]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/YoutubeDL.py
4493fd0e1aabc0452041a68262772b9b5d223d79
[youtubedl] / 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 string import ascii_letters
30
31 from .compat import (
32 compat_basestring,
33 compat_cookiejar,
34 compat_get_terminal_size,
35 compat_http_client,
36 compat_kwargs,
37 compat_numeric_types,
38 compat_os_name,
39 compat_str,
40 compat_tokenize_tokenize,
41 compat_urllib_error,
42 compat_urllib_request,
43 compat_urllib_request_DataHandler,
44 )
45 from .utils import (
46 age_restricted,
47 args_to_str,
48 ContentTooShortError,
49 date_from_str,
50 DateRange,
51 DEFAULT_OUTTMPL,
52 determine_ext,
53 determine_protocol,
54 DownloadError,
55 encode_compat_str,
56 encodeFilename,
57 error_to_compat_str,
58 expand_path,
59 ExtractorError,
60 format_bytes,
61 formatSeconds,
62 GeoRestrictedError,
63 int_or_none,
64 ISO3166Utils,
65 locked_file,
66 make_HTTPS_handler,
67 MaxDownloadsReached,
68 orderedSet,
69 PagedList,
70 parse_filesize,
71 PerRequestProxyHandler,
72 platform_name,
73 PostProcessingError,
74 preferredencoding,
75 prepend_extension,
76 register_socks_protocols,
77 render_table,
78 replace_extension,
79 SameFileError,
80 sanitize_filename,
81 sanitize_path,
82 sanitize_url,
83 sanitized_Request,
84 std_headers,
85 subtitles_filename,
86 UnavailableVideoError,
87 url_basename,
88 version_tuple,
89 write_json_file,
90 write_string,
91 YoutubeDLCookieJar,
92 YoutubeDLCookieProcessor,
93 YoutubeDLHandler,
94 )
95 from .cache import Cache
96 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
97 from .extractor.openload import PhantomJSwrapper
98 from .downloader import get_suitable_downloader
99 from .downloader.rtmp import rtmpdump_version
100 from .postprocessor import (
101 FFmpegFixupM3u8PP,
102 FFmpegFixupM4aPP,
103 FFmpegFixupStretchedPP,
104 FFmpegMergerPP,
105 FFmpegPostProcessor,
106 get_postprocessor,
107 )
108 from .version import __version__
109
110 if compat_os_name == 'nt':
111 import ctypes
112
113
114 class YoutubeDL(object):
115 """YoutubeDL class.
116
117 YoutubeDL objects are the ones responsible of downloading the
118 actual video file and writing it to disk if the user has requested
119 it, among some other tasks. In most cases there should be one per
120 program. As, given a video URL, the downloader doesn't know how to
121 extract all the needed information, task that InfoExtractors do, it
122 has to pass the URL to one of them.
123
124 For this, YoutubeDL objects have a method that allows
125 InfoExtractors to be registered in a given order. When it is passed
126 a URL, the YoutubeDL object handles it to the first InfoExtractor it
127 finds that reports being able to handle it. The InfoExtractor extracts
128 all the information about the video or videos the URL refers to, and
129 YoutubeDL process the extracted information, possibly using a File
130 Downloader to download the video.
131
132 YoutubeDL objects accept a lot of parameters. In order not to saturate
133 the object constructor with arguments, it receives a dictionary of
134 options instead. These options are available through the params
135 attribute for the InfoExtractors to use. The YoutubeDL also
136 registers itself as the downloader in charge for the InfoExtractors
137 that are added to it, so this is a "mutual registration".
138
139 Available options:
140
141 username: Username for authentication purposes.
142 password: Password for authentication purposes.
143 videopassword: Password for accessing a video.
144 ap_mso: Adobe Pass multiple-system operator identifier.
145 ap_username: Multiple-system operator account username.
146 ap_password: Multiple-system operator account password.
147 usenetrc: Use netrc for authentication instead.
148 verbose: Print additional info to stdout.
149 quiet: Do not print messages to stdout.
150 no_warnings: Do not print out anything for warnings.
151 forceurl: Force printing final URL.
152 forcetitle: Force printing title.
153 forceid: Force printing ID.
154 forcethumbnail: Force printing thumbnail URL.
155 forcedescription: Force printing description.
156 forcefilename: Force printing final filename.
157 forceduration: Force printing duration.
158 forcejson: Force printing info_dict as JSON.
159 dump_single_json: Force printing the info_dict of the whole playlist
160 (or video) as a single JSON line.
161 simulate: Do not download the video files.
162 format: Video format code. See options.py for more information.
163 outtmpl: Template for output names.
164 restrictfilenames: Do not allow "&" and spaces in file names
165 ignoreerrors: Do not stop on download errors.
166 force_generic_extractor: Force downloader to use the generic extractor
167 nooverwrites: Prevent overwriting files.
168 playliststart: Playlist item to start at.
169 playlistend: Playlist item to end at.
170 playlist_items: Specific indices of playlist to download.
171 playlistreverse: Download playlist items in reverse order.
172 playlistrandom: Download playlist items in random order.
173 matchtitle: Download only matching titles.
174 rejecttitle: Reject downloads for matching titles.
175 logger: Log messages to a logging.Logger instance.
176 logtostderr: Log messages to stderr instead of stdout.
177 writedescription: Write the video description to a .description file
178 writeinfojson: Write the video description to a .info.json file
179 writeannotations: Write the video annotations to a .annotations.xml file
180 writethumbnail: Write the thumbnail image to a file
181 write_all_thumbnails: Write all thumbnail formats to files
182 writesubtitles: Write the video subtitles to a file
183 writeautomaticsub: Write the automatically generated subtitles to a file
184 allsubtitles: Downloads all the subtitles of the video
185 (requires writesubtitles or writeautomaticsub)
186 listsubtitles: Lists all available subtitles for the video
187 subtitlesformat: The format code for subtitles
188 subtitleslangs: List of languages of the subtitles to download
189 keepvideo: Keep the video file after post-processing
190 daterange: A DateRange object, download only if the upload_date is in the range.
191 skip_download: Skip the actual download of the video file
192 cachedir: Location of the cache files in the filesystem.
193 False to disable filesystem cache.
194 noplaylist: Download single video instead of a playlist if in doubt.
195 age_limit: An integer representing the user's age in years.
196 Unsuitable videos for the given age are skipped.
197 min_views: An integer representing the minimum view count the video
198 must have in order to not be skipped.
199 Videos without view count information are always
200 downloaded. None for no limit.
201 max_views: An integer representing the maximum view count.
202 Videos that are more popular than that are not
203 downloaded.
204 Videos without view count information are always
205 downloaded. None for no limit.
206 download_archive: File name of a file where all downloads are recorded.
207 Videos already present in the file are not downloaded
208 again.
209 cookiefile: File name where cookies should be read from and dumped to.
210 nocheckcertificate:Do not verify SSL certificates
211 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
212 At the moment, this is only supported by YouTube.
213 proxy: URL of the proxy server to use
214 geo_verification_proxy: URL of the proxy to use for IP address verification
215 on geo-restricted sites.
216 socket_timeout: Time to wait for unresponsive hosts, in seconds
217 bidi_workaround: Work around buggy terminals without bidirectional text
218 support, using fridibi
219 debug_printtraffic:Print out sent and received HTTP traffic
220 include_ads: Download ads as well
221 default_search: Prepend this string if an input url is not valid.
222 'auto' for elaborate guessing
223 encoding: Use this encoding instead of the system-specified.
224 extract_flat: Do not resolve URLs, return the immediate result.
225 Pass in 'in_playlist' to only show this behavior for
226 playlist items.
227 postprocessors: A list of dictionaries, each with an entry
228 * key: The name of the postprocessor. See
229 youtube_dl/postprocessor/__init__.py for a list.
230 as well as any further keyword arguments for the
231 postprocessor.
232 progress_hooks: A list of functions that get called on download
233 progress, with a dictionary with the entries
234 * status: One of "downloading", "error", or "finished".
235 Check this first and ignore unknown values.
236
237 If status is one of "downloading", or "finished", the
238 following properties may also be present:
239 * filename: The final filename (always present)
240 * tmpfilename: The filename we're currently writing to
241 * downloaded_bytes: Bytes on disk
242 * total_bytes: Size of the whole file, None if unknown
243 * total_bytes_estimate: Guess of the eventual file size,
244 None if unavailable.
245 * elapsed: The number of seconds since download started.
246 * eta: The estimated time in seconds, None if unknown
247 * speed: The download speed in bytes/second, None if
248 unknown
249 * fragment_index: The counter of the currently
250 downloaded video fragment.
251 * fragment_count: The number of fragments (= individual
252 files that will be merged)
253
254 Progress hooks are guaranteed to be called at least once
255 (with status "finished") if the download is successful.
256 merge_output_format: Extension to use when merging formats.
257 fixup: Automatically correct known faults of the file.
258 One of:
259 - "never": do nothing
260 - "warn": only emit a warning
261 - "detect_or_warn": check whether we can do anything
262 about it, warn otherwise (default)
263 source_address: Client-side IP address to bind to.
264 call_home: Boolean, true iff we are allowed to contact the
265 youtube-dl servers for debugging.
266 sleep_interval: Number of seconds to sleep before each download when
267 used alone or a lower bound of a range for randomized
268 sleep before each download (minimum possible number
269 of seconds to sleep) when used along with
270 max_sleep_interval.
271 max_sleep_interval:Upper bound of a range for randomized sleep before each
272 download (maximum possible number of seconds to sleep).
273 Must only be used along with sleep_interval.
274 Actual sleep time will be a random float from range
275 [sleep_interval; max_sleep_interval].
276 listformats: Print an overview of available video formats and exit.
277 list_thumbnails: Print a table of all thumbnails and exit.
278 match_filter: A function that gets called with the info_dict of
279 every video.
280 If it returns a message, the video is ignored.
281 If it returns None, the video is downloaded.
282 match_filter_func in utils.py is one example for this.
283 no_color: Do not emit color codes in output.
284 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
285 HTTP header
286 geo_bypass_country:
287 Two-letter ISO 3166-2 country code that will be used for
288 explicit geographic restriction bypassing via faking
289 X-Forwarded-For HTTP header
290 geo_bypass_ip_block:
291 IP range in CIDR notation that will be used similarly to
292 geo_bypass_country
293
294 The following options determine which downloader is picked:
295 external_downloader: Executable of the external downloader to call.
