4 from __future__ import absolute_import, unicode_literals
29 from string import ascii_letters
34 compat_get_terminal_size,
40 compat_tokenize_tokenize,
42 compat_urllib_request,
43 compat_urllib_request_DataHandler,
71 PerRequestProxyHandler,
76 register_socks_protocols,
86 UnavailableVideoError,
91 YoutubeDLCookieProcessor,
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 (
102 FFmpegFixupStretchedPP,
107 from .version import __version__
109 if compat_os_name == 'nt':
113 class YoutubeDL(object):
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.
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.
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".
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
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
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
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
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.
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,
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
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)
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.
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
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
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)
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)
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.
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.
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
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
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',
330 _download_retcode = None
331 _num_downloads = None
334 def __init__(self, params=None, auto_init=True):
335 """Create a FileDownloader object with the given options."""
339 self._ies_instances = {}
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
348 'nocheckcertificate': False,
350 self.params.update(params)
351 self.cache = Cache(self)
353 def check_deprecated(param, option, suggestion):
354 if self.params.get(param) is not None:
356 '%s is deprecated. Use %s instead.' % (option, suggestion))
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']
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"')
368 if params.get('bidi_workaround', False):
371 master, slave = pty.openpty()
372 width = compat_get_terminal_size().columns
376 width_args = ['-w', str(width)]
378 stdin=subprocess.PIPE,
380 stderr=self._err_file)
382 self._output_process = subprocess.Popen(
383 ['bidiv'] + width_args, **sp_kwargs
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 $PATH.')
395 if (sys.platform != 'win32' and
396 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
397 not params.get('restrictfilenames', False)):
398 # Unicode filesystem API will throw errors (#1474, #13027)
400 'Assuming --restrict-filenames since file system encoding '
401 'cannot encode all characters. '
402 'Set the LC_ALL environment variable to fix this.')
403 self.params['restrictfilenames'] = True
405 if isinstance(params.get('outtmpl'), bytes):
407 'Parameter outtmpl is bytes, but should be a unicode string. '
408 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
413 self.print_debug_header()
414 self.add_default_info_extractors()
416 for pp_def_raw in self.params.get('postprocessors', []):
417 pp_class = get_postprocessor(pp_def_raw['key'])
418 pp_def = dict(pp_def_raw)
420 pp = pp_class(self, **compat_kwargs(pp_def))
421 self.add_post_processor(pp)
423 for ph in self.params.get('progress_hooks', []):
424 self.add_progress_hook(ph)
426 register_socks_protocols()
428 def warn_if_short_id(self, argv):
429 # short YouTube ID starting with dash?
431 i for i, a in enumerate(argv)
432 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
436 [a for i, a in enumerate(argv) if i not in idxs] +
437 ['--'] + [argv[i] for i in idxs]
440 'Long argument string detected. '
441 'Use -- to separate parameters and URLs, like this:\n%s\n' %
442 args_to_str(correct_argv))
444 def add_info_extractor(self, ie):
445 """Add an InfoExtractor object to the end of the list."""
447 if not isinstance(ie, type):
448 self._ies_instances[ie.ie_key()] = ie
449 ie.set_downloader(self)
451 def get_info_extractor(self, ie_key):
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.
457 ie = self._ies_instances.get(ie_key)
459 ie = get_info_extractor(ie_key)()
460 self.add_info_extractor(ie)
463 def add_default_info_extractors(self):
465 Add the InfoExtractors returned by gen_extractors to the end of the list
467 for ie in gen_extractor_classes():
468 self.add_info_extractor(ie)
470 def add_post_processor(self, pp):
471 """Add a PostProcessor object to the end of the chain."""
473 pp.set_downloader(self)
475 def add_progress_hook(self, ph):
476 """Add the progress hook (currently only for the file downloader)"""
477 self._progress_hooks.append(ph)
479 def _bidi_workaround(self, message):
480 if not hasattr(self, '_output_channel'):
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')]
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)
496 def _write_string(self, s, out=None):
497 write_string(s, out=out, encoding=self.params.get('encoding'))
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
508 self._write_string(output, self._screen_file)
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)
516 message = self._bidi_workaround(message)
517 output = message + '\n'
518 self._write_string(output, self._err_file)
520 def to_console_title(self, message):
521 if not self.params.get('consoletitle', False):
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)
531 def save_console_title(self):
532 if not self.params.get('consoletitle', False):
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)
538 def restore_console_title(self):
539 if not self.params.get('consoletitle', False):
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)
546 self.save_console_title()
549 def __exit__(self, *args):
550 self.restore_console_title()
552 if self.params.get('cookiefile') is not None:
553 self.cookiejar.save()
555 def trouble(self, message=None, tb=None):
556 """Determine action to take when a download problem appears.
