 Returns the resolved ie_result.
812 """
813 result_type = ie_result.get('_type', 'video')
814
815 if result_type in ('url', 'url_transparent'):
816 ie_result['url'] = sanitize_url(ie_result['url'])
817 extract_flat = self.params.get('extract_flat', False)
818 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
819 extract_flat is True):
820 if self.params.get('forcejson', False):
821 self.to_stdout(json.dumps(ie_result))
822 return ie_result
823
824 if result_type == 'video':
825 self.add_extra_info(ie_result, extra_info)
826 return self.process_video_result(ie_result, download=download)
827 elif result_type == 'url':
828 # We have to add extra_info to the results because it may be
829 # contained in a playlist
830 return self.extract_info(ie_result['url'],
831 download,
832 ie_key=ie_result.get('ie_key'),
833 extra_info=extra_info)
834 elif result_type == 'url_transparent':
835 # Use the information from the embedding page
836 info = self.extract_info(
837 ie_result['url'], ie_key=ie_result.get('ie_key'),
838 extra_info=extra_info, download=False, process=False)
839
840 # extract_info may return None when ignoreerrors is enabled and
841 # extraction failed with an error, don't crash and return early
842 # in this case
843 if not info:
844 return info
845
846 force_properties = dict(
847 (k, v) for k, v in ie_result.items() if v is not None)
848 for f in ('_type', 'url', 'ie_key'):
849 if f in force_properties:
850 del force_properties[f]
851 new_result = info.copy()
852 new_result.update(force_properties)
853
854 # Extracted info may not be a video result (i.e.
855 # info.get('_type', 'video') != video) but rather an url or
856 # url_transparent. In such cases outer metadata (from ie_result)
857 # should be propagated to inner one (info). For this to happen
858 # _type of info should be overridden with url_transparent. This
859 # fixes issue from https://github.com/rg3/youtube-dl/pull/11163.
860 if new_result.get('_type') == 'url':
861 new_result['_type'] = 'url_transparent'
862
863 return self.process_ie_result(
864 new_result, download=download, extra_info=extra_info)
865 elif result_type in ('playlist', 'multi_video'):
866 # We process each entry in the playlist
867 playlist = ie_result.get('title') or ie_result.get('id')
868 self.to_screen('[download] Downloading playlist: %s' % playlist)
869
870 playlist_results = []
871
872 playliststart = self.params.get('playliststart', 1) - 1
873 playlistend = self.params.get('playlistend')
874 # For backwards compatibility, interpret -1 as whole list
875 if playlistend == -1:
876 playlistend = None
877
878 playlistitems_str = self.params.get('playlist_items')
879 playlistitems = None
880 if playlistitems_str is not None:
881 def iter_playlistitems(format):
882 for string_segment in format.split(','):
883 if '-' in string_segment:
884 start, end = string_segment.split('-')
885 for item in range(int(start), int(end) + 1):
886 yield int(item)
887 else:
888 yield int(string_segment)
889 playlistitems = iter_playlistitems(playlistitems_str)
890
891 ie_entries = ie_result['entries']
892 if isinstance(ie_entries, list):
893 n_all_entries = len(ie_entries)
894 if playlistitems:
895 entries = [
896 ie_entries[i - 1] for i in playlistitems
897 if -n_all_entries <= i - 1 < n_all_entries]
898 else:
899 entries = ie_entries[playliststart:playlistend]
900 n_entries = len(entries)
901 self.to_screen(
902 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
903 (ie_result['extractor'], playlist, n_all_entries, n_entries))
904 elif isinstance(ie_entries, PagedList):
905 if playlistitems:
906 entries = []
907 for item in playlistitems:
908 entries.extend(ie_entries.getslice(
909 item - 1, item
910 ))
911 else:
912 entries = ie_entries.getslice(
913 playliststart, playlistend)
914 n_entries = len(entries)
915 self.to_screen(
916 '[%s] playlist %s: Downloading %d videos' %
917 (ie_result['extractor'], playlist, n_entries))
918 else: # iterable
919 if playlistitems:
920 entry_list = list(ie_entries)
921 entries = [entry_list[i - 1] for i in playlistitems]
922 else:
923 entries = list(itertools.islice(
924 ie_entries, playliststart, playlistend))
925 n_entries = len(entries)
926 self.to_screen(
927 '[%s] playlist %s: Downloading %d videos' %
928 (ie_result['extractor'], playlist, n_entries))
929
930 if self.params.get('playlistreverse', False):
931 entries = entries[::-1]
932
933 if self.params.get('playlistrandom', False):
934 random.shuffle(entries)
935
936 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
937
938 for i, entry in enumerate(entries, 1):
939 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
940 # This __x_forwarded_for_ip thing is a bit ugly but requires
941 # minimal changes
942 if x_forwarded_for:
943 entry['__x_forwarded_for_ip'] = x_forwarded_for
944 extra = {
945 'n_entries': n_entries,
946 'playlist': playlist,
947 'playlist_id': ie_result.get('id'),
948 'playlist_title': ie_result.get('title'),
949 'playlist_index': i + playliststart,
950 'extractor': ie_result['extractor'],
951 'webpage_url': ie_result['webpage_url'],
952 'webpage_url_basename': url_basename(ie_result['webpage_url']),
953 'extractor_key': ie_result['extractor_key'],
954 }
955
956 reason = self._match_entry(entry, incomplete=True)
957 if reason is not None:
958 self.to_screen('[download] ' + reason)
959 continue
960
961 entry_result = self.process_ie_result(entry,
962 download=download,
963 extra_info=extra)
964 playlist_results.append(entry_result)
965 ie_result['entries'] = playlist_results
966 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
967 return ie_result
968 elif result_type == 'compat_list':
969 self.report_warning(
970 'Extractor %s returned a compat_list result. '
971 'It needs to be updated.' % ie_result.get('extractor'))
972
973 def _fixup(r):
974 self.add_extra_info(
975 r,
976 {
977 'extractor': ie_result['extractor'],
978 'webpage_url': ie_result['webpage_url'],
979 'webpage_url_basename': url_basename(ie_result['webpage_url']),
980 'extractor_key': ie_result['extractor_key'],
981 }
982 )
983 return r
984 ie_result['entries'] = [
985 self.process_ie_result(_fixup(r), download, extra_info)
986 for r in ie_result['entries']
987 ]
988 return ie_result
989 else:
990 raise Exception('Invalid result type: %s' % result_type)
991
992 def _build_format_filter(self, filter_spec):
993 " Returns a function to filter the formats according to the filter_spec "
994
995 OPERATORS = {
996 '<': operator.lt,
997 '<=': operator.le,
998 '>': operator.gt,
999 '>=': operator.ge,
1000 '=': operator.eq,
1001 '!=': operator.ne,
1002 }
1003 operator_rex = re.compile(r'''(?x)\s*
1004 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
1005 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1006 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
1007 $