296 None or unset for standard (built-in) downloader.
297 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
298 if True, otherwise use ffmpeg/avconv if False, otherwise
299 use downloader suggested by extractor if None.
300
301 The following parameters are not used by YoutubeDL itself, they are used by
302 the downloader (see youtube_dl/downloader/common.py):
303 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
304 noresizebuffer, retries, continuedl, noprogress, consoletitle,
305 xattr_set_filesize, external_downloader_args, hls_use_mpegts,
306 http_chunk_size.
307
308 The following options are used by the post processors:
309 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
310 otherwise prefer ffmpeg.
311 postprocessor_args: A list of additional command-line arguments for the
312 postprocessor.
313
314 The following options are used by the Youtube extractor:
315 youtube_include_dash_manifest: If True (default), DASH manifests and related
316 data will be downloaded and processed by extractor.
317 You can reduce network I/O by disabling it if you don't
318 care about DASH.
319 """
320
321 _NUMERIC_FIELDS = set((
322 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
323 'timestamp', 'upload_year', 'upload_month', 'upload_day',
324 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
325 'average_rating', 'comment_count', 'age_limit',
326 'start_time', 'end_time',
327 'chapter_number', 'season_number', 'episode_number',
328 'track_number', 'disc_number', 'release_year',
329 'playlist_index',
330 ))
331
332 params = None
333 _ies = []
334 _pps = []
335 _download_retcode = None
336 _num_downloads = None
337 _screen_file = None
338
339 def __init__(self, params=None, auto_init=True):
340 """Create a FileDownloader object with the given options."""
341 if params is None:
342 params = {}
343 self._ies = []
344 self._ies_instances = {}
345 self._pps = []
346 self._progress_hooks = []
347 self._download_retcode = 0
348 self._num_downloads = 0
349 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
350 self._err_file = sys.stderr
351 self.params = {
352 # Default parameters
353 'nocheckcertificate': False,
354 }
355 self.params.update(params)
356 self.cache = Cache(self)
357
358 def check_deprecated(param, option, suggestion):
359 if self.params.get(param) is not None:
360 self.report_warning(
361 '%s is deprecated. Use %s instead.' % (option, suggestion))
362 return True
363 return False
364
365 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
366 if self.params.get('geo_verification_proxy') is None:
367 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
368
369 check_deprecated('autonumber_size', '--autonumber-size', 'output template with %(autonumber)0Nd, where N in the number of digits')
370 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
371 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
372
373 if params.get('bidi_workaround', False):
374 try:
375 import pty
376 master, slave = pty.openpty()
377 width = compat_get_terminal_size().columns
378 if width is None:
379 width_args = []
380 else:
381 width_args = ['-w', str(width)]
382 sp_kwargs = dict(
383 stdin=subprocess.PIPE,
384 stdout=slave,
385 stderr=self._err_file)
386 try:
387 self._output_process = subprocess.Popen(
388 ['bidiv'] + width_args, **sp_kwargs
389 )
390 except OSError:
391 self._output_process = subprocess.Popen(
392 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
393 self._output_channel = os.fdopen(master, 'rb')
394 except OSError as ose:
395 if ose.errno == errno.ENOENT:
396 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 $PATH.')
397 else:
398 raise
399
400 if (sys.platform != 'win32' and
401 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
402 not params.get('restrictfilenames', False)):
403 # Unicode filesystem API will throw errors (#1474, #13027)
404 self.report_warning(
405 'Assuming --restrict-filenames since file system encoding '
406 'cannot encode all characters. '
407 'Set the LC_ALL environment variable to fix this.')
408 self.params['restrictfilenames'] = True
409
410 if isinstance(params.get('outtmpl'), bytes):
411 self.report_warning(
412 'Parameter outtmpl is bytes, but should be a unicode string. '
413 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
414
415 self._setup_opener()
416
417 if auto_init:
418 self.print_debug_header()
419 self.add_default_info_extractors()
420
421 for pp_def_raw in self.params.get('postprocessors', []):
422 pp_class = get_postprocessor(pp_def_raw['key'])
423 pp_def = dict(pp_def_raw)
424 del pp_def['key']
425 pp = pp_class(self, **compat_kwargs(pp_def))
426 self.add_post_processor(pp)
427
428 for ph in self.params.get('progress_hooks', []):
429 self.add_progress_hook(ph)
430
431 register_socks_protocols()
432
433 def warn_if_short_id(self, argv):
434 # short YouTube ID starting with dash?
435 idxs = [
436 i for i, a in enumerate(argv)
437 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
438 if idxs:
439 correct_argv = (
440 ['youtube-dl'] +
441 [a for i, a in enumerate(argv) if i not in idxs] +
442 ['--'] + [argv[i] for i in idxs]
443 )
444 self.report_warning(
445 'Long argument string detected. '
446 'Use -- to separate parameters and URLs, like this:\n%s\n' %
447 args_to_str(correct_argv))
448
449 def add_info_extractor(self, ie):
450 """Add an InfoExtractor object to the end of the list."""
451 self._ies.append(ie)
452 if not isinstance(ie, type):
453 self._ies_instances[ie.ie_key()] = ie
454 ie.set_downloader(self)
455
456 def get_info_extractor(self, ie_key):
457 """
458 Get an instance of an IE with name ie_key, it will try to get one from
459 the _ies list, if there's no instance it will create a new one and add
460 it to the extractor list.
461 """
462 ie = self._ies_instances.get(ie_key)
463 if ie is None:
464 ie = get_info_extractor(ie_key)()
465 self.add_info_extractor(ie)
466 return ie
467
468 def add_default_info_extractors(self):
469 """
470 Add the InfoExtractors returned by gen_extractors to the end of the list
471 """
472 for ie in gen_extractor_classes():
473 self.add_info_extractor(ie)
474
475 def add_post_processor(self, pp):
476 """Add a PostProcessor object to the end of the chain."""
477 self._pps.append(pp)
478 pp.set_downloader(self)
479
480 def add_progress_hook(self, ph):
481 """Add the progress hook (currently only for the file downloader)"""
482 self._progress_hooks.append(ph)
483
484 def _bidi_workaround(self, message):
485 if not hasattr(self, '_output_channel'):
486 return message
487
488 assert hasattr(self, '_output_process')
489 assert isinstance(message, compat_str)
490 line_count = message.count('\n') + 1
491 self._output_process.stdin.write((message + '\n').encode('utf-8'))
492 self._output_process.stdin.flush()
493 res = ''.join(self._output_channel.readline().decode('utf-8')
494 for _ in range(line_count))
495 return res[:-len('\n')]
496
497 def to_screen(self, message, skip_eol=False):
498 """Print message to stdout if not in quiet mode."""
499 return self.to_stdout(message, skip_eol, check_quiet=True)
500
501 def _write_string(self, s, out=None):
502 write_string(s, out=out, encoding=self.params.get('encoding'))
503
504 def to_stdout(self, message, skip_eol=False, check_quiet=False):
505 """Print message to stdout if not in quiet mode."""
506 if self.params.get('logger'):
507 self.params['logger'].debug(message)
508 elif not check_quiet or not self.params.get('quiet', False):
509 message = self._bidi_workaround(message)
510 terminator = ['\n', ''][skip_eol]
511 output = message + terminator
512
513 self._write_string(output, self._screen_file)
514
515 def to_stderr(self, message):
516 """Print message to stderr."""
517 assert isinstance(message, compat_str)
518 if self.params.get('logger'):
519 self.params['logger'].error(message)
520 else:
521 message = self._bidi_workaround(message)
522 output = message + '\n'
523 self._write_string(output, self._err_file)
524
525 def to_console_title(self, message):
526 if not self.params.get('consoletitle', False):
527 return
528 if compat_os_name == 'nt':
529 if ctypes.windll.kernel32.GetConsoleWindow():
530 # c_wchar_p() might not be necessary if `message` is
531 # already of type unicode()
532 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
533 elif 'TERM' in os.environ:
534 self._write_string('\033]0;%s\007' % message, self._screen_file)
535
536 def save_console_title(self):
537 if not self.params.get('consoletitle', False):
538 return
539 if self.params.get('simulate', False):
540 return
541 if compat_os_name != 'nt' and 'TERM' in os.environ:
542 # Save the title on stack
543 self._write_string('\033[22;0t', self._screen_file)
544
545 def restore_console_title(self):
546 if not self.params.get('consoletitle', False):
547 return
548 if self.params.get('simulate', False):
549 return
550 if compat_os_name != 'nt' and 'TERM' in os.environ:
551 # Restore the title from stack
552 self._write_string('\033[23;0t', self._screen_file)
553
554 def __enter__(self):
555 self.save_console_title()
556 return self
557
558 def __exit__(self, *args):
559 self.restore_console_title()
560
561 if self.params.get('cookiefile') is not None:
562 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
563
564 def trouble(self, message=None, tb=None):
565 """Determine action to take when a download problem appears.
566
567 Depending on if the downloader has been configured to ignore
568 download errors or not, this method may throw an exception or
569 not when errors are found, after printing the message.
570
571 tb, if given, is additional traceback information.
572 """
573 if message is not None:
574 self.to_stderr(message)
575 if self.params.get('verbose'):
576 if tb is None:
577 if sys.exc_info()[0]: # if .trouble has been called from an except block
578 tb = ''
579 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
580 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
581 tb += encode_compat_str(traceback.format_exc())
582 else:
583 tb_data = traceback.format_list(traceback.extract_stack())
584 tb = ''.join(tb_data)
585 self.to_stderr(tb)
586 if not self.params.get('ignoreerrors', False):
587 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
588 exc_info = sys.exc_info()[1].exc_info
589 else:
590 exc_info = sys.exc_info()
591 raise DownloadError(message, exc_info)
592 self._download_retcode = 1
593
594 def report_warning(self, message):
595 '''
596 Print the message to stderr, it will be prefixed with 'WARNING:'
597 If stderr is a tty file the 'WARNING:' will be colored
598 '''
599 if self.params.get('logger') is not None:
600 self.params['logger'].warning(message)
601 else:
602 if self.params.get('no_warnings'):
603 return
604 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
605 _msg_header = '\033[0;33mWARNING:\033[0m'
606 else:
607 _msg_header = 'WARNING:'
608 warning_message = '%s %s' % (_msg_header, message)
609 self.to_stderr(warning_message)
610
611 def report_error(self, message, tb=None):
612 '''
613 Do the same as trouble, but prefixes the message with 'ERROR:', colored
614 in red if stderr is a tty file.