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.
562 tb, if given, is additional traceback information.
564 if message is not None:
565 self.to_stderr(message)
566 if self.params.get('verbose'):
568 if sys.exc_info()[0]: # if .trouble has been called from an except block
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())
574 tb_data = traceback.format_list(traceback.extract_stack())
575 tb = ''.join(tb_data)
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
581 exc_info = sys.exc_info()
582 raise DownloadError(message, exc_info)
583 self._download_retcode = 1
585 def report_warning(self, message):
587 Print the message to stderr, it will be prefixed with 'WARNING:'
588 If stderr is a tty file the 'WARNING:' will be colored
590 if self.params.get('logger') is not None:
591 self.params['logger'].warning(message)
593 if self.params.get('no_warnings'):
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'
598 _msg_header = 'WARNING:'
599 warning_message = '%s %s' % (_msg_header, message)
600 self.to_stderr(warning_message)
602 def report_error(self, message, tb=None):
604 Do the same as trouble, but prefixes the message with 'ERROR:', colored
605 in red if stderr is a tty file.
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'
610 _msg_header = 'ERROR:'
611 error_message = '%s %s' % (_msg_header, message)
612 self.trouble(error_message, tb)
614 def report_file_already_downloaded(self, file_name):
615 """Report file has already been fully downloaded."""
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')
621 def prepare_filename(self, info_dict):
622 """Generate the output filename."""
624 template_dict = dict(info_dict)
626 template_dict['epoch'] = int(time.time())
627 autonumber_size = self.params.get('autonumber_size')
628 if autonumber_size is None:
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']
639 sanitize = lambda k, v: sanitize_filename(
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)
648 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
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,
656 FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
657 mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
660 FIELD_SIZE_COMPAT_RE,
661 r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
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
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
684 FORMAT_RE.format(numeric_field),
685 r'%({0})s'.format(numeric_field), outtmpl)
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))
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
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()) + ')')
710 def _match_entry(self, info_dict, incomplete):
711 """ Returns None iff the file should be downloaded """
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)
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)
723 if re.search(rejecttitle, title, re.IGNORECASE):
724 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
725 date = info_dict.get('upload_date')
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
744 match_filter = self.params.get('match_filter')
745 if match_filter is not None:
746 ret = match_filter(info_dict)
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)
758 def extract_info(self, url, download=True, ie_key=None, extra_info={},
759 process=True, force_generic_extractor=False):
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
766 if not ie_key and force_generic_extractor:
770 ies = [self.get_info_extractor(ie_key)]
775 if not ie.suitable(url):
778 ie = self.get_info_extractor(ie.ie_key())
780 self.report_warning('The program functionality for this site has been marked as broken, '
781 'and will probably not work.')
784 ie_result = ie.extract(url)
785 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
787 if isinstance(ie_result, list):
788 # Backwards compatibility: old IE result format
790 '_type': 'compat_list',
791 'entries': ie_result,
793 self.add_default_extra_info(ie_result, ie, url)
795 return self.process_ie_result(ie_result, download, extra_info)
798 except GeoRestrictedError as e:
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)
806 except ExtractorError as e: # An error we somewhat expected
807 self.report_error(compat_str(e), e.format_traceback())
809 except MaxDownloadsReached:
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()))
818 self.report_error('no suitable InfoExtractor for URL %s' % url)
820 def add_default_extra_info(self, ie_result, ie, url):
821 self.add_extra_info(ie_result, {
822 'extractor': ie.IE_NAME,
824 'webpage_url_basename': url_basename(url),
825 'extractor_key': ie.ie_key(),
828 def process_ie_result(self, ie_result, download=True, extra_info={}):
830 Take the result of the ie(may be modified) and resolve all unresolved
831 references (URLs, playlist items).