1008 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1009 m = operator_rex.search(filter_spec)
1010 if m:
1011 try:
1012 comparison_value = int(m.group('value'))
1013 except ValueError:
1014 comparison_value = parse_filesize(m.group('value'))
1015 if comparison_value is None:
1016 comparison_value = parse_filesize(m.group('value') + 'B')
1017 if comparison_value is None:
1018 raise ValueError(
1019 'Invalid value %r in format specification %r' % (
1020 m.group('value'), filter_spec))
1021 op = OPERATORS[m.group('op')]
1022
1023 if not m:
1024 STR_OPERATORS = {
1025 '=': operator.eq,
1026 '!=': operator.ne,
1027 '^=': lambda attr, value: attr.startswith(value),
1028 '$=': lambda attr, value: attr.endswith(value),
1029 '*=': lambda attr, value: value in attr,
1030 }
1031 str_operator_rex = re.compile(r'''(?x)
1032 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
1033 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
1034 \s*(?P<value>[a-zA-Z0-9._-]+)
1035 \s*$
1036 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1037 m = str_operator_rex.search(filter_spec)
1038 if m:
1039 comparison_value = m.group('value')
1040 op = STR_OPERATORS[m.group('op')]
1041
1042 if not m:
1043 raise ValueError('Invalid filter specification %r' % filter_spec)
1044
1045 def _filter(f):
1046 actual_value = f.get(m.group('key'))
1047 if actual_value is None:
1048 return m.group('none_inclusive')
1049 return op(actual_value, comparison_value)
1050 return _filter
1051
1052 def build_format_selector(self, format_spec):
1053 def syntax_error(note, start):
1054 message = (
1055 'Invalid format specification: '
1056 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1057 return SyntaxError(message)
1058
1059 PICKFIRST = 'PICKFIRST'
1060 MERGE = 'MERGE'
1061 SINGLE = 'SINGLE'
1062 GROUP = 'GROUP'
1063 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1064
1065 def _parse_filter(tokens):
1066 filter_parts = []
1067 for type, string, start, _, _ in tokens:
1068 if type == tokenize.OP and string == ']':
1069 return ''.join(filter_parts)
1070 else:
1071 filter_parts.append(string)
1072
1073 def _remove_unused_ops(tokens):
1074 # Remove operators that we don't use and join them with the surrounding strings
1075 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1076 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1077 last_string, last_start, last_end, last_line = None, None, None, None
1078 for type, string, start, end, line in tokens:
1079 if type == tokenize.OP and string == '[':
1080 if last_string:
1081 yield tokenize.NAME, last_string, last_start, last_end, last_line
1082 last_string = None
1083 yield type, string, start, end, line
1084 # everything inside brackets will be handled by _parse_filter
1085 for type, string, start, end, line in tokens:
1086 yield type, string, start, end, line
1087 if type == tokenize.OP and string == ']':
1088 break
1089 elif type == tokenize.OP and string in ALLOWED_OPS:
1090 if last_string:
1091 yield tokenize.NAME, last_string, last_start, last_end, last_line
1092 last_string = None
1093 yield type, string, start, end, line
1094 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1095 if not last_string:
1096 last_string = string
1097 last_start = start
1098 last_end = end
1099 else:
1100 last_string += string
1101 if last_string:
1102 yield tokenize.NAME, last_string, last_start, last_end, last_line
1103
1104 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1105 selectors = []
1106 current_selector = None
1107 for type, string, start, _, _ in tokens:
1108 # ENCODING is only defined in python 3.x
1109 if type == getattr(tokenize, 'ENCODING', None):
1110 continue
1111 elif type in [tokenize.NAME, tokenize.NUMBER]:
1112 current_selector = FormatSelector(SINGLE, string, [])
1113 elif type == tokenize.OP:
1114 if string == ')':
1115 if not inside_group:
1116 # ')' will be handled by the parentheses group
1117 tokens.restore_last_token()
1118 break
1119 elif inside_merge and string in ['/', ',']:
1120 tokens.restore_last_token()
1121 break
1122 elif inside_choice and string == ',':
1123 tokens.restore_last_token()
1124 break
1125 elif string == ',':
1126 if not current_selector:
1127 raise syntax_error('"," must follow a format selector', start)
1128 selectors.append(current_selector)
1129 current_selector = None
1130 elif string == '/':
1131 if not current_selector:
1132 raise syntax_error('"/" must follow a format selector', start)
1133 first_choice = current_selector
1134 second_choice = _parse_format_selection(tokens, inside_choice=True)
1135 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1136 elif string == '[':
1137 if not current_selector:
1138 current_selector = FormatSelector(SINGLE, 'best', [])
1139 format_filter = _parse_filter(tokens)
1140 current_selector.filters.append(format_filter)
1141 elif string == '(':
1142 if current_selector:
1143 raise syntax_error('Unexpected "("', start)
1144 group = _parse_format_selection(tokens, inside_group=True)
1145 current_selector = FormatSelector(GROUP, group, [])
1146 elif string == '+':
1147 video_selector = current_selector
1148 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1149 if not video_selector or not audio_selector:
1150 raise syntax_error('"+" must be between two format selectors', start)
1151 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1152 else:
1153 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1154 elif type == tokenize.ENDMARKER:
1155 break
1156 if current_selector:
1157 selectors.append(current_selector)
1158 return selectors
1159
1160 def _build_selector_function(selector):
1161 if isinstance(selector, list):
1162 fs = [_build_selector_function(s) for s in selector]
1163
1164 def selector_function(ctx):
1165 for f in fs:
1166 for format in f(ctx):
1167 yield format
1168 return selector_function
1169 elif selector.type == GROUP:
1170 selector_function = _build_selector_function(selector.selector)
1171 elif selector.type == PICKFIRST:
1172 fs = [_build_selector_function(s) for s in selector.selector]
1173
1174 def selector_function(ctx):
1175 for f in fs:
1176 picked_formats = list(f(ctx))
1177 if picked_formats:
1178 return picked_formats
1179 return []
1180 elif selector.type == SINGLE:
1181 format_spec = selector.selector
1182
1183 def selector_function(ctx):
1184 formats = list(ctx['formats'])
1185 if not formats:
1186 return
1187 if format_spec == 'all':
1188 for f in formats:
1189 yield f
1190 elif format_spec in ['best', 'worst', None]:
1191 format_idx = 0 if format_spec == 'worst' else -1
1192 audiovideo_formats = [
1193 f for f in formats
1194 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1195 if audiovideo_formats:
1196 yield audiovideo_formats[format_idx]
1197 # for extractors with incomplete formats (audio only (soundcloud)
1198 # or video only (imgur)) we will fallback to best/worst