615 '''
616 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
617 _msg_header = '\033[0;31mERROR:\033[0m'
618 else:
619 _msg_header = 'ERROR:'
620 error_message = '%s %s' % (_msg_header, message)
621 self.trouble(error_message, tb)
622
623 def report_file_already_downloaded(self, file_name):
624 """Report file has already been fully downloaded."""
625 try:
626 self.to_screen('[download] %s has already been downloaded' % file_name)
627 except UnicodeEncodeError:
628 self.to_screen('[download] The file has already been downloaded')
629
630 def prepare_filename(self, info_dict):
631 """Generate the output filename."""
632 try:
633 template_dict = dict(info_dict)
634
635 template_dict['epoch'] = int(time.time())
636 autonumber_size = self.params.get('autonumber_size')
637 if autonumber_size is None:
638 autonumber_size = 5
639 template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
640 if template_dict.get('resolution') is None:
641 if template_dict.get('width') and template_dict.get('height'):
642 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
643 elif template_dict.get('height'):
644 template_dict['resolution'] = '%sp' % template_dict['height']
645 elif template_dict.get('width'):
646 template_dict['resolution'] = '%dx?' % template_dict['width']
647
648 sanitize = lambda k, v: sanitize_filename(
649 compat_str(v),
650 restricted=self.params.get('restrictfilenames'),
651 is_id=(k == 'id' or k.endswith('_id')))
652 template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
653 for k, v in template_dict.items()
654 if v is not None and not isinstance(v, (list, tuple, dict)))
655 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
656
657 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
658
659 # For fields playlist_index and autonumber convert all occurrences
660 # of %(field)s to %(field)0Nd for backward compatibility
661 field_size_compat_map = {
662 'playlist_index': len(str(template_dict['n_entries'])),
663 'autonumber': autonumber_size,
664 }
665 FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
666 mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
667 if mobj:
668 outtmpl = re.sub(
669 FIELD_SIZE_COMPAT_RE,
670 r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
671 outtmpl)
672
673 # Missing numeric fields used together with integer presentation types
674 # in format specification will break the argument substitution since
675 # string 'NA' is returned for missing fields. We will patch output
676 # template for missing fields to meet string presentation type.
677 for numeric_field in self._NUMERIC_FIELDS:
678 if numeric_field not in template_dict:
679 # As of [1] format syntax is:
680 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
681 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
682 FORMAT_RE = r'''(?x)
683 (?<!%)
684 %
685 \({0}\) # mapping key
686 (?:[#0\-+ ]+)? # conversion flags (optional)
687 (?:\d+)? # minimum field width (optional)
688 (?:\.\d+)? # precision (optional)
689 [hlL]? # length modifier (optional)
690 [diouxXeEfFgGcrs%] # conversion type
691 '''
692 outtmpl = re.sub(
693 FORMAT_RE.format(numeric_field),
694 r'%({0})s'.format(numeric_field), outtmpl)
695
696 # expand_path translates '%%' into '%' and '$$' into '$'
697 # correspondingly that is not what we want since we need to keep
698 # '%%' intact for template dict substitution step. Working around
699 # with boundary-alike separator hack.
700 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
701 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
702
703 # outtmpl should be expand_path'ed before template dict substitution
704 # because meta fields may contain env variables we don't want to
705 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
706 # title "Hello $PATH", we don't want `$PATH` to be expanded.
707 filename = expand_path(outtmpl).replace(sep, '') % template_dict
708
709 # Temporary fix for #4787
710 # 'Treat' all problem characters by passing filename through preferredencoding
711 # to workaround encoding issues with subprocess on python2 @ Windows
712 if sys.version_info < (3, 0) and sys.platform == 'win32':
713 filename = encodeFilename(filename, True).decode(preferredencoding())
714 return sanitize_path(filename)
715 except ValueError as err:
716 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
717 return None
718
719 def _match_entry(self, info_dict, incomplete):
720 """ Returns None iff the file should be downloaded """
721
722 video_title = info_dict.get('title', info_dict.get('id', 'video'))
723 if 'title' in info_dict:
724 # This can happen when we're just evaluating the playlist
725 title = info_dict['title']
726 matchtitle = self.params.get('matchtitle', False)
727 if matchtitle:
728 if not re.search(matchtitle, title, re.IGNORECASE):
729 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
730 rejecttitle = self.params.get('rejecttitle', False)
731 if rejecttitle:
732 if re.search(rejecttitle, title, re.IGNORECASE):
733 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
734 date = info_dict.get('upload_date')
735 if date is not None:
736 dateRange = self.params.get('daterange', DateRange())
737 if date not in dateRange:
738 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
739 view_count = info_dict.get('view_count')
740 if view_count is not None:
741 min_views = self.params.get('min_views')
742 if min_views is not None and view_count < min_views:
743 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
744 max_views = self.params.get('max_views')
745 if max_views is not None and view_count > max_views:
746 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
747 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
748 return 'Skipping "%s" because it is age restricted' % video_title
749 if self.in_download_archive(info_dict):
750 return '%s has already been recorded in archive' % video_title
751
752 if not incomplete:
753 match_filter = self.params.get('match_filter')
754 if match_filter is not None:
755 ret = match_filter(info_dict)
756 if ret is not None:
757 return ret
758
759 return None
760
761 @staticmethod
762 def add_extra_info(info_dict, extra_info):
763 '''Set the keys from extra_info in info dict if they are missing'''
764 for key, value in extra_info.items():
765 info_dict.setdefault(key, value)
766
767 def extract_info(self, url, download=True, ie_key=None, extra_info={},
768 process=True, force_generic_extractor=False):
769 '''
770 Returns a list with a dictionary for each video we find.
771 If 'download', also downloads the videos.
772 extra_info is a dict containing the extra values to add to each result
773 '''
774
775 if not ie_key and force_generic_extractor:
776 ie_key = 'Generic'
777
778 if ie_key:
779 ies = [self.get_info_extractor(ie_key)]
780 else:
781 ies = self._ies
782
783 for ie in ies:
784 if not ie.suitable(url):
785 continue
786
787 ie = self.get_info_extractor(ie.ie_key())
788 if not ie.working():
789 self.report_warning('The program functionality for this site has been marked as broken, '
790 'and will probably not work.')
791
792 try:
793 ie_result = ie.extract(url)
794 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
795 break
796 if isinstance(ie_result, list):
797 # Backwards compatibility: old IE result format
798 ie_result = {
799 '_type': 'compat_list',
800 'entries': ie_result,
801 }
802 self.add_default_extra_info(ie_result, ie, url)
803 if process:
804 return self.process_ie_result(ie_result, download, extra_info)
805 else:
806 return ie_result
807 except GeoRestrictedError as e:
808 msg = e.msg
809 if e.countries:
810 msg += '\nThis video is available in %s.' % ', '.join(
811 map(ISO3166Utils.short2full, e.countries))
812 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
813 self.report_error(msg)
814 break
815 except ExtractorError as e: # An error we somewhat expected
816 self.report_error(compat_str(e), e.format_traceback())
817 break
818 except MaxDownloadsReached:
819 raise
820 except Exception as e:
821 if self.params.get('ignoreerrors', False):
822 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
823 break
824 else:
825 raise
826 else:
827 self.report_error('no suitable InfoExtractor for URL %s' % url)
828
829 def add_default_extra_info(self, ie_result, ie, url):
830 self.add_extra_info(ie_result, {
831 'extractor': ie.IE_NAME,
832 'webpage_url': url,
833 'webpage_url_basename': url_basename(url),
834 'extractor_key': ie.ie_key(),
835 })
836
837 def process_ie_result(self, ie_result, download=True, extra_info={}):
838 """
839 Take the result of the ie(may be modified) and resolve all unresolved
840 references (URLs, playlist items).
841
842 It will also download the videos if 'download'.
843 Returns the resolved ie_result.
844 """
845 result_type = ie_result.get('_type', 'video')
846
847 if result_type in ('url', 'url_transparent'):
848 ie_result['url'] = sanitize_url(ie_result['url'])
849 extract_flat = self.params.get('extract_flat', False)
850 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
851 extract_flat is True):
852 if self.params.get('forcejson', False):
853 self.to_stdout(json.dumps(ie_result))
854 return ie_result
855
856 if result_type == 'video':
857 self.add_extra_info(ie_result, extra_info)
858 return self.process_video_result(ie_result, download=download)
859 elif result_type == 'url':
860 # We have to add extra_info to the results because it may be
861 # contained in a playlist
862 return self.extract_info(ie_result['url'],
863 download,
864 ie_key=ie_result.get('ie_key'),
865 extra_info=extra_info)
866 elif result_type == 'url_transparent':
867 # Use the information from the embedding page
868 info = self.extract_info(
869 ie_result['url'], ie_key=ie_result.get('ie_key'),
870 extra_info=extra_info, download=False, process=False)
871
872 # extract_info may return None when ignoreerrors is enabled and
873 # extraction failed with an error, don't crash and return early
874 # in this case
875 if not info:
876 return info
877
878 force_properties = dict(
879 (k, v) for k, v in ie_result.items() if v is not None)
880 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
881 if f in force_properties:
882 del force_properties[f]
883 new_result = info.copy()
884 new_result.update(force_properties)
885
886 # Extracted info may not be a video result (i.e.
887 # info.get('_type', 'video') != video) but rather an url or
888 # url_transparent. In such cases outer metadata (from ie_result)
889 # should be propagated to inner one (info). For this to happen
890 # _type of info should be overridden with url_transparent. This
891 # fixes issue from https://github.com/rg3/youtube-dl/pull/11163.