833 It will also download the videos if 'download'.
834 Returns the resolved ie_result.
836 result_type = ie_result.get('_type', 'video')
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))
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'],
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)
863 # extract_info may return None when ignoreerrors is enabled and
864 # extraction failed with an error, don't crash and return early
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)
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'
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)
893 playlist_results = []
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:
901 playlistitems_str = self.params.get('playlist_items')
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):
911 yield int(string_segment)
912 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
914 ie_entries = ie_result['entries']
916 def make_playlistitems_entries(list_ie_entries):
917 num_entries = len(list_ie_entries)
919 list_ie_entries[i - 1] for i in playlistitems
920 if -num_entries <= i - 1 < num_entries]
922 def report_download(num_entries):
924 '[%s] playlist %s: Downloading %d videos' %
925 (ie_result['extractor'], playlist, num_entries))
927 if isinstance(ie_entries, list):
928 n_all_entries = len(ie_entries)
930 entries = make_playlistitems_entries(ie_entries)
932 entries = ie_entries[playliststart:playlistend]
933 n_entries = len(entries)
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):
940 for item in playlistitems:
941 entries.extend(ie_entries.getslice(
945 entries = ie_entries.getslice(
946 playliststart, playlistend)
947 n_entries = len(entries)
948 report_download(n_entries)
951 entries = make_playlistitems_entries(list(ie_entries))
953 entries = list(itertools.islice(
954 ie_entries, playliststart, playlistend))
955 n_entries = len(entries)
956 report_download(n_entries)
958 if self.params.get('playlistreverse', False):
959 entries = entries[::-1]
961 if self.params.get('playlistrandom', False):
962 random.shuffle(entries)
964 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
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
971 entry['__x_forwarded_for_ip'] = x_forwarded_for
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'],
984 reason = self._match_entry(entry, incomplete=True)
985 if reason is not None:
986 self.to_screen('[download] ' + reason)
989 entry_result = self.process_ie_result(entry,
992 playlist_results.append(entry_result)
993 ie_result['entries'] = playlist_results
994 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
996 elif result_type == 'compat_list':
998 'Extractor %s returned a compat_list result. '
999 'It needs to be updated.' % ie_result.get('extractor'))
1002 self.add_extra_info(
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'],
1012 ie_result['entries'] = [
1013 self.process_ie_result(_fixup(r), download, extra_info)
1014 for r in ie_result['entries']
1018 raise Exception('Invalid result type: %s' % result_type)
1020 def _build_format_filter(self, filter_spec):
1021 " Returns a function to filter the formats according to the filter_spec "
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]?)?)
1036 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1037 m = operator_rex.search(filter_spec)
1040 comparison_value = int(m.group('value'))
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:
1047 'Invalid value %r in format specification %r' % (
1048 m.group('value'), filter_spec))
1049 op = OPERATORS[m.group('op')]
1055 '^=': lambda attr, value: attr.startswith(value),
1056 '$=': lambda attr, value: attr.endswith(value),
1057 '*=': lambda attr, value: value in attr,
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._-]+)
1064 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1065 m = str_operator_rex.search(filter_spec)
1067 comparison_value = m.group('value')
1068 op = STR_OPERATORS[m.group('op')]
1071 raise ValueError('Invalid filter specification %r' % filter_spec)
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)
1080 def _default_format_spec(self, info_dict, download=True):
1081 req_format_list = []
1083 def can_have_partial_formats():
1084 if self.params.get('simulate', False):
1088 if self.params.get('outtmpl', DEFAULT_OUTTMPL) == '-':
1090 if info_dict.get('is_live'):
1092 merger = FFmpegMergerPP(self)
1093 return merger.available and merger.can_merge()
1094 if can_have_partial_formats():
1095 req_format_list.append('bestvideo+bestaudio')
1096 req_format_list.append('best')
1097 return '/'.join(req_format_list)
1099 def build_format_selector(self, format_spec):
1100 def syntax_error(note, start):
1102 'Invalid format specification: '
1103 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1104 return SyntaxError(message)
1106 PICKFIRST = 'PICKFIRST'
1110 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1112 def _parse_filter(tokens):
1114 for type, string, start, _, _ in tokens:
1115 if type == tokenize.OP and string == ']':
1116 return ''.join(filter_parts)
1118 filter_parts.append(string)
1120 def _remove_unused_ops(tokens):
1121 # Remove operators that we don't use and join them with the surrounding strings
1122 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1123 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1124 last_string, last_start, last_end, last_line = None, None, None, None
1125 for type, string, start, end, line in tokens:
1126 if type == tokenize.OP and string == '[':
1128 yield tokenize.NAME, last_string, last_start, last_end, last_line
1130 yield type, string, start, end, line
1131 # everything inside brackets will be handled by _parse_filter
1132 for type, string, start, end, line in tokens:
1133 yield type, string, start, end, line
1134 if type == tokenize.OP and string == ']':
1136 elif type == tokenize.OP and string in ALLOWED_OPS:
1138 yield tokenize.NAME, last_string, last_start, last_end, last_line
1140 yield type, string, start, end, line
1141 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1143 last_string = string
1147 last_string += string
1149 yield tokenize.NAME, last_string, last_start, last_end, last_line
1151 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1153 current_selector = None
1154 for type, string, start, _, _ in tokens:
1155 # ENCODING is only defined in python 3.x
1156 if type == getattr(tokenize, 'ENCODING', None):
1158 elif type in [tokenize.NAME, tokenize.NUMBER]:
1159 current_selector = FormatSelector(SINGLE, string, [])
1160 elif type == tokenize.OP:
1162 if not inside_group:
1163 # ')' will be handled by the parentheses group
1164 tokens.restore_last_token()
1166 elif inside_merge and string in ['/', ',']:
1167 tokens.restore_last_token()
1169 elif inside_choice and string == ',':
1170 tokens.restore_last_token()
1173 if not current_selector:
1174 raise syntax_error('"," must follow a format selector', start)
1175 selectors.append(current_selector)
1176 current_selector = None
1178 if not current_selector:
1179 raise syntax_error('"/" must follow a format selector', start)
1180 first_choice = current_selector
1181 second_choice = _parse_format_selection(tokens, inside_choice=True)
1182 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1184 if not current_selector:
1185 current_selector = FormatSelector(SINGLE, 'best', [])
1186 format_filter = _parse_filter(tokens)
1187 current_selector.filters.append(format_filter)
1189 if current_selector:
1190 raise syntax_error('Unexpected "("', start)
1191 group = _parse_format_selection(tokens, inside_group=True)
1192 current_selector = FormatSelector(GROUP, group, [])
1194 video_selector = current_selector
1195 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1196 if not video_selector or not audio_selector:
1197 raise syntax_error('"+" must be between two format selectors', start)
1198 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1200 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1201 elif type == tokenize.ENDMARKER:
1203 if current_selector:
1204 selectors.append(current_selector)
1207 def _build_selector_function(selector):
1208 if isinstance(selector, list):
1209 fs = [_build_selector_function(s) for s in selector]
1211 def selector_function(ctx):
1213 for format in f(ctx):
1215 return selector_function
1216 elif selector.type == GROUP:
1217 selector_function = _build_selector_function(selector.selector)
1218 elif selector.type == PICKFIRST:
1219 fs = [_build_selector_function(s) for s in selector.selector]
1221 def selector_function(ctx):
1223 picked_formats = list(f(ctx))
1225 return picked_formats
1227 elif selector.type == SINGLE:
1228 format_spec = selector.selector
1230 def selector_function(ctx):
1231 formats = list(ctx['formats'])
1234 if format_spec == 'all':
1237 elif format_spec in ['best', 'worst', None]:
1238 format_idx = 0 if format_spec == 'worst' else -1
1239 audiovideo_formats = [
1241 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1242 if audiovideo_formats:
1243 yield audiovideo_formats[format_idx]
1244 # for extractors with incomplete formats (audio only (soundcloud)
1245 # or video only (imgur)) we will fallback to best/worst
1246 # {video,audio}-only format
1247 elif ctx['incomplete_formats']:
1248 yield formats[format_idx]
1249 elif format_spec == 'bestaudio':
1252 if f.get('vcodec') == 'none']
1254 yield audio_formats[-1]
1255 elif format_spec == 'worstaudio':
1258 if f.get('vcodec') == 'none']
1260 yield audio_formats[0]
1261 elif format_spec == 'bestvideo':
1264 if f.get('acodec') == 'none']
1266 yield video_formats[-1]
1267 elif format_spec == 'worstvideo':
1270 if f.get('acodec') == 'none']
1272 yield video_formats[0]
1274 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1275 if format_spec in extensions:
1276 filter_f = lambda f: f['ext'] == format_spec
1278 filter_f = lambda f: f['format_id'] == format_spec
1279 matches = list(filter(filter_f, formats))