1199 # {video,audio}-only format
1200 elif ctx['incomplete_formats']:
1201 yield formats[format_idx]
1202 elif format_spec == 'bestaudio':
1203 audio_formats = [
1204 f for f in formats
1205 if f.get('vcodec') == 'none']
1206 if audio_formats:
1207 yield audio_formats[-1]
1208 elif format_spec == 'worstaudio':
1209 audio_formats = [
1210 f for f in formats
1211 if f.get('vcodec') == 'none']
1212 if audio_formats:
1213 yield audio_formats[0]
1214 elif format_spec == 'bestvideo':
1215 video_formats = [
1216 f for f in formats
1217 if f.get('acodec') == 'none']
1218 if video_formats:
1219 yield video_formats[-1]
1220 elif format_spec == 'worstvideo':
1221 video_formats = [
1222 f for f in formats
1223 if f.get('acodec') == 'none']
1224 if video_formats:
1225 yield video_formats[0]
1226 else:
1227 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1228 if format_spec in extensions:
1229 filter_f = lambda f: f['ext'] == format_spec
1230 else:
1231 filter_f = lambda f: f['format_id'] == format_spec
1232 matches = list(filter(filter_f, formats))
1233 if matches:
1234 yield matches[-1]
1235 elif selector.type == MERGE:
1236 def _merge(formats_info):
1237 format_1, format_2 = [f['format_id'] for f in formats_info]
1238 # The first format must contain the video and the
1239 # second the audio
1240 if formats_info[0].get('vcodec') == 'none':
1241 self.report_error('The first format must '
1242 'contain the video, try using '
1243 '"-f %s+%s"' % (format_2, format_1))
1244 return
1245 # Formats must be opposite (video+audio)
1246 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1247 self.report_error(
1248 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1249 % (format_1, format_2))
1250 return
1251 output_ext = (
1252 formats_info[0]['ext']
1253 if self.params.get('merge_output_format') is None
1254 else self.params['merge_output_format'])
1255 return {
1256 'requested_formats': formats_info,
1257 'format': '%s+%s' % (formats_info[0].get('format'),
1258 formats_info[1].get('format')),
1259 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1260 formats_info[1].get('format_id')),
1261 'width': formats_info[0].get('width'),
1262 'height': formats_info[0].get('height'),
1263 'resolution': formats_info[0].get('resolution'),
1264 'fps': formats_info[0].get('fps'),
1265 'vcodec': formats_info[0].get('vcodec'),
1266 'vbr': formats_info[0].get('vbr'),
1267 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1268 'acodec': formats_info[1].get('acodec'),
1269 'abr': formats_info[1].get('abr'),
1270 'ext': output_ext,
1271 }
1272 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1273
1274 def selector_function(ctx):
1275 for pair in itertools.product(
1276 video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
1277 yield _merge(pair)
1278
1279 filters = [self._build_format_filter(f) for f in selector.filters]
1280
1281 def final_selector(ctx):
1282 ctx_copy = copy.deepcopy(ctx)
1283 for _filter in filters:
1284 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1285 return selector_function(ctx_copy)
1286 return final_selector
1287
1288 stream = io.BytesIO(format_spec.encode('utf-8'))
1289 try:
1290 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1291 except tokenize.TokenError:
1292 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1293
1294 class TokenIterator(object):
1295 def __init__(self, tokens):
1296 self.tokens = tokens
1297 self.counter = 0
1298
1299 def __iter__(self):
1300 return self
1301
1302 def __next__(self):
1303 if self.counter >= len(self.tokens):
1304 raise StopIteration()
1305 value = self.tokens[self.counter]
1306 self.counter += 1
1307 return value
1308
1309 next = __next__
1310
1311 def restore_last_token(self):
1312 self.counter -= 1
1313
1314 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1315 return _build_selector_function(parsed_selector)
1316
1317 def _calc_headers(self, info_dict):
1318 res = std_headers.copy()
1319
1320 add_headers = info_dict.get('http_headers')
1321 if add_headers:
1322 res.update(add_headers)
1323
1324 cookies = self._calc_cookies(info_dict)
1325 if cookies:
1326 res['Cookie'] = cookies
1327
1328 if 'X-Forwarded-For' not in res:
1329 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1330 if x_forwarded_for_ip:
1331 res['X-Forwarded-For'] = x_forwarded_for_ip
1332
1333 return res
1334
1335 def _calc_cookies(self, info_dict):
1336 pr = sanitized_Request(info_dict['url'])
1337 self.cookiejar.add_cookie_header(pr)
1338 return pr.get_header('Cookie')
1339
1340 def process_video_result(self, info_dict, download=True):
1341 assert info_dict.get('_type', 'video') == 'video'
1342
1343 if 'id' not in info_dict:
1344 raise ExtractorError('Missing "id" field in extractor result')
1345 if 'title' not in info_dict:
1346 raise ExtractorError('Missing "title" field in extractor result')
1347
1348 if not isinstance(info_dict['id'], compat_str):
1349 self.report_warning('"id" field is not a string - forcing string conversion')
1350 info_dict['id'] = compat_str(info_dict['id'])
1351
1352 if 'playlist' not in info_dict:
1353 # It isn't part of a playlist
1354 info_dict['playlist'] = None
1355 info_dict['playlist_index'] = None
1356
1357 thumbnails = info_dict.get('thumbnails')
1358 if thumbnails is None:
1359 thumbnail = info_dict.get('thumbnail')
1360 if thumbnail:
1361 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1362 if thumbnails:
1363 thumbnails.sort(key=lambda t: (
1364 t.get('preference') if t.get('preference') is not None else -1,
1365 t.get('width') if t.get('width') is not None else -1,
1366 t.get('height') if t.get('height') is not None else -1,
1367 t.get('id') if t.get('id') is not None else '', t.get('url')))
1368 for i, t in enumerate(thumbnails):
1369 t['url'] = sanitize_url(t['url'])
1370 if t.get('width') and t.get('height'):
1371 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1372 if t.get('id') is None:
1373 t['id'] = '%d' % i
1374
1375 if self.params.get('list_thumbnails'):
1376 self.list_thumbnails(info_dict)
1377 return
1378
1379 thumbnail = info_dict.get('thumbnail')
1380 if thumbnail:
1381 info_dict['thumbnail'] = sanitize_url(thumbnail)
1382 elif thumbnails:
1383 info_dict['thumbnail'] = thumbnails[-1]['url']
1384
1385 if 'display_id' not in info_dict and 'id' in info_dict:
1386 info_dict['display_id'] = info_dict['id']
1387
1388 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1389 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1390 # see http://bugs.python.org/issue1646728)
1391 try:
1392 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1393 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1394 except (ValueError, OverflowError, OSError):
1395 pass
1396
1397 # Auto generate title fields corresponding to the *_number fields when missing
1398 # in order to always have clean titles. This is very common for TV series.