892 if new_result.get('_type') == 'url':
893 new_result['_type'] = 'url_transparent'
894
895 return self.process_ie_result(
896 new_result, download=download, extra_info=extra_info)
897 elif result_type in ('playlist', 'multi_video'):
898 # We process each entry in the playlist
899 playlist = ie_result.get('title') or ie_result.get('id')
900 self.to_screen('[download] Downloading playlist: %s' % playlist)
901
902 playlist_results = []
903
904 playliststart = self.params.get('playliststart', 1) - 1
905 playlistend = self.params.get('playlistend')
906 # For backwards compatibility, interpret -1 as whole list
907 if playlistend == -1:
908 playlistend = None
909
910 playlistitems_str = self.params.get('playlist_items')
911 playlistitems = None
912 if playlistitems_str is not None:
913 def iter_playlistitems(format):
914 for string_segment in format.split(','):
915 if '-' in string_segment:
916 start, end = string_segment.split('-')
917 for item in range(int(start), int(end) + 1):
918 yield int(item)
919 else:
920 yield int(string_segment)
921 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
922
923 ie_entries = ie_result['entries']
924
925 def make_playlistitems_entries(list_ie_entries):
926 num_entries = len(list_ie_entries)
927 return [
928 list_ie_entries[i - 1] for i in playlistitems
929 if -num_entries <= i - 1 < num_entries]
930
931 def report_download(num_entries):
932 self.to_screen(
933 '[%s] playlist %s: Downloading %d videos' %
934 (ie_result['extractor'], playlist, num_entries))
935
936 if isinstance(ie_entries, list):
937 n_all_entries = len(ie_entries)
938 if playlistitems:
939 entries = make_playlistitems_entries(ie_entries)
940 else:
941 entries = ie_entries[playliststart:playlistend]
942 n_entries = len(entries)
943 self.to_screen(
944 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
945 (ie_result['extractor'], playlist, n_all_entries, n_entries))
946 elif isinstance(ie_entries, PagedList):
947 if playlistitems:
948 entries = []
949 for item in playlistitems:
950 entries.extend(ie_entries.getslice(
951 item - 1, item
952 ))
953 else:
954 entries = ie_entries.getslice(
955 playliststart, playlistend)
956 n_entries = len(entries)
957 report_download(n_entries)
958 else: # iterable
959 if playlistitems:
960 entries = make_playlistitems_entries(list(itertools.islice(
961 ie_entries, 0, max(playlistitems))))
962 else:
963 entries = list(itertools.islice(
964 ie_entries, playliststart, playlistend))
965 n_entries = len(entries)
966 report_download(n_entries)
967
968 if self.params.get('playlistreverse', False):
969 entries = entries[::-1]
970
971 if self.params.get('playlistrandom', False):
972 random.shuffle(entries)
973
974 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
975
976 for i, entry in enumerate(entries, 1):
977 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
978 # This __x_forwarded_for_ip thing is a bit ugly but requires
979 # minimal changes
980 if x_forwarded_for:
981 entry['__x_forwarded_for_ip'] = x_forwarded_for
982 extra = {
983 'n_entries': n_entries,
984 'playlist': playlist,
985 'playlist_id': ie_result.get('id'),
986 'playlist_title': ie_result.get('title'),
987 'playlist_uploader': ie_result.get('uploader'),
988 'playlist_uploader_id': ie_result.get('uploader_id'),
989 'playlist_index': i + playliststart,
990 'extractor': ie_result['extractor'],
991 'webpage_url': ie_result['webpage_url'],
992 'webpage_url_basename': url_basename(ie_result['webpage_url']),
993 'extractor_key': ie_result['extractor_key'],
994 }
995
996 reason = self._match_entry(entry, incomplete=True)
997 if reason is not None:
998 self.to_screen('[download] ' + reason)
999 continue
1000
1001 entry_result = self.process_ie_result(entry,
1002 download=download,
1003 extra_info=extra)
1004 playlist_results.append(entry_result)
1005 ie_result['entries'] = playlist_results
1006 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1007 return ie_result
1008 elif result_type == 'compat_list':
1009 self.report_warning(
1010 'Extractor %s returned a compat_list result. '
1011 'It needs to be updated.' % ie_result.get('extractor'))
1012
1013 def _fixup(r):
1014 self.add_extra_info(
1015 r,
1016 {
1017 'extractor': ie_result['extractor'],
1018 'webpage_url': ie_result['webpage_url'],
1019 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1020 'extractor_key': ie_result['extractor_key'],
1021 }
1022 )
1023 return r
1024 ie_result['entries'] = [
1025 self.process_ie_result(_fixup(r), download, extra_info)
1026 for r in ie_result['entries']
1027 ]
1028 return ie_result
1029 else:
1030 raise Exception('Invalid result type: %s' % result_type)
1031
1032 def _build_format_filter(self, filter_spec):
1033 " Returns a function to filter the formats according to the filter_spec "
1034
1035 OPERATORS = {
1036 '<': operator.lt,
1037 '<=': operator.le,
1038 '>': operator.gt,
1039 '>=': operator.ge,
1040 '=': operator.eq,
1041 '!=': operator.ne,
1042 }
1043 operator_rex = re.compile(r'''(?x)\s*
1044 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)
1045 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1046 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
1047 $
1048 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1049 m = operator_rex.search(filter_spec)
1050 if m:
1051 try:
1052 comparison_value = int(m.group('value'))
1053 except ValueError:
1054 comparison_value = parse_filesize(m.group('value'))
1055 if comparison_value is None:
1056 comparison_value = parse_filesize(m.group('value') + 'B')
1057 if comparison_value is None:
1058 raise ValueError(
1059 'Invalid value %r in format specification %r' % (
1060 m.group('value'), filter_spec))
1061 op = OPERATORS[m.group('op')]
1062
1063 if not m:
1064 STR_OPERATORS = {
1065 '=': operator.eq,
1066 '!=': operator.ne,
1067 '^=': lambda attr, value: attr.startswith(value),
1068 '$=': lambda attr, value: attr.endswith(value),
1069 '*=': lambda attr, value: value in attr,
1070 }
1071 str_operator_rex = re.compile(r'''(?x)
1072 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
1073 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
1074 \s*(?P<value>[a-zA-Z0-9._-]+)
1075 \s*$
1076 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1077 m = str_operator_rex.search(filter_spec)
1078 if m:
1079 comparison_value = m.group('value')
1080 op = STR_OPERATORS[m.group('op')]
1081
1082 if not m:
1083 raise ValueError('Invalid filter specification %r' % filter_spec)
1084
1085 def _filter(f):
1086 actual_value = f.get(m.group('key'))
1087 if actual_value is None:
1088 return m.group('none_inclusive')
1089 return op(actual_value, comparison_value)
1090 return _filter
1091
1092 def _default_format_spec(self, info_dict, download=True):
1093
1094 def can_merge():
1095 merger = FFmpegMergerPP(self)
1096 return merger.available and merger.can_merge()
1097
1098 def prefer_best():
1099 if self.params.get('simulate', False):
1100 return False
1101 if not download:
1102 return False
1103 if self.params.get('outtmpl', DEFAULT_OUTTMPL) == '-':
1104 return True
1105 if info_dict.get('is_live'):
1106 return True
1107 if not can_merge():
1108 return True
1109 return False
1110
1111 req_format_list = ['bestvideo+bestaudio', 'best']
1112 if prefer_best():
1113 req_format_list.reverse()
1114 return '/'.join(req_format_list)
1115
1116 def build_format_selector(self, format_spec):
1117 def syntax_error(note, start):
1118 message = (
1119 'Invalid format specification: '
1120 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1121 return SyntaxError(message)
1122
1123 PICKFIRST = 'PICKFIRST'
1124 MERGE = 'MERGE'
1125 SINGLE = 'SINGLE'
1126 GROUP = 'GROUP'
1127 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1128
1129 def _parse_filter(tokens):
1130 filter_parts = []
1131 for type, string, start, _, _ in tokens:
1132 if type == tokenize.OP and string == ']':
1133 return ''.join(filter_parts)
1134 else:
1135 filter_parts.append(string)
1136
1137 def _remove_unused_ops(tokens):
1138 # Remove operators that we don't use and join them with the surrounding strings
1139 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1140 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1141 last_string, last_start, last_end, last_line = None, None, None, None
1142 for type, string, start, end, line in tokens:
1143 if type == tokenize.OP and string == '[':
1144 if last_string:
1145 yield tokenize.NAME, last_string, last_start, last_end, last_line
1146 last_string = None
1147 yield type, string, start, end, line
1148 # everything inside brackets will be handled by _parse_filter
1149 for type, string, start, end, line in tokens:
1150 yield type, string, start, end, line
1151 if type == tokenize.OP and string == ']':
1152 break
1153 elif type == tokenize.OP and string in ALLOWED_OPS:
1154 if last_string:
1155 yield tokenize.NAME, last_string, last_start, last_end, last_line
1156 last_string = None
1157 yield type, string, start, end, line
1158 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1159 if not last_string:
1160 last_string = string
1161 last_start = start
1162 last_end = end
1163 else:
1164 last_string += string
1165 if last_string:
1166 yield tokenize.NAME, last_string, last_start, last_end, last_line
1167
1168 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1169 selectors = []
1170 current_selector = None
1171 for type, string, start, _, _ in tokens:
1172 # ENCODING is only defined in python 3.x
1173 if type == getattr(tokenize, 'ENCODING', None):
1174 continue
1175 elif type in [tokenize.NAME, tokenize.NUMBER]:
1176 current_selector = FormatSelector(SINGLE, string, [])
1177 elif type == tokenize.OP:
1178 if string == ')':
1179 if not inside_group:
1180 # ')' will be handled by the parentheses group
1181 tokens.restore_last_token()
1182 break
1183 elif inside_merge and string in ['/', ',']:
1184 tokens.restore_last_token()
1185 break
1186 elif inside_choice and string == ',':
1187 tokens.restore_last_token()
1188 break
1189 elif string == ',':
1190 if not current_selector:
1191 raise syntax_error('"," must follow a format selector', start)
1192 selectors.append(current_selector)
1193 current_selector = None
1194 elif string == '/':
1195 if not current_selector:
1196 raise syntax_error('"/" must follow a format selector', start)
1197 first_choice = current_selector
1198 second_choice = _parse_format_selection(tokens, inside_choice=True)
1199 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1200 elif string == '[':
1201 if not current_selector:
1202 current_selector = FormatSelector(SINGLE, 'best', [])
1203 format_filter = _parse_filter(tokens)
1204 current_selector.filters.append(format_filter)
1205 elif string == '(':
1206 if current_selector:
1207 raise syntax_error('Unexpected "("', start)