1282 elif selector.type == MERGE:
1283 def _merge(formats_info):
1284 format_1, format_2 = [f['format_id'] for f in formats_info]
1285 # The first format must contain the video and the
1287 if formats_info[0].get('vcodec') == 'none':
1288 self.report_error('The first format must '
1289 'contain the video, try using '
1290 '"-f %s+%s"' % (format_2, format_1))
1292 # Formats must be opposite (video+audio)
1293 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1295 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1296 % (format_1, format_2))
1299 formats_info[0]['ext']
1300 if self.params.get('merge_output_format') is None
1301 else self.params['merge_output_format'])
1303 'requested_formats': formats_info,
1304 'format': '%s+%s' % (formats_info[0].get('format'),
1305 formats_info[1].get('format')),
1306 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1307 formats_info[1].get('format_id')),
1308 'width': formats_info[0].get('width'),
1309 'height': formats_info[0].get('height'),
1310 'resolution': formats_info[0].get('resolution'),
1311 'fps': formats_info[0].get('fps'),
1312 'vcodec': formats_info[0].get('vcodec'),
1313 'vbr': formats_info[0].get('vbr'),
1314 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1315 'acodec': formats_info[1].get('acodec'),
1316 'abr': formats_info[1].get('abr'),
1319 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1321 def selector_function(ctx):
1322 for pair in itertools.product(
1323 video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
1326 filters = [self._build_format_filter(f) for f in selector.filters]
1328 def final_selector(ctx):
1329 ctx_copy = copy.deepcopy(ctx)
1330 for _filter in filters:
1331 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1332 return selector_function(ctx_copy)
1333 return final_selector
1335 stream = io.BytesIO(format_spec.encode('utf-8'))
1337 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1338 except tokenize.TokenError:
1339 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1341 class TokenIterator(object):
1342 def __init__(self, tokens):
1343 self.tokens = tokens
1350 if self.counter >= len(self.tokens):
1351 raise StopIteration()
1352 value = self.tokens[self.counter]
1358 def restore_last_token(self):
1361 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1362 return _build_selector_function(parsed_selector)
1364 def _calc_headers(self, info_dict):
1365 res = std_headers.copy()
1367 add_headers = info_dict.get('http_headers')
1369 res.update(add_headers)
1371 cookies = self._calc_cookies(info_dict)
1373 res['Cookie'] = cookies
1375 if 'X-Forwarded-For' not in res:
1376 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1377 if x_forwarded_for_ip:
1378 res['X-Forwarded-For'] = x_forwarded_for_ip
1382 def _calc_cookies(self, info_dict):
1383 pr = sanitized_Request(info_dict['url'])
1384 self.cookiejar.add_cookie_header(pr)
1385 return pr.get_header('Cookie')
1387 def process_video_result(self, info_dict, download=True):
1388 assert info_dict.get('_type', 'video') == 'video'
1390 if 'id' not in info_dict:
1391 raise ExtractorError('Missing "id" field in extractor result')
1392 if 'title' not in info_dict:
1393 raise ExtractorError('Missing "title" field in extractor result')
1395 def report_force_conversion(field, field_not, conversion):
1396 self.report_warning(
1397 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
1398 % (field, field_not, conversion))
1400 def sanitize_string_field(info, string_field):
1401 field = info.get(string_field)
1402 if field is None or isinstance(field, compat_str):
1404 report_force_conversion(string_field, 'a string', 'string')
1405 info[string_field] = compat_str(field)
1407 def sanitize_numeric_fields(info):
1408 for numeric_field in self._NUMERIC_FIELDS:
1409 field = info.get(numeric_field)
1410 if field is None or isinstance(field, compat_numeric_types):
1412 report_force_conversion(numeric_field, 'numeric', 'int')
1413 info[numeric_field] = int_or_none(field)
1415 sanitize_string_field(info_dict, 'id')
1416 sanitize_numeric_fields(info_dict)
1418 if 'playlist' not in info_dict:
1419 # It isn't part of a playlist
1420 info_dict['playlist'] = None
1421 info_dict['playlist_index'] = None
1423 thumbnails = info_dict.get('thumbnails')
1424 if thumbnails is None:
1425 thumbnail = info_dict.get('thumbnail')
1427 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1429 thumbnails.sort(key=lambda t: (
1430 t.get('preference') if t.get('preference') is not None else -1,
1431 t.get('width') if t.get('width') is not None else -1,
1432 t.get('height') if t.get('height') is not None else -1,
1433 t.get('id') if t.get('id') is not None else '', t.get('url')))
1434 for i, t in enumerate(thumbnails):
1435 t['url'] = sanitize_url(t['url'])
1436 if t.get('width') and t.get('height'):
1437 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1438 if t.get('id') is None:
1441 if self.params.get('list_thumbnails'):