1399 for field in ('chapter', 'season', 'episode'):
1400 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1401 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1402
1403 subtitles = info_dict.get('subtitles')
1404 if subtitles:
1405 for _, subtitle in subtitles.items():
1406 for subtitle_format in subtitle:
1407 if subtitle_format.get('url'):
1408 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1409 if subtitle_format.get('ext') is None:
1410 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1411
1412 if self.params.get('listsubtitles', False):
1413 if 'automatic_captions' in info_dict:
1414 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1415 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1416 return
1417 info_dict['requested_subtitles'] = self.process_subtitles(
1418 info_dict['id'], subtitles,
1419 info_dict.get('automatic_captions'))
1420
1421 # We now pick which formats have to be downloaded
1422 if info_dict.get('formats') is None:
1423 # There's only one format available
1424 formats = [info_dict]
1425 else:
1426 formats = info_dict['formats']
1427
1428 if not formats:
1429 raise ExtractorError('No video formats found!')
1430
1431 formats_dict = {}
1432
1433 # We check that all the formats have the format and format_id fields
1434 for i, format in enumerate(formats):
1435 if 'url' not in format:
1436 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1437
1438 format['url'] = sanitize_url(format['url'])
1439
1440 if format.get('format_id') is None:
1441 format['format_id'] = compat_str(i)
1442 else:
1443 # Sanitize format_id from characters used in format selector expression
1444 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
1445 format_id = format['format_id']
1446 if format_id not in formats_dict:
1447 formats_dict[format_id] = []
1448 formats_dict[format_id].append(format)
1449
1450 # Make sure all formats have unique format_id
1451 for format_id, ambiguous_formats in formats_dict.items():
1452 if len(ambiguous_formats) > 1:
1453 for i, format in enumerate(ambiguous_formats):
1454 format['format_id'] = '%s-%d' % (format_id, i)
1455
1456 for i, format in enumerate(formats):
1457 if format.get('format') is None:
1458 format['format'] = '{id} - {res}{note}'.format(
1459 id=format['format_id'],
1460 res=self.format_resolution(format),
1461 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1462 )
1463 # Automatically determine file extension if missing
1464 if format.get('ext') is None:
1465 format['ext'] = determine_ext(format['url']).lower()
1466 # Automatically determine protocol if missing (useful for format
1467 # selection purposes)
1468 if format.get('protocol') is None:
1469 format['protocol'] = determine_protocol(format)
1470 # Add HTTP headers, so that external programs can use them from the
1471 # json output
1472 full_format_info = info_dict.copy()
1473 full_format_info.update(format)
1474 format['http_headers'] = self._calc_headers(full_format_info)
1475 # Remove private housekeeping stuff
1476 if '__x_forwarded_for_ip' in info_dict:
1477 del info_dict['__x_forwarded_for_ip']
1478
1479 # TODO Central sorting goes here
1480
1481 if formats[0] is not info_dict:
1482 # only set the 'formats' fields if the original info_dict list them
1483 # otherwise we end up with a circular reference, the first (and unique)
1484 # element in the 'formats' field in info_dict is info_dict itself,
1485 # which can't be exported to json
1486 info_dict['formats'] = formats
1487 if self.params.get('listformats'):
1488 self.list_formats(info_dict)
1489 return
1490
1491 req_format = self.params.get('format')
1492 if req_format is None:
1493 req_format_list = []
1494 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1495 not info_dict.get('is_live')):
1496 merger = FFmpegMergerPP(self)
1497 if merger.available and merger.can_merge():
1498 req_format_list.append('bestvideo+bestaudio')
1499 req_format_list.append('best')
1500 req_format = '/'.join(req_format_list)
1501 format_selector = self.build_format_selector(req_format)
1502
1503 # While in format selection we may need to have an access to the original
1504 # format set in order to calculate some metrics or do some processing.
1505 # For now we need to be able to guess whether original formats provided
1506 # by extractor are incomplete or not (i.e. whether extractor provides only
1507 # video-only or audio-only formats) for proper formats selection for
1508 # extractors with such incomplete formats (see
1509 # https://github.com/rg3/youtube-dl/pull/5556).
1510 # Since formats may be filtered during format selection and may not match
1511 # the original formats the results may be incorrect. Thus original formats
1512 # or pre-calculated metrics should be passed to format selection routines
1513 # as well.
1514 # We will pass a context object containing all necessary additional data
1515 # instead of just formats.
1516 # This fixes incorrect format selection issue (see
1517 # https://github.com/rg3/youtube-dl/issues/10083).