1208 group = _parse_format_selection(tokens, inside_group=True)
1209 current_selector = FormatSelector(GROUP, group, [])
1210 elif string == '+':
1211 video_selector = current_selector
1212 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1213 if not video_selector or not audio_selector:
1214 raise syntax_error('"+" must be between two format selectors', start)
1215 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1216 else:
1217 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1218 elif type == tokenize.ENDMARKER:
1219 break
1220 if current_selector:
1221 selectors.append(current_selector)
1222 return selectors
1223
1224 def _build_selector_function(selector):
1225 if isinstance(selector, list):
1226 fs = [_build_selector_function(s) for s in selector]
1227
1228 def selector_function(ctx):
1229 for f in fs:
1230 for format in f(ctx):
1231 yield format
1232 return selector_function
1233 elif selector.type == GROUP:
1234 selector_function = _build_selector_function(selector.selector)
1235 elif selector.type == PICKFIRST:
1236 fs = [_build_selector_function(s) for s in selector.selector]
1237
1238 def selector_function(ctx):
1239 for f in fs:
1240 picked_formats = list(f(ctx))
1241 if picked_formats:
1242 return picked_formats
1243 return []
1244 elif selector.type == SINGLE:
1245 format_spec = selector.selector
1246
1247 def selector_function(ctx):
1248 formats = list(ctx['formats'])
1249 if not formats:
1250 return
1251 if format_spec == 'all':
1252 for f in formats:
1253 yield f
1254 elif format_spec in ['best', 'worst', None]:
1255 format_idx = 0 if format_spec == 'worst' else -1
1256 audiovideo_formats = [
1257 f for f in formats
1258 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1259 if audiovideo_formats:
1260 yield audiovideo_formats[format_idx]
1261 # for extractors with incomplete formats (audio only (soundcloud)
1262 # or video only (imgur)) we will fallback to best/worst
1263 # {video,audio}-only format
1264 elif ctx['incomplete_formats']:
1265 yield formats[format_idx]
1266 elif format_spec == 'bestaudio':
1267 audio_formats = [
1268 f for f in formats
1269 if f.get('vcodec') == 'none']
1270 if audio_formats:
1271 yield audio_formats[-1]
1272 elif format_spec == 'worstaudio':
1273 audio_formats = [
1274 f for f in formats
1275 if f.get('vcodec') == 'none']
1276 if audio_formats:
1277 yield audio_formats[0]
1278 elif format_spec == 'bestvideo':
1279 video_formats = [
1280 f for f in formats
1281 if f.get('acodec') == 'none']
1282 if video_formats:
1283 yield video_formats[-1]
1284 elif format_spec == 'worstvideo':
1285 video_formats = [
1286 f for f in formats
1287 if f.get('acodec') == 'none']
1288 if video_formats:
1289 yield video_formats[0]
1290 else:
1291 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1292 if format_spec in extensions:
1293 filter_f = lambda f: f['ext'] == format_spec
1294 else:
1295 filter_f = lambda f: f['format_id'] == format_spec
1296 matches = list(filter(filter_f, formats))
1297 if matches:
1298 yield matches[-1]
1299 elif selector.type == MERGE:
1300 def _merge(formats_info):
1301 format_1, format_2 = [f['format_id'] for f in formats_info]
1302 # The first format must contain the video and the
1303 # second the audio
1304 if formats_info[0].get('vcodec') == 'none':
1305 self.report_error('The first format must '
1306 'contain the video, try using '
1307 '"-f %s+%s"' % (format_2, format_1))
1308 return
1309 # Formats must be opposite (video+audio)
1310 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1311 self.report_error(
1312 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1313 % (format_1, format_2))
1314 return
1315 output_ext = (
1316 formats_info[0]['ext']
1317 if self.params.get('merge_output_format') is None
1318 else self.params['merge_output_format'])
1319 return {
1320 'requested_formats': formats_info,
1321 'format': '%s+%s' % (formats_info[0].get('format'),
1322 formats_info[1].get('format')),
1323 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1324 formats_info[1].get('format_id')),
1325 'width': formats_info[0].get('width'),
1326 'height': formats_info[0].get('height'),
1327 'resolution': formats_info[0].get('resolution'),
1328 'fps': formats_info[0].get('fps'),
1329 'vcodec': formats_info[0].get('vcodec'),
1330 'vbr': formats_info[0].get('vbr'),
1331 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1332 'acodec': formats_info[1].get('acodec'),
1333 'abr': formats_info[1].get('abr'),
1334 'ext': output_ext,
1335 }
1336 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1337
1338 def selector_function(ctx):
1339 for pair in itertools.product(
1340 video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
1341 yield _merge(pair)
1342
1343 filters = [self._build_format_filter(f) for f in selector.filters]
1344
1345 def final_selector(ctx):
1346 ctx_copy = copy.deepcopy(ctx)
1347 for _filter in filters:
1348 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1349 return selector_function(ctx_copy)
1350 return final_selector
1351
1352 stream = io.BytesIO(format_spec.encode('utf-8'))
1353 try:
1354 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1355 except tokenize.TokenError:
1356 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1357
1358 class TokenIterator(object):
1359 def __init__(self, tokens):
1360 self.tokens = tokens
1361 self.counter = 0
1362
1363 def __iter__(self):
1364 return self
1365
1366 def __next__(self):
1367 if self.counter >= len(self.tokens):
1368 raise StopIteration()
1369 value = self.tokens[self.counter]
1370 self.counter += 1
1371 return value
1372
1373 next = __next__
1374
1375 def restore_last_token(self):
1376 self.counter -= 1
1377
1378 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1379 return _build_selector_function(parsed_selector)
1380
1381 def _calc_headers(self, info_dict):
1382 res = std_headers.copy()
1383
1384 add_headers = info_dict.get('http_headers')
1385 if add_headers:
1386 res.update(add_headers)
1387
1388 cookies = self._calc_cookies(info_dict)
1389 if cookies:
1390 res['Cookie'] = cookies
1391
1392 if 'X-Forwarded-For' not in res:
1393 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1394 if x_forwarded_for_ip:
1395 res['X-Forwarded-For'] = x_forwarded_for_ip
1396
1397 return res
1398
1399 def _calc_cookies(self, info_dict):
1400 pr = sanitized_Request(info_dict['url'])
1401 self.cookiejar.add_cookie_header(pr)
1402 return pr.get_header('Cookie')
1403
1404 def process_video_result(self, info_dict, download=True):
1405 assert info_dict.get('_type', 'video') == 'video'
1406
1407 if 'id' not in info_dict:
1408 raise ExtractorError('Missing "id" field in extractor result')
1409 if 'title' not in info_dict:
1410 raise ExtractorError('Missing "title" field in extractor result')
1411
1412 def report_force_conversion(field, field_not, conversion):
1413 self.report_warning(
1414 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
1415 % (field, field_not, conversion))
1416
1417 def sanitize_string_field(info, string_field):
1418 field = info.get(string_field)
1419 if field is None or isinstance(field, compat_str):
1420 return
1421 report_force_conversion(string_field, 'a string', 'string')
1422 info[string_field] = compat_str(field)
1423
1424 def sanitize_numeric_fields(info):
1425 for numeric_field in self._NUMERIC_FIELDS:
1426 field = info.get(numeric_field)
1427 if field is None or isinstance(field, compat_numeric_types):
1428 continue
1429 report_force_conversion(numeric_field, 'numeric', 'int')
1430 info[numeric_field] = int_or_none(field)
1431
1432 sanitize_string_field(info_dict, 'id')
1433 sanitize_numeric_fields(info_dict)
1434
1435 if 'playlist' not in info_dict:
1436 # It isn't part of a playlist
1437 info_dict['playlist'] = None
1438 info_dict['playlist_index'] = None
1439
1440 thumbnails = info_dict.get('thumbnails')
1441 if thumbnails is None:
1442 thumbnail = info_dict.get('thumbnail')
1443 if thumbnail:
1444 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1445 if thumbnails:
1446 thumbnails.sort(key=lambda t: (
1447 t.get('preference') if t.get('preference') is not None else -1,
1448 t.get('width') if t.get('width') is not None else -1,
1449 t.get('height') if t.get('height') is not None else -1,
1450 t.get('id') if t.get('id') is not None else '', t.get('url')))
1451 for i, t in enumerate(thumbnails):
1452 t['url'] = sanitize_url(t['url'])
1453 if t.get('width') and t.get('height'):
1454 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1455 if t.get('id') is None:
1456 t['id'] = '%d' % i
1457
1458 if self.params.get('list_thumbnails'):
1459 self.list_thumbnails(info_dict)
1460 return
1461
1462 thumbnail = info_dict.get('thumbnail')
1463 if thumbnail:
1464 info_dict['thumbnail'] = sanitize_url(thumbnail)
1465 elif thumbnails:
1466 info_dict['thumbnail'] = thumbnails[-1]['url']
1467
1468 if 'display_id' not in info_dict and 'id' in info_dict:
1469 info_dict['display_id'] = info_dict['id']
1470
1471 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1472 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1473 # see http://bugs.python.org/issue1646728)
1474 try:
1475 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1476 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1477 except (ValueError, OverflowError, OSError):
1478 pass
1479
1480 # Auto generate title fields corresponding to the *_number fields when missing
1481 # in order to always have clean titles. This is very common for TV series.
1482 for field in ('chapter', 'season', 'episode'):
1483 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1484 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1485
1486 for cc_kind in ('subtitles', 'automatic_captions'):
1487 cc = info_dict.get(cc_kind)
1488 if cc:
1489 for _, subtitle in cc.items():
1490 for subtitle_format in subtitle:
1491 if subtitle_format.get('url'):
1492 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1493 if subtitle_format.get('ext') is None:
1494 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1495
1496 automatic_captions = info_dict.get('automatic_captions')
1497 subtitles = info_dict.get('subtitles')
1498
1499 if self.params.get('listsubtitles', False):
1500 if 'automatic_captions' in info_dict:
1501 self.list_subtitles(
1502 info_dict['id'], automatic_captions, 'automatic captions')
1503 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1504 return
1505
1506 info_dict['requested_subtitles'] = self.process_subtitles(
1507 info_dict['id'], subtitles, automatic_captions)
1508
1509 # We now pick which formats have to be downloaded
1510 if info_dict.get('formats') is None:
1511 # There's only one format available
1512 formats = [info_dict]
1513 else:
1514 formats = info_dict['formats']
1515
1516 if not formats:
1517 raise ExtractorError('No video formats found!')