1442 self.list_thumbnails(info_dict)
1445 thumbnail = info_dict.get('thumbnail')
1447 info_dict['thumbnail'] = sanitize_url(thumbnail)
1449 info_dict['thumbnail'] = thumbnails[-1]['url']
1451 if 'display_id' not in info_dict and 'id' in info_dict:
1452 info_dict['display_id'] = info_dict['id']
1454 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1455 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1456 # see http://bugs.python.org/issue1646728)
1458 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1459 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1460 except (ValueError, OverflowError, OSError):
1463 # Auto generate title fields corresponding to the *_number fields when missing
1464 # in order to always have clean titles. This is very common for TV series.
1465 for field in ('chapter', 'season', 'episode'):
1466 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1467 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1469 subtitles = info_dict.get('subtitles')
1471 for _, subtitle in subtitles.items():
1472 for subtitle_format in subtitle:
1473 if subtitle_format.get('url'):
1474 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1475 if subtitle_format.get('ext') is None:
1476 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1478 if self.params.get('listsubtitles', False):
1479 if 'automatic_captions' in info_dict:
1480 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1481 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1483 info_dict['requested_subtitles'] = self.process_subtitles(
1484 info_dict['id'], subtitles,
1485 info_dict.get('automatic_captions'))
1487 # We now pick which formats have to be downloaded
1488 if info_dict.get('formats') is None:
1489 # There's only one format available
1490 formats = [info_dict]
1492 formats = info_dict['formats']
1495 raise ExtractorError('No video formats found!')
1497 def is_wellformed(f):
1500 self.report_warning(
1501 '"url" field is missing or empty - skipping format, '
1502 'there is an error in extractor')
1504 if isinstance(url, bytes):
1505 sanitize_string_field(f, 'url')
1508 # Filter out malformed formats for better extraction robustness
1509 formats = list(filter(is_wellformed, formats))
1513 # We check that all the formats have the format and format_id fields
1514 for i, format in enumerate(formats):
1515 sanitize_string_field(format, 'format_id')
1516 sanitize_numeric_fields(format)
1517 format['url'] = sanitize_url(format['url'])
1518 if not format.get('format_id'):
1519 format['format_id'] = compat_str(i)
1521 # Sanitize format_id from characters used in format selector expression
1522 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
1523 format_id = format['format_id']
1524 if format_id not in formats_dict:
1525 formats_dict[format_id] = []
1526 formats_dict[format_id].append(format)
1528 # Make sure all formats have unique format_id
1529 for format_id, ambiguous_formats in formats_dict.items():
1530 if len(ambiguous_formats) > 1:
1531 for i, format in enumerate(ambiguous_formats):
1532 format['format_id'] = '%s-%d' % (format_id, i)
1534 for i, format in enumerate(formats):
1535 if format.get('format') is None:
1536 format['format'] = '{id} - {res}{note}'.format(
1537 id=format['format_id'],
1538 res=self.format_resolution(format),
1539 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1541 # Automatically determine file extension if missing
1542 if format.get('ext') is None:
1543 format['ext'] = determine_ext(format['url']).lower()
1544 # Automatically determine protocol if missing (useful for format
1545 # selection purposes)
1546 if format.get('protocol') is None:
1547 format['protocol'] = determine_protocol(format)
1548 # Add HTTP headers, so that external programs can use them from the
1550 full_format_info = info_dict.copy()
1551 full_format_info.update(format)
1552 format['http_headers'] = self._calc_headers(full_format_info)
1553 # Remove private housekeeping stuff
1554 if '__x_forwarded_for_ip' in info_dict:
1555 del info_dict['__x_forwarded_for_ip']
1557 # TODO Central sorting goes here
1559 if formats[0] is not info_dict:
1560 # only set the 'formats' fields if the original info_dict list them
1561 # otherwise we end up with a circular reference, the first (and unique)
1562 # element in the 'formats' field in info_dict is info_dict itself,
1563 # which can't be exported to json
1564 info_dict['formats'] = formats
1565 if self.params.get('listformats'):
1566 self.list_formats(info_dict)
1569 req_format = self.params.get('format')
1570 if req_format is None:
1571 req_format = self._default_format_spec(info_dict, download=download)
1572 if self.params.get('verbose'):
1573 self.to_stdout('[debug] Default format spec: %s' % req_format)
1575 format_selector = self.build_format_selector(req_format)
1577 # While in format selection we may need to have an access to the original
1578 # format set in order to calculate some metrics or do some processing.