1518 incomplete_formats = (
1519 # All formats are video-only or
1520 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) or
1521 # all formats are audio-only
1522 all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1523
1524 ctx = {
1525 'formats': formats,
1526 'incomplete_formats': incomplete_formats,
1527 }
1528
1529 formats_to_download = list(format_selector(ctx))
1530 if not formats_to_download:
1531 raise ExtractorError('requested format not available',
1532 expected=True)
1533
1534 if download:
1535 if len(formats_to_download) > 1:
1536 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1537 for format in formats_to_download:
1538 new_info = dict(info_dict)
1539 new_info.update(format)
1540 self.process_info(new_info)
1541 # We update the info dict with the best quality format (backwards compatibility)
1542 info_dict.update(formats_to_download[-1])
1543 return info_dict
1544
1545 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1546 """Select the requested subtitles and their format"""
1547 available_subs = {}
1548 if normal_subtitles and self.params.get('writesubtitles'):
1549 available_subs.update(normal_subtitles)
1550 if automatic_captions and self.params.get('writeautomaticsub'):
1551 for lang, cap_info in automatic_captions.items():
1552 if lang not in available_subs:
1553 available_subs[lang] = cap_info
1554
1555 if (not self.params.get('writesubtitles') and not
1556 self.params.get('writeautomaticsub') or not
1557 available_subs):
1558 return None
1559
1560 if self.params.get('allsubtitles', False):
1561 requested_langs = available_subs.keys()
1562 else:
1563 if self.params.get('subtitleslangs', False):
1564 requested_langs = self.params.get('subtitleslangs')
1565 elif 'en' in available_subs:
1566 requested_langs = ['en']
1567 else:
1568 requested_langs = [list(available_subs.keys())[0]]
1569
1570 formats_query = self.params.get('subtitlesformat', 'best')
1571 formats_preference = formats_query.split('/') if formats_query else []
1572 subs = {}
1573 for lang in requested_langs:
1574 formats = available_subs.get(lang)
1575 if formats is None:
1576 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1577 continue
1578 for ext in formats_preference:
1579 if ext == 'best':
1580 f = formats[-1]
1581 break
1582 matches = list(filter(lambda f: f['ext'] == ext, formats))
1583 if matches:
1584 f = matches[-1]
1585 break
1586 else:
1587 f = formats[-1]
1588 self.report_warning(
1589 'No subtitle format found matching "%s" for language %s, '
1590 'using %s' % (formats_query, lang, f['ext']))
1591 subs[lang] = f
1592 return subs
1593
1594 def process_info(self, info_dict):
1595 """Process a single resolved IE result."""
1596
1597 assert info_dict.get('_type', 'video') == 'video'
1598
1599 max_downloads = self.params.get('max_downloads')
1600 if max_downloads is not None:
1601 if self._num_downloads >= int(max_downloads):
1602 raise MaxDownloadsReached()
1603
1604 info_dict['fulltitle'] = info_dict['title']
1605 if len(info_dict['title']) > 200:
1606 info_dict['title'] = info_dict['title'][:197] + '...'
1607
1608 if 'format' not in info_dict:
1609 info_dict['format'] = info_dict['ext']
1610
1611 reason = self._match_entry(info_dict, incomplete=False)
1612 if reason is not None:
1613 self.to_screen('[download] ' + reason)
1614 return
1615
1616 self._num_downloads += 1
1617
1618 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1619
1620 # Forced printings
1621 if self.params.get('forcetitle', False):
1622 self.to_stdout(info_dict['fulltitle'])
1623 if self.params.get('forceid', False):
1624 self.to_stdout(info_dict['id'])
1625 if self.params.get('forceurl', False):
1626 if info_dict.get('requested_formats') is not None:
1627 for f in info_dict['requested_formats']:
1628 self.to_stdout(f['url'] + f.get('play_path', ''))
1629 else:
1630 # For RTMP URLs, also include the playpath
1631 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1632 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1633 self.to_stdout(info_dict['thumbnail'])
1634 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1635 self.to_stdout(info_dict['description'])
1636 if self.params.get('forcefilename', False) and filename is not None:
1637 self.to_stdout(filename)
1638 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1639 self.to_stdout(formatSeconds(info_dict['duration']))
1640 if self.params.get('forceformat', False):
1641 self.to_stdout(info_dict['format'])
1642 if self.params.get('forcejson', False):
1643 self.to_stdout(json.dumps(info_dict))
1644
1645 # Do nothing else if in simulate mode
1646 if self.params.get('simulate', False):
1647 return
1648
1649 if filename is None:
1650 return
1651
1652 try:
1653 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1654 if dn and not os.path.exists(dn):
1655 os.makedirs(dn)
1656 except (OSError, IOError) as err:
1657 self.report_error('unable to create directory ' + error_to_compat_str(err))
1658 return
1659
1660 if self.params.get('writedescription', False):
1661 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1662 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1663 self.to_screen('[info] Video description is already present')
1664 elif info_dict.get('description') is None:
1665 self.report_warning('There\'s no description to write.')
1666 else:
1667 try:
1668 self.to_screen('[info] Writing video description to: ' + descfn)
1669 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1670 descfile.write(info_dict['description'])
1671 except (OSError, IOError):
1672 self.report_error('Cannot write description file ' + descfn)
1673 return
1674
1675 if self.params.get('writeannotations', False):
1676 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1677 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1678 self.to_screen('[info] Video annotations are already present')
1679 else:
1680 try:
1681 self.to_screen('[info] Writing video annotations to: ' + annofn)
1682 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1683 annofile.write(info_dict['annotations'])
1684 except (KeyError, TypeError):
1685 self.report_warning('There are no annotations to write.')
1686 except (OSError, IOError):
1687 self.report_error('Cannot write annotations file: ' + annofn)
1688 return
1689
1690 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1691 self.params.get('writeautomaticsub')])
1692
1693 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1694 # subtitles download errors are already managed as troubles in relevant IE
1695 # that way it will silently go on when used with unsupporting IE
1696 subtitles = info_dict['requested_subtitles']
1697 ie = self.get_info_extractor(info_dict['extractor_key'])
1698 for sub_lang, sub_info in subtitles.items():
1699 sub_format = sub_info['ext']
1700 if sub_info.get('data') is not None:
1701 sub_data = sub_info['data']
1702 else:
1703 try:
1704 sub_data = ie._download_webpage(
1705 sub_info['url'], info_dict['id'], note=False)
1706 except ExtractorError as err:
1707 self.report_warning('Unable to download subtitle for "%s": %s' %
1708 (sub_lang, error_to_compat_str(err.cause)))
1709 continue
1710 try:
1711 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1712 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1713 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1714 else:
1715 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1716 # Use newline='' to prevent conversion of newline characters
1717 # See https://github.com/rg3/youtube-dl/issues/10268
1718 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
1719 subfile.write(sub_data)
1720 except (OSError, IOError):
1721 self.report_error('Cannot write subtitles file ' + sub_filename)
1722 return
1723
1724 if self.params.get('writeinfojson', False):
1725 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1726 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1727 self.to_screen('[info] Video description metadata is already present')
1728 else:
1729 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1730 try:
1731 write_json_file(self.filter_requested_info(info_dict), infofn)
1732 except (OSError, IOError):
1733 self.report_error('Cannot write metadata to JSON file ' + infofn)
1734 return
1735
1736 self._write_thumbnails(info_dict, filename)
1737
1738 if not self.params.get('skip_download', False):
1739 try:
1740 def dl(name, info):
1741 fd = get_suitable_downloader(info, self.params)(self, self.params)
1742 for ph in self._progress_hooks:
1743 fd.add_progress_hook(ph)
1744 if self.params.get('verbose'):
1745 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1746 return fd.download(name, info)
1747
1748 if info_dict.get('requested_formats') is not None:
1749 downloaded = []
1750 success = True
1751 merger = FFmpegMergerPP(self)
1752 if not merger.available:
1753 postprocessors = []
1754 self.report_warning('You have requested multiple '
1755 'formats but ffmpeg or avconv are not installed.'