1518
1519 def is_wellformed(f):
1520 url = f.get('url')
1521 if not url:
1522 self.report_warning(
1523 '"url" field is missing or empty - skipping format, '
1524 'there is an error in extractor')
1525 return False
1526 if isinstance(url, bytes):
1527 sanitize_string_field(f, 'url')
1528 return True
1529
1530 # Filter out malformed formats for better extraction robustness
1531 formats = list(filter(is_wellformed, formats))
1532
1533 formats_dict = {}
1534
1535 # We check that all the formats have the format and format_id fields
1536 for i, format in enumerate(formats):
1537 sanitize_string_field(format, 'format_id')
1538 sanitize_numeric_fields(format)
1539 format['url'] = sanitize_url(format['url'])
1540 if not format.get('format_id'):
1541 format['format_id'] = compat_str(i)
1542 else:
1543 # Sanitize format_id from characters used in format selector expression
1544 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
1545 format_id = format['format_id']
1546 if format_id not in formats_dict:
1547 formats_dict[format_id] = []
1548 formats_dict[format_id].append(format)
1549
1550 # Make sure all formats have unique format_id
1551 for format_id, ambiguous_formats in formats_dict.items():
1552 if len(ambiguous_formats) > 1:
1553 for i, format in enumerate(ambiguous_formats):
1554 format['format_id'] = '%s-%d' % (format_id, i)
1555
1556 for i, format in enumerate(formats):
1557 if format.get('format') is None:
1558 format['format'] = '{id} - {res}{note}'.format(
1559 id=format['format_id'],
1560 res=self.format_resolution(format),
1561 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1562 )
1563 # Automatically determine file extension if missing
1564 if format.get('ext') is None:
1565 format['ext'] = determine_ext(format['url']).lower()
1566 # Automatically determine protocol if missing (useful for format
1567 # selection purposes)
1568 if format.get('protocol') is None:
1569 format['protocol'] = determine_protocol(format)
1570 # Add HTTP headers, so that external programs can use them from the
1571 # json output
1572 full_format_info = info_dict.copy()
1573 full_format_info.update(format)
1574 format['http_headers'] = self._calc_headers(full_format_info)
1575 # Remove private housekeeping stuff
1576 if '__x_forwarded_for_ip' in info_dict:
1577 del info_dict['__x_forwarded_for_ip']
1578
1579 # TODO Central sorting goes here
1580
1581 if formats[0] is not info_dict:
1582 # only set the 'formats' fields if the original info_dict list them
1583 # otherwise we end up with a circular reference, the first (and unique)
1584 # element in the 'formats' field in info_dict is info_dict itself,
1585 # which can't be exported to json
1586 info_dict['formats'] = formats
1587 if self.params.get('listformats'):
1588 self.list_formats(info_dict)
1589 return
1590
1591 req_format = self.params.get('format')
1592 if req_format is None:
1593 req_format = self._default_format_spec(info_dict, download=download)
1594 if self.params.get('verbose'):
1595 self.to_stdout('[debug] Default format spec: %s' % req_format)
1596
1597 format_selector = self.build_format_selector(req_format)
1598
1599 # While in format selection we may need to have an access to the original
1600 # format set in order to calculate some metrics or do some processing.
1601 # For now we need to be able to guess whether original formats provided
1602 # by extractor are incomplete or not (i.e. whether extractor provides only
1603 # video-only or audio-only formats) for proper formats selection for
1604 # extractors with such incomplete formats (see
1605 # https://github.com/rg3/youtube-dl/pull/5556).
1606 # Since formats may be filtered during format selection and may not match
1607 # the original formats the results may be incorrect. Thus original formats
1608 # or pre-calculated metrics should be passed to format selection routines
1609 # as well.
1610 # We will pass a context object containing all necessary additional data
1611 # instead of just formats.
1612 # This fixes incorrect format selection issue (see
1613 # https://github.com/rg3/youtube-dl/issues/10083).
1614 incomplete_formats = (
1615 # All formats are video-only or
1616 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) or
1617 # all formats are audio-only
1618 all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1619
1620 ctx = {
1621 'formats': formats,
1622 'incomplete_formats': incomplete_formats,
1623 }
1624
1625 formats_to_download = list(format_selector(ctx))
1626 if not formats_to_download:
1627 raise ExtractorError('requested format not available',
1628 expected=True)
1629
1630 if download:
1631 if len(formats_to_download) > 1:
1632 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1633 for format in formats_to_download:
1634 new_info = dict(info_dict)
1635 new_info.update(format)
1636 self.process_info(new_info)
1637 # We update the info dict with the best quality format (backwards compatibility)
1638 info_dict.update(formats_to_download[-1])
1639 return info_dict
1640
1641 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1642 """Select the requested subtitles and their format"""
1643 available_subs = {}
1644 if normal_subtitles and self.params.get('writesubtitles'):
1645 available_subs.update(normal_subtitles)
1646 if automatic_captions and self.params.get('writeautomaticsub'):
1647 for lang, cap_info in automatic_captions.items():
1648 if lang not in available_subs:
1649 available_subs[lang] = cap_info
1650
1651 if (not self.params.get('writesubtitles') and not
1652 self.params.get('writeautomaticsub') or not
1653 available_subs):
1654 return None
1655
1656 if self.params.get('allsubtitles', False):
1657 requested_langs = available_subs.keys()
1658 else:
1659 if self.params.get('subtitleslangs', False):
1660 requested_langs = self.params.get('subtitleslangs')
1661 elif 'en' in available_subs:
1662 requested_langs = ['en']
1663 else:
1664 requested_langs = [list(available_subs.keys())[0]]
1665
1666 formats_query = self.params.get('subtitlesformat', 'best')
1667 formats_preference = formats_query.split('/') if formats_query else []
1668 subs = {}
1669 for lang in requested_langs:
1670 formats = available_subs.get(lang)
1671 if formats is None:
1672 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1673 continue
1674 for ext in formats_preference:
1675 if ext == 'best':
1676 f = formats[-1]
1677 break
1678 matches = list(filter(lambda f: f['ext'] == ext, formats))
1679 if matches:
1680 f = matches[-1]
1681 break
1682 else:
1683 f = formats[-1]
1684 self.report_warning(
1685 'No subtitle format found matching "%s" for language %s, '
1686 'using %s' % (formats_query, lang, f['ext']))
1687 subs[lang] = f
1688 return subs
1689
1690 def process_info(self, info_dict):
1691 """Process a single resolved IE result."""
1692
1693 assert info_dict.get('_type', 'video') == 'video'
1694
1695 max_downloads = self.params.get('max_downloads')
1696 if max_downloads is not None:
1697 if self._num_downloads >= int(max_downloads):
1698 raise MaxDownloadsReached()
1699
1700 info_dict['fulltitle'] = info_dict['title']
1701 if len(info_dict['title']) > 200:
1702 info_dict['title'] = info_dict['title'][:197] + '...'
1703
1704 if 'format' not in info_dict:
1705 info_dict['format'] = info_dict['ext']
1706
1707 reason = self._match_entry(info_dict, incomplete=False)
1708 if reason is not None:
1709 self.to_screen('[download] ' + reason)
1710 return
1711
1712 self._num_downloads += 1
1713
1714 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1715
1716 # Forced printings
1717 if self.params.get('forcetitle', False):
1718 self.to_stdout(info_dict['fulltitle'])
1719 if self.params.get('forceid', False):
1720 self.to_stdout(info_dict['id'])
1721 if self.params.get('forceurl', False):
1722 if info_dict.get('requested_formats') is not None:
1723 for f in info_dict['requested_formats']:
1724 self.to_stdout(f['url'] + f.get('play_path', ''))
1725 else:
1726 # For RTMP URLs, also include the playpath
1727 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1728 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1729 self.to_stdout(info_dict['thumbnail'])
1730 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1731 self.to_stdout(info_dict['description'])
1732 if self.params.get('forcefilename', False) and filename is not None:
1733 self.to_stdout(filename)
1734 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1735 self.to_stdout(formatSeconds(info_dict['duration']))
1736 if self.params.get('forceformat', False):
1737 self.to_stdout(info_dict['format'])
1738 if self.params.get('forcejson', False):
1739 self.to_stdout(json.dumps(info_dict))
1740
1741 # Do nothing else if in simulate mode
1742 if self.params.get('simulate', False):
1743 return
1744
1745 if filename is None:
1746 return
1747
1748 def ensure_dir_exists(path):
1749 try:
1750 dn = os.path.dirname(path)
1751 if dn and not os.path.exists(dn):
1752 os.makedirs(dn)
1753 return True
1754 except (OSError, IOError) as err:
1755 self.report_error('unable to create directory ' + error_to_compat_str(err))
1756 return False
1757
1758 if not ensure_dir_exists(sanitize_path(encodeFilename(filename))):
1759 return
1760
1761 if self.params.get('writedescription', False):
1762 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1763 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1764 self.to_screen('[info] Video description is already present')
1765 elif info_dict.get('description') is None:
1766 self.report_warning('There\'s no description to write.')
1767 else:
1768 try:
1769 self.to_screen('[info] Writing video description to: ' + descfn)
1770 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1771 descfile.write(info_dict['description'])
1772 except (OSError, IOError):
1773 self.report_error('Cannot write description file ' + descfn)
1774 return
1775
1776 if self.params.get('writeannotations', False):
1777 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1778 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1779 self.to_screen('[info] Video annotations are already present')
1780 else:
1781 try:
1782 self.to_screen('[info] Writing video annotations to: ' + annofn)
1783 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1784 annofile.write(info_dict['annotations'])
1785 except (KeyError, TypeError):
1786 self.report_warning('There are no annotations to write.')
