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