1756 ' The formats won\'t be merged.')
1757 else:
1758 postprocessors = [merger]
1759
1760 def compatible_formats(formats):
1761 video, audio = formats
1762 # Check extension
1763 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1764 if video_ext and audio_ext:
1765 COMPATIBLE_EXTS = (
1766 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'),
1767 ('webm')
1768 )
1769 for exts in COMPATIBLE_EXTS:
1770 if video_ext in exts and audio_ext in exts:
1771 return True
1772 # TODO: Check acodec/vcodec
1773 return False
1774
1775 filename_real_ext = os.path.splitext(filename)[1][1:]
1776 filename_wo_ext = (
1777 os.path.splitext(filename)[0]
1778 if filename_real_ext == info_dict['ext']
1779 else filename)
1780 requested_formats = info_dict['requested_formats']
1781 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1782 info_dict['ext'] = 'mkv'
1783 self.report_warning(
1784 'Requested formats are incompatible for merge and will be merged into mkv.')
1785 # Ensure filename always has a correct extension for successful merge
1786 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1787 if os.path.exists(encodeFilename(filename)):
1788 self.to_screen(
1789 '[download] %s has already been downloaded and '
1790 'merged' % filename)
1791 else:
1792 for f in requested_formats:
1793 new_info = dict(info_dict)
1794 new_info.update(f)
1795 fname = self.prepare_filename(new_info)
1796 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1797 downloaded.append(fname)
1798 partial_success = dl(fname, new_info)
1799 success = success and partial_success
1800 info_dict['__postprocessors'] = postprocessors
1801 info_dict['__files_to_merge'] = downloaded
1802 else:
1803 # Just a single file
1804 success = dl(filename, info_dict)
1805 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1806 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1807 return
1808 except (OSError, IOError) as err:
1809 raise UnavailableVideoError(err)
1810 except (ContentTooShortError, ) as err:
1811 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1812 return
1813
1814 if success and filename != '-':
1815 # Fixup content
1816 fixup_policy = self.params.get('fixup')
1817 if fixup_policy is None:
1818 fixup_policy = 'detect_or_warn'
1819
1820 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1821
1822 stretched_ratio = info_dict.get('stretched_ratio')
1823 if stretched_ratio is not None and stretched_ratio != 1:
1824 if fixup_policy == 'warn':
1825 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1826 info_dict['id'], stretched_ratio))
1827 elif fixup_policy == 'detect_or_warn':
1828 stretched_pp = FFmpegFixupStretchedPP(self)
1829 if stretched_pp.available:
1830 info_dict.setdefault('__postprocessors', [])
1831 info_dict['__postprocessors'].append(stretched_pp)
1832 else:
1833 self.report_warning(
1834 '%s: Non-uniform pixel ratio (%s). %s'
1835 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1836 else:
1837 assert fixup_policy in ('ignore', 'never')
1838
1839 if (info_dict.get('requested_formats') is None and
1840 info_dict.get('container') == 'm4a_dash'):
1841 if fixup_policy == 'warn':
1842 self.report_warning(
1843 '%s: writing DASH m4a. '
1844 'Only some players support this container.'
1845 % info_dict['id'])
1846 elif fixup_policy == 'detect_or_warn':
1847 fixup_pp = FFmpegFixupM4aPP(self)
1848 if fixup_pp.available:
1849 info_dict.setdefault('__postprocessors', [])
1850 info_dict['__postprocessors'].append(fixup_pp)
1851 else:
1852 self.report_warning(
1853 '%s: writing DASH m4a. '
1854 'Only some players support this container. %s'
1855 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1856 else:
1857 assert fixup_policy in ('ignore', 'never')
1858
1859 if (info_dict.get('protocol') == 'm3u8_native' or
1860 info_dict.get('protocol') == 'm3u8' and
1861 self.params.get('hls_prefer_native')):
1862 if fixup_policy == 'warn':
1863 self.report_warning('%s: malformated aac bitstream.' % (
1864 info_dict['id']))
1865 elif fixup_policy == 'detect_or_warn':
1866 fixup_pp = FFmpegFixupM3u8PP(self)
1867 if fixup_pp.available:
1868 info_dict.setdefault('__postprocessors', [])
1869 info_dict['__postprocessors'].append(fixup_pp)
1870 else:
1871 self.report_warning(
1872 '%s: malformated aac bitstream. %s'
1873 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1874 else:
1875 assert fixup_policy in ('ignore', 'never')
1876
1877 try:
1878 self.post_process(filename, info_dict)
1879 except (PostProcessingError) as err:
1880 self.report_error('postprocessing: %s' % str(err))
1881 return
1882 self.record_download_archive(info_dict)
1883
1884 def download(self, url_list):
1885 """Download a given list of URLs."""
1886 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1887 if (len(url_list) > 1 and
1888 outtmpl != '-' and
1889 '%' not in outtmpl and
1890 self.params.get('max_downloads') != 1):
1891 raise SameFileError(outtmpl)
1892
1893 for url in url_list:
1894 try:
1895 # It also downloads the videos
1896 res = self.extract_info(
1897 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1898 except UnavailableVideoError:
1899 self.report_error('unable to download video')
1900 except MaxDownloadsReached:
1901 self.to_screen('[info] Maximum number of downloaded files reached.')