1787 except (OSError, IOError):
1788 self.report_error('Cannot write annotations file: ' + annofn)
1789 return
1790
1791 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1792 self.params.get('writeautomaticsub')])
1793
1794 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1795 # subtitles download errors are already managed as troubles in relevant IE
1796 # that way it will silently go on when used with unsupporting IE
1797 subtitles = info_dict['requested_subtitles']
1798 ie = self.get_info_extractor(info_dict['extractor_key'])
1799 for sub_lang, sub_info in subtitles.items():
1800 sub_format = sub_info['ext']
1801 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1802 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1803 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
1804 else:
1805 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1806 if sub_info.get('data') is not None:
1807 try:
1808 # Use newline='' to prevent conversion of newline characters
1809 # See https://github.com/rg3/youtube-dl/issues/10268
1810 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
1811 subfile.write(sub_info['data'])
1812 except (OSError, IOError):
1813 self.report_error('Cannot write subtitles file ' + sub_filename)
1814 return
1815 else:
1816 try:
1817 sub_data = ie._request_webpage(
1818 sub_info['url'], info_dict['id'], note=False).read()
1819 with io.open(encodeFilename(sub_filename), 'wb') as subfile:
1820 subfile.write(sub_data)
1821 except (ExtractorError, IOError, OSError, ValueError) as err:
1822 self.report_warning('Unable to download subtitle for "%s": %s' %
1823 (sub_lang, error_to_compat_str(err)))
1824 continue
1825
1826 if self.params.get('writeinfojson', False):
1827 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1828 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1829 self.to_screen('[info] Video description metadata is already present')
1830 else:
1831 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1832 try:
1833 write_json_file(self.filter_requested_info(info_dict), infofn)
1834 except (OSError, IOError):
1835 self.report_error('Cannot write metadata to JSON file ' + infofn)
1836 return
1837
1838 self._write_thumbnails(info_dict, filename)
1839
1840 if not self.params.get('skip_download', False):
1841 try:
1842 def dl(name, info):
1843 fd = get_suitable_downloader(info, self.params)(self, self.params)
1844 for ph in self._progress_hooks:
1845 fd.add_progress_hook(ph)
1846 if self.params.get('verbose'):
1847 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1848 return fd.download(name, info)
1849
1850 if info_dict.get('requested_formats') is not None:
1851 downloaded = []
1852 success = True
1853 merger = FFmpegMergerPP(self)
1854 if not merger.available:
1855 postprocessors = []
1856 self.report_warning('You have requested multiple '
1857 'formats but ffmpeg or avconv are not installed.'
1858 ' The formats won\'t be merged.')
1859 else:
1860 postprocessors = [merger]
1861
1862 def compatible_formats(formats):
1863 video, audio = formats
1864 # Check extension
1865 video_ext, audio_ext = video.get('ext'), audio.get('ext')
1866 if video_ext and audio_ext:
1867 COMPATIBLE_EXTS = (
1868 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'),
1869 ('webm')
1870 )
1871 for exts in COMPATIBLE_EXTS:
1872 if video_ext in exts and audio_ext in exts:
1873 return True
1874 # TODO: Check acodec/vcodec
1875 return False
1876
1877 filename_real_ext = os.path.splitext(filename)[1][1:]
1878 filename_wo_ext = (
1879 os.path.splitext(filename)[0]
1880 if filename_real_ext == info_dict['ext']
1881 else filename)
1882 requested_formats = info_dict['requested_formats']
1883 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1884 info_dict['ext'] = 'mkv'
1885 self.report_warning(
1886 'Requested formats are incompatible for merge and will be merged into mkv.')
1887 # Ensure filename always has a correct extension for successful merge
1888 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1889 if os.path.exists(encodeFilename(filename)):
1890 self.to_screen(
1891 '[download] %s has already been downloaded and '
1892 'merged' % filename)
1893 else:
1894 for f in requested_formats:
1895 new_info = dict(info_dict)
1896 new_info.update(f)
1897 fname = prepend_extension(
1898 self.prepare_filename(new_info),
1899 'f%s' % f['format_id'], new_info['ext'])
1900 if not ensure_dir_exists(fname):
1901 return
1902 downloaded.append(fname)
1903 partial_success = dl(fname, new_info)
1904 success = success and partial_success
1905 info_dict['__postprocessors'] = postprocessors
1906 info_dict['__files_to_merge'] = downloaded
1907 else:
1908 # Just a single file
1909 success = dl(filename, info_dict)
1910 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1911 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1912 return
1913 except (OSError, IOError) as err:
1914 raise UnavailableVideoError(err)
1915 except (ContentTooShortError, ) as err:
1916 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1917 return
1918
1919 if success and filename != '-':
1920 # Fixup content
1921 fixup_policy = self.params.get('fixup')
1922 if fixup_policy is None:
1923 fixup_policy = 'detect_or_warn'
1924
1925 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1926
1927 stretched_ratio = info_dict.get('stretched_ratio')
1928 if stretched_ratio is not None and stretched_ratio != 1:
1929 if fixup_policy == 'warn':
1930 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1931 info_dict['id'], stretched_ratio))
1932 elif fixup_policy == 'detect_or_warn':
1933 stretched_pp = FFmpegFixupStretchedPP(self)
1934 if stretched_pp.available:
1935 info_dict.setdefault('__postprocessors', [])
1936 info_dict['__postprocessors'].append(stretched_pp)
1937 else:
1938 self.report_warning(
1939 '%s: Non-uniform pixel ratio (%s). %s'
1940 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1941 else:
1942 assert fixup_policy in ('ignore', 'never')
1943
1944 if (info_dict.get('requested_formats') is None and
1945 info_dict.get('container') == 'm4a_dash'):
1946 if fixup_policy == 'warn':
1947 self.report_warning(
1948 '%s: writing DASH m4a. '
1949 'Only some players support this container.'
1950 % info_dict['id'])
1951 elif fixup_policy == 'detect_or_warn':
1952 fixup_pp = FFmpegFixupM4aPP(self)
1953 if fixup_pp.available:
1954 info_dict.setdefault('__postprocessors', [])
1955 info_dict['__postprocessors'].append(fixup_pp)
1956 else:
1957 self.report_warning(
1958 '%s: writing DASH m4a. '
1959 'Only some players support this container. %s'
1960 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1961 else:
1962 assert fixup_policy in ('ignore', 'never')
1963
1964 if (info_dict.get('protocol') == 'm3u8_native' or
1965 info_dict.get('protocol') == 'm3u8' and
1966 self.params.get('hls_prefer_native')):
1967 if fixup_policy == 'warn':
1968 self.report_warning('%s: malformed AAC bitstream detected.' % (
1969 info_dict['id']))
1970 elif fixup_policy == 'detect_or_warn':
1971 fixup_pp = FFmpegFixupM3u8PP(self)
1972 if fixup_pp.available:
1973 info_dict.setdefault('__postprocessors', [])
1974 info_dict['__postprocessors'].append(fixup_pp)
1975 else:
1976 self.report_warning(
1977 '%s: malformed AAC bitstream detected. %s'
1978 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1979 else:
1980 assert fixup_policy in ('ignore', 'never')
1981
1982 try:
1983 self.post_process(filename, info_dict)
1984 except (PostProcessingError) as err:
1985 self.report_error('postprocessing: %s' % str(err))
1986 return
1987 self.record_download_archive(info_dict)
1988
1989 def download(self, url_list):
1990 """Download a given list of URLs."""
1991 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1992 if (len(url_list) > 1 and
1993 outtmpl != '-' and
1994 '%' not in outtmpl and
1995 self.params.get('max_downloads') != 1):
1996 raise SameFileError(outtmpl)
1997
1998 for url in url_list:
1999 try:
2000 # It also downloads the videos
2001 res = self.extract_info(
2002 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2003 except UnavailableVideoError:
2004 self.report_error('unable to download video')
2005 except MaxDownloadsReached:
2006 self.to_screen('[info] Maximum number of downloaded files reached.')
2007 raise
2008 else:
2009 if self.params.get('dump_single_json', False):
2010 self.to_stdout(json.dumps(res))
2011
2012 return self._download_retcode
2013
2014 def download_with_info_file(self, info_filename):
2015 with contextlib.closing(fileinput.FileInput(
2016 [info_filename], mode='r',
2017 openhook=fileinput.hook_encoded('utf-8'))) as f:
2018 # FileInput doesn't have a read method, we can't call json.load
2019 info = self.filter_requested_info(json.loads('\n'.join(f)))
2020 try:
2021 self.process_ie_result(info, download=True)
2022 except DownloadError:
2023 webpage_url = info.get('webpage_url')
2024 if webpage_url is not None:
2025 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2026 return self.download([webpage_url])
2027 else:
2028 raise
2029 return self._download_retcode
2030
2031 @staticmethod
2032 def filter_requested_info(info_dict):
2033 return dict(
2034 (k, v) for k, v in info_dict.items()
2035 if k not in ['requested_formats', 'requested_subtitles'])
2036
2037 def post_process(self, filename, ie_info):
2038 """Run all the postprocessors on the given file."""