1579 # For now we need to be able to guess whether original formats provided
1580 # by extractor are incomplete or not (i.e. whether extractor provides only
1581 # video-only or audio-only formats) for proper formats selection for
1582 # extractors with such incomplete formats (see
1583 # https://github.com/rg3/youtube-dl/pull/5556).
1584 # Since formats may be filtered during format selection and may not match
1585 # the original formats the results may be incorrect. Thus original formats
1586 # or pre-calculated metrics should be passed to format selection routines
1588 # We will pass a context object containing all necessary additional data
1589 # instead of just formats.
1590 # This fixes incorrect format selection issue (see
1591 # https://github.com/rg3/youtube-dl/issues/10083).
1592 incomplete_formats = (
1593 # All formats are video-only or
1594 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) or
1595 # all formats are audio-only
1596 all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1600 'incomplete_formats': incomplete_formats,
1603 formats_to_download = list(format_selector(ctx))
1604 if not formats_to_download:
1605 raise ExtractorError('requested format not available',
1609 if len(formats_to_download) > 1:
1610 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1611 for format in formats_to_download:
1612 new_info = dict(info_dict)
1613 new_info.update(format)
1614 self.process_info(new_info)
1615 # We update the info dict with the best quality format (backwards compatibility)
1616 info_dict.update(formats_to_download[-1])
1619 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1620 """Select the requested subtitles and their format"""
1622 if normal_subtitles and self.params.get('writesubtitles'):
1623 available_subs.update(normal_subtitles)
1624 if automatic_captions and self.params.get('writeautomaticsub'):
1625 for lang, cap_info in automatic_captions.items():
1626 if lang not in available_subs:
1627 available_subs[lang] = cap_info
1629 if (not self.params.get('writesubtitles') and not
1630 self.params.get('writeautomaticsub') or not
1634 if self.params.get('allsubtitles', False):
1635 requested_langs = available_subs.keys()
1637 if self.params.get('subtitleslangs', False):
1638 requested_langs = self.params.get('subtitleslangs')
1639 elif 'en' in available_subs:
1640 requested_langs = ['en']
1642 requested_langs = [list(available_subs.keys())[0]]
1644 formats_query = self.params.get('subtitlesformat', 'best')
1645 formats_preference = formats_query.split('/') if formats_query else []
1647 for lang in requested_langs:
1648 formats = available_subs.get(lang)
1650 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1652 for ext in formats_preference:
1656 matches = list(filter(lambda f: f['ext'] == ext, formats))
1662 self.report_warning(
1663 'No subtitle format found matching "%s" for language %s, '
1664 'using %s' % (formats_query, lang, f['ext']))
1668 def process_info(self, info_dict):
1669 """Process a single resolved IE result."""
1671 assert info_dict.get('_type', 'video') == 'video'
1673 max_downloads = self.params.get('max_downloads')
1674 if max_downloads is not None:
1675 if self._num_downloads >= int(max_downloads):
1676 raise MaxDownloadsReached()
1678 info_dict['fulltitle'] = info_dict['title']
1679 if len(info_dict['title']) > 200:
1680