1902 raise
1903 else:
1904 if self.params.get('dump_single_json', False):
1905 self.to_stdout(json.dumps(res))
1906
1907 return self._download_retcode
1908
1909 def download_with_info_file(self, info_filename):
1910 with contextlib.closing(fileinput.FileInput(
1911 [info_filename], mode='r',
1912 openhook=fileinput.hook_encoded('utf-8'))) as f:
1913 # FileInput doesn't have a read method, we can't call json.load
1914 info = self.filter_requested_info(json.loads('\n'.join(f)))
1915 try:
1916 self.process_ie_result(info, download=True)
1917 except DownloadError:
1918 webpage_url = info.get('webpage_url')
1919 if webpage_url is not None:
1920 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1921 return self.download([webpage_url])
1922 else:
1923 raise
1924 return self._download_retcode
1925
1926 @staticmethod
1927 def filter_requested_info(info_dict):
1928 return dict(
1929 (k, v) for k, v in info_dict.items()
1930 if k not in ['requested_formats', 'requested_subtitles'])
1931
1932 def post_process(self, filename, ie_info):
1933 """Run all the postprocessors on the given file."""
1934 info = dict(ie_info)
1935 info['filepath'] = filename
1936 pps_chain = []
1937 if ie_info.get('__postprocessors') is not None:
1938 pps_chain.extend(ie_info['__postprocessors'])
1939 pps_chain.extend(self._pps)
1940 for pp in pps_chain:
1941 files_to_delete = []
1942 try:
1943 files_to_delete, info = pp.run(info)
1944 except PostProcessingError as e:
1945 self.report_error(e.msg)
1946 if files_to_delete and not self.params.get('keepvideo', False):
1947 for old_filename in files_to_delete:
1948 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1949 try:
1950 os.remove(encodeFilename(old_filename))
1951 except (IOError, OSError):
1952 self.report_warning('Unable to remove downloaded original file')
1953
1954 def _make_archive_id(self, info_dict):
1955 # Future-proof against any change in case
1956 # and backwards compatibility with prior versions
1957 extractor = info_dict.get('extractor_key')
1958 if extractor is None:
1959 if 'id' in info_dict:
1960 extractor = info_dict.get('ie_key') # key in a playlist
1961 if extractor is None:
1962 return None # Incomplete video information
1963 return extractor.lower() + ' ' + info_dict['id']
1964
1965 def in_download_archive(self, info_dict):
1966 fn = self.params.get('download_archive')
1967 if fn is None:
1968 return False
1969
1970 vid_id = self._make_archive_id(info_dict)
1971 if vid_id is None:
1972 return False # Incomplete video information
1973
1974 try:
1975 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1976 for line in archive_file:
1977 if line.strip() == vid_id:
1978 return True
1979 except IOError as ioe:
1980 if ioe.errno != errno.ENOENT:
1981 raise
1982 return False
1983
1984 def record_download_archive(self, info_dict):
1985 fn = self.params.get('download_archive')
1986 if fn is None:
1987 return
1988 vid_id = self._make_archive_id(info_dict)
1989 assert vid_id
1990 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1991 archive_file.write(vid_id + '\n')
1992
1993 @staticmethod
1994 def format_resolution(format, default='unknown'):
1995 if format.get('vcodec') == 'none':
1996 return 'audio only'
1997 if format.get('resolution') is not None:
1998 return format['resolution']
1999 if format.get('height') is not None:
2000 if format.get('width') is not None:
2001 res = '%sx%s' % (format['width'], format['height'])
2002 else:
2003 res = '%sp' % format['height']
2004 elif format.get('width') is not None:
2005 res = '%dx?' % format['width']
2006 else:
2007 res = default
2008 return res
2009
2010 def _format_note(self, fdict):
2011 res = ''
2012 if fdict.get('ext') in ['f4f', 'f4m']:
2013 res += '(unsupported) '
2014 if fdict.get('language'):
2015 if res:
2016 res += ' '
2017 res += '[%s] ' % fdict['language']
2018 if fdict.get('format_note') is not None:
2019 res += fdict['format_note'] + ' '
2020 if fdict.get('tbr') is not None:
2021 res += '%4dk ' % fdict['tbr']
2022 if fdict.get('container') is not None:
2023 if res:
2024 res += ', '
2025 res += '%s container' % fdict['container']
2026 if (fdict.get('vcodec') is not None and
2027 fdict.get('vcodec') != 'none'):
2028 if res:
2029 res += ', '
2030 res += fdict['vcodec']
2031 if fdict.get('vbr') is not None:
2032 res += '@'
2033 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2034 res += 'video@'
2035 if fdict.get('vbr') is not None:
2036 res += '%4dk' % fdict['vbr']
2037 if fdict.get('fps') is not None:
2038 if res:
2039 res += ', '
2040 res += '%sfps' % fdict['fps']
2041 if fdict.get('acodec') is not None:
2042 if res:
2043 res += ', '
2044 if fdict['acodec'] == 'none':
2045 res += 'video only'
2046 else:
2047 res += '%-5s' % fdict['acodec']
2048 elif fdict.get('abr') is not None:
2049 if res:
2050 res += ', '
2051 res += 'audio'
2052 if fdict.get('abr') is not None:
2053 res += '@%3dk' % fdict['abr']
2054 if fdict.get('asr') is not None:
2055 res += ' (%5dHz)' % fdict['asr']
2056 if fdict.get('filesize') is not None:
2057 if res:
2058 res += ', '
2059 res += format_bytes(fdict['filesize'])
2060 elif fdict.get('filesize_approx') is not None:
2061 if res:
2062 res += ', '
2063 res += '~' + format_bytes(fdict['filesize_approx'])
2064 return res
2065
2066 def list_formats(self, info_dict):
2067 formats = info_dict.get('formats', [info_dict])
2068 table = [
2069 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
2070 for f in formats
2071 if f.get('preference') is None or f['preference'] >= -1000]
2072 if len(formats) > 1:
2073 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
2074
2075 header_line = ['format code', 'extension', 'resolution', 'note']
2076 self.to_screen(
2077 '[info] Available formats for %s:\n%s' %
2078 (info_dict['id'], render_table(header_line, table)))
2079
2080 def list_thumbnails(self, info_dict):
2081 thumbnails = info_dict.get('thumbnails')
2082 if not thumbnails:
2083 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
2084 return
2085
2086 self.to_screen(
2087 '[info] Thumbnails for %s:' % info_dict['id'])
2088 self.to_screen(render_table(
2089 ['ID', 'width', 'height', 'URL'],
2090 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
2091
2092 def list_subtitles(self, video_id, subtitles, name='subtitles'):
2093 if not subtitles:
2094 self.to_screen('%s has no %s' % (video_id, name))
2095 return
2096 self.to_screen(
2097 'Available %s for %s:' % (name, video_id))
2098 self.to_screen(render_table(
2099 ['Language', 'formats'],
2100 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