2039 info = dict(ie_info)
2040 info['filepath'] = filename
2041 pps_chain = []
2042 if ie_info.get('__postprocessors') is not None:
2043 pps_chain.extend(ie_info['__postprocessors'])
2044 pps_chain.extend(self._pps)
2045 for pp in pps_chain:
2046 files_to_delete = []
2047 try:
2048 files_to_delete, info = pp.run(info)
2049 except PostProcessingError as e:
2050 self.report_error(e.msg)
2051 if files_to_delete and not self.params.get('keepvideo', False):
2052 for old_filename in files_to_delete:
2053 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2054 try:
2055 os.remove(encodeFilename(old_filename))
2056 except (IOError, OSError):
2057 self.report_warning('Unable to remove downloaded original file')
2058
2059 def _make_archive_id(self, info_dict):
2060 # Future-proof against any change in case
2061 # and backwards compatibility with prior versions
2062 extractor = info_dict.get('extractor_key')
2063 if extractor is None:
2064 if 'id' in info_dict:
2065 extractor = info_dict.get('ie_key') # key in a playlist
2066 if extractor is None:
2067 return None # Incomplete video information
2068 return extractor.lower() + ' ' + info_dict['id']
2069
2070 def in_download_archive(self, info_dict):
2071 fn = self.params.get('download_archive')
2072 if fn is None:
2073 return False
2074
2075 vid_id = self._make_archive_id(info_dict)
2076 if vid_id is None:
2077 return False # Incomplete video information
2078
2079 try:
2080 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
2081 for line in archive_file:
2082 if line.strip() == vid_id:
2083 return True
2084 except IOError as ioe:
2085 if ioe.errno != errno.ENOENT:
2086 raise
2087 return False
2088
2089 def record_download_archive(self, info_dict):
2090 fn = self.params.get('download_archive')
2091 if fn is None:
2092 return
2093 vid_id = self._make_archive_id(info_dict)
2094 assert vid_id
2095 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2096 archive_file.write(vid_id + '\n')
2097
2098 @staticmethod
2099 def format_resolution(format, default='unknown'):
2100 if format.get('vcodec') == 'none':
2101 return 'audio only'
2102 if format.get('resolution') is not None:
2103 return format['resolution']
2104 if format.get('height') is not None:
2105 if format.get('width') is not None:
2106 res = '%sx%s' % (format['width'], format['height'])
2107 else:
2108 res = '%sp' % format['height']
2109 elif format.get('width') is not None:
2110 res = '%dx?' % format['width']
2111 else:
2112 res = default
2113 return res
2114
2115 def _format_note(self, fdict):
2116 res = ''
2117 if fdict.get('ext') in ['f4f', 'f4m']:
2118 res += '(unsupported) '
2119 if fdict.get('language'):
2120 if res:
2121 res += ' '
2122 res += '[%s] ' % fdict['language']
2123 if fdict.get('format_note') is not None:
2124 res += fdict['format_note'] + ' '
2125 if fdict.get('tbr') is not None:
2126 res += '%4dk ' % fdict['tbr']
2127 if fdict.get('container') is not None:
2128 if res:
2129 res += ', '
2130 res += '%s container' % fdict['container']
2131 if (fdict.get('vcodec') is not None and
2132 fdict.get('vcodec') != 'none'):
2133 if res:
2134 res += ', '
2135 res += fdict['vcodec']
2136 if fdict.get('vbr') is not None:
2137 res += '@'
2138 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2139 res += 'video@'
2140 if fdict.get('vbr') is not None:
2141 res += '%4dk' % fdict['vbr']
2142 if fdict.get('fps') is not None:
2143 if res:
2144 res += ', '
2145 res += '%sfps' % fdict['fps']
2146 if fdict.get('acodec') is not None:
2147 if res:
2148 res += ', '
2149 if fdict['acodec'] == 'none':
2150 res += 'video only'
2151 else:
2152 res += '%-5s' % fdict['acodec']
2153 elif fdict.get('abr') is not None:
2154 if res:
2155 res += ', '
2156 res += 'audio'
2157 if fdict.get('abr') is not None:
2158 res += '@%3dk' % fdict['abr']
2159 if fdict.get('asr') is not None:
2160 res += ' (%5dHz)' % fdict['asr']
2161 if fdict.get('filesize') is not None:
2162 if res:
2163 res += ', '
2164 res += format_bytes(fdict['filesize'])
2165 elif fdict.get('filesize_approx') is not None:
2166 if res:
2167 res += ', '
2168 res += '~' + format_bytes(fdict['filesize_approx'])
2169 return res
2170
2171 def list_formats(self, info_dict):
2172 formats = info_dict.get('formats', [info_dict])
2173 table = [
2174 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
2175 for f in formats
2176 if f.get('preference') is None or f['preference'] >= -1000]
2177 if len(formats) > 1:
2178 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
2179
2180 header_line = ['format code', 'extension', 'resolution', 'note']
2181 self.to_screen(
2182 '[info] Available formats for %s:\n%s' %
2183 (info_dict['id'], render_table(header_line, table)))
2184
2185 def list_thumbnails(self, info_dict):
2186 thumbnails = info_dict.get('thumbnails')
2187 if not thumbnails:
2188 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
2189 return
2190
2191 self.to_screen(
2192 '[info] Thumbnails for %s:' % info_dict['id'])
2193 self.to_screen(render_table(
2194 ['ID', 'width', 'height', 'URL'],
2195 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
2196
2197 def list_subtitles(self, video_id, subtitles, name='subtitles'):
2198 if not subtitles:
2199 self.to_screen('%s has no %s' % (video_id, name))
2200 return
2201 self.to_screen(
2202 'Available %s for %s:' % (name, video_id))
2203 self.to_screen(render_table(
2204 ['Language', 'formats'],
2205 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
2206 for lang, formats in subtitles.items()]))
2207
2208 def urlopen(self, req):
2209 """ Start an HTTP download """
2210 if isinstance(req, compat_basestring):
2211 req = sanitized_Request(req)
2212 return self._opener.open(req, timeout=self._socket_timeout)
2213
2214 def print_debug_header(self):
2215 if not self.params.get('verbose'):
2216 return
2217
2218 if type('') is not compat_str:
2219 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
2220 self.report_warning(
2221 'Your Python is broken! Update to a newer and supported version')
2222
2223 stdout_encoding = getattr(
2224 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2225 encoding_str = (
2226 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2227 locale.getpreferredencoding(),
2228 sys.getfilesystemencoding(),
2229 stdout_encoding,
2230 self.get_encoding()))
2231 write_string(encoding_str, encoding=None)
2232
2233 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
2234 if _LAZY_LOADER:
2235 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2236 try:
2237 sp = subprocess.Popen(
2238 ['git', 'rev-parse', '--short', 'HEAD'],
2239 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2240 cwd=os.path.dirname(os.path.abspath(__file__)))
2241 out, err = sp.communicate()
2242 out = out.decode().strip()
2243 if re.match('[0-9a-f]+', out):
2244 self._write_string('[debug] Git HEAD: ' + out + '\n')
2245 except Exception:
2246 try:
2247 sys.exc_clear()
2248 except Exception:
2249 pass
2250
2251 def python_implementation():
2252 impl_name = platform.python_implementation()
2253 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
2254 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
2255 return impl_name
2256
2257 self._write_string('[debug] Python version %s (%s) - %s\n' % (
2258 platform.python_version(), python_implementation(),
2259 platform_name()))
2260
2261 exe_versions = FFmpegPostProcessor.get_versions(self)
2262 exe_versions['rtmpdump'] = rtmpdump_version()
2263 exe_versions['phantomjs'] = PhantomJSwrapper._version()
2264 exe_str = ', '.join(
2265 '%s %s' % (exe, v)
2266 for exe, v in sorted(exe_versions.items())
2267 if v
2268 )
2269 if not exe_str:
2270 exe_str = 'none'
2271 self._write_string('[debug] exe versions: %s\n' % exe_str)
2272
2273 proxy_map = {}
2274 for handler in self._opener.handlers:
2275 if hasattr(handler, 'proxies'):
2276 proxy_map.update(handler.proxies)
2277 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2278
2279 if self.params.get('call_home', False):
2280 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2281 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2282 latest_version = self.urlopen(
2283 'https://yt-dl.org/latest/version').read().decode('utf-8')
2284 if version_tuple(latest_version) > version_tuple(__version__):
2285 self.report_warning(
2286 'You are using an outdated version (newest version: %s)! '
2287 'See https://yt-dl.org/update if you need help updating.' %
2288 latest_version)
2289
2290 def _setup_opener(self):
2291 timeout_val = self.params.get('socket_timeout')
2292 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2293
2294 opts_cookiefile = self.params.get('cookiefile')
2295 opts_proxy = self.params.get('proxy')
2296
2297 if opts_cookiefile is None:
2298 self.cookiejar = compat_cookiejar.CookieJar()
2299 else:
2300 opts_cookiefile = expand_path(opts_cookiefile)
2301 self.cookiejar = YoutubeDLCookieJar(opts_cookiefile)
2302 if os.access(opts_cookiefile, os.R_OK):
2303 self.cookiejar.load(ignore_discard=True, ignore_expires=True)
2304
2305 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2306 if opts_proxy is not None:
2307 if opts_proxy == '':
2308 proxies = {}
2309 else:
2310 proxies = {'http': opts_proxy, 'https': opts_proxy}
2311 else:
2312 proxies = compat_urllib_request.getproxies()
2313 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2314 if 'http' in proxies and 'https' not in proxies:
2315 proxies['https'] = proxies['http']
2316 proxy_handler = PerRequestProxyHandler(proxies)
2317
2318 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2319 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2320 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2321 data_handler = compat_urllib_request_DataHandler()
2322
2323 # When passing our own FileHandler instance, build_opener won't add the
2324 # default FileHandler and allows us to disable the file protocol, which
2325 # can be used for malicious purposes (see
2326 # https://github.com/rg3/youtube-dl/issues/8227)
2327 file_handler = compat_urllib_request.FileHandler()
2328
2329 def file_open(*args, **kwargs):
2330 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2331 file_handler.file_open = file_open
2332
2333 opener = compat_urllib_request.build_opener(
2334 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2335
2336 # Delete the default user-agent header, which would otherwise apply in
2337 # cases where our custom HTTP handler doesn't come into play
2338 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2339 opener.addheaders = []
2340 self._opener = opener
2341
2342 def encode(self, s):
2343 if isinstance(s, bytes):
2344 return s # Already encoded
2345
2346 try:
2347 return s.encode(self.get_encoding())
2348 except UnicodeEncodeError as err:
2349 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2350 raise
2351
2352 def get_encoding(self):
2353 encoding = self.params.get('encoding')
2354 if encoding is None:
2355 encoding = preferredencoding()
2356 return encoding
2357
2358 def _write_thumbnails(self, info_dict, filename):
2359 if self.params.get('writethumbnail', False):
2360 thumbnails = info_dict.get('thumbnails')
2361 if thumbnails:
2362 thumbnails = [thumbnails[-1]]
2363 elif self.params.get('write_all_thumbnails', False):
2364 thumbnails = info_dict.get('thumbnails')
2365 else:
2366 return
2367
2368 if not thumbnails:
2369 # No thumbnails present, so return immediately
2370 return
2371
2372 for t in thumbnails:
2373 thumb_ext = determine_ext(t['url'], 'jpg')
2374 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2375 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2376 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2377
2378 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2379 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2380 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2381 else:
2382 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2383 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2384 try:
2385 uf = self.urlopen(t['url'])
2386 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2387 shutil.copyfileobj(uf, thumbf)
2388 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2389 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2390 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2391 self.report_warning('Unable to download thumbnail "%s": %s' %
2392 (t['url'], error_to_compat_str(err)))