2101 for lang, formats in subtitles.items()]))
2102
2103 def urlopen(self, req):
2104 """ Start an HTTP download """
2105 if isinstance(req, compat_basestring):
2106 req = sanitized_Request(req)
2107 return self._opener.open(req, timeout=self._socket_timeout)
2108
2109 def print_debug_header(self):
2110 if not self.params.get('verbose'):
2111 return
2112
2113 if type('') is not compat_str:
2114 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
2115 self.report_warning(
2116 'Your Python is broken! Update to a newer and supported version')
2117
2118 stdout_encoding = getattr(
2119 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2120 encoding_str = (
2121 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2122 locale.getpreferredencoding(),
2123 sys.getfilesystemencoding(),
2124 stdout_encoding,
2125 self.get_encoding()))
2126 write_string(encoding_str, encoding=None)
2127
2128 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
2129 if _LAZY_LOADER:
2130 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2131 try:
2132 sp = subprocess.Popen(
2133 ['git', 'rev-parse', '--short', 'HEAD'],
2134 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2135 cwd=os.path.dirname(os.path.abspath(__file__)))
2136 out, err = sp.communicate()
2137 out = out.decode().strip()
2138 if re.match('[0-9a-f]+', out):
2139 self._write_string('[debug] Git HEAD: ' + out + '\n')
2140 except Exception:
2141 try:
2142 sys.exc_clear()
2143 except Exception:
2144 pass
2145 self._write_string('[debug] Python version %s - %s\n' % (
2146 platform.python_version(), platform_name()))
2147
2148 exe_versions = FFmpegPostProcessor.get_versions(self)
2149 exe_versions['rtmpdump'] = rtmpdump_version()
2150 exe_str = ', '.join(
2151 '%s %s' % (exe, v)
2152 for exe, v in sorted(exe_versions.items())
2153 if v
2154 )
2155 if not exe_str:
2156 exe_str = 'none'
2157 self._write_string('[debug] exe versions: %s\n' % exe_str)
2158
2159 proxy_map = {}
2160 for handler in self._opener.handlers:
2161 if hasattr(handler, 'proxies'):
2162 proxy_map.update(handler.proxies)
2163 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2164
2165 if self.params.get('call_home', False):
2166 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2167 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2168 latest_version = self.urlopen(
2169 'https://yt-dl.org/latest/version').read().decode('utf-8')
2170 if version_tuple(latest_version) > version_tuple(__version__):
2171 self.report_warning(
2172 'You are using an outdated version (newest version: %s)! '
2173 'See https://yt-dl.org/update if you need help updating.' %
2174 latest_version)
2175
2176 def _setup_opener(self):
2177 timeout_val = self.params.get('socket_timeout')
2178 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2179
2180 opts_cookiefile = self.params.get('cookiefile')
2181 opts_proxy = self.params.get('proxy')
2182
2183 if opts_cookiefile is None:
2184 self.cookiejar = compat_cookiejar.CookieJar()
2185 else:
2186 opts_cookiefile = expand_path(opts_cookiefile)
2187 self.cookiejar = compat_cookiejar.MozillaCookieJar(
2188 opts_cookiefile)
2189 if os.access(opts_cookiefile, os.R_OK):
2190 self.cookiejar.load()
2191
2192 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2193 if opts_proxy is not None:
2194 if opts_proxy == '':
2195 proxies = {}
2196 else:
2197 proxies = {'http': opts_proxy, 'https': opts_proxy}
2198 else:
2199 proxies = compat_urllib_request.getproxies()
2200 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2201 if 'http' in proxies and 'https' not in proxies:
2202 proxies['https'] = proxies['http']
2203 proxy_handler = PerRequestProxyHandler(proxies)
2204
2205 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2206 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2207 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2208 data_handler = compat_urllib_request_DataHandler()
2209
2210 # When passing our own FileHandler instance, build_opener won't add the
2211 # default FileHandler and allows us to disable the file protocol, which
2212 # can be used for malicious purposes (see
2213 # https://github.com/rg3/youtube-dl/issues/8227)
2214 file_handler = compat_urllib_request.FileHandler()
2215
2216 def file_open(*args, **kwargs):
2217 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2218 file_handler.file_open = file_open
2219
2220 opener = compat_urllib_request.build_opener(
2221 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2222
2223 # Delete the default user-agent header, which would otherwise apply in
2224 # cases where our custom HTTP handler doesn't come into play
2225 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2226 opener.addheaders = []
2227 self._opener = opener
2228
2229 def encode(self, s):
2230 if isinstance(s, bytes):
2231 return s # Already encoded
2232
2233 try:
2234 return s.encode(self.get_encoding())
2235 except UnicodeEncodeError as err:
2236 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2237 raise
2238
2239 def get_encoding(self):
2240 encoding = self.params.get('encoding')
2241 if encoding is None:
2242 encoding = preferredencoding()
2243 return encoding
2244
2245 def _write_thumbnails(self, info_dict, filename):
2246 if self.params.get('writethumbnail', False):
2247 thumbnails = info_dict.get('thumbnails')
2248 if thumbnails:
2249 thumbnails = [thumbnails[-1]]
2250 elif self.params.get('write_all_thumbnails', False):
2251 thumbnails = info_dict.get('thumbnails')
2252 else:
2253 return
2254
2255 if not thumbnails:
2256 # No thumbnails present, so return immediately
2257 return
2258
2259 for t in thumbnails:
2260 thumb_ext = determine_ext(t['url'], 'jpg')
2261 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2262 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2263 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2264
2265 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2266 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2267 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2268 else:
2269 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2270 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2271 try:
2272 uf = self.urlopen(t['url'])
2273 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2274 shutil.copyfileobj(uf, thumbf)
2275 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2276 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2277 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2278 self.report_warning('Unable to download thumbnail "%s": %s' %
2279 (t['url'], error_to_compat_str(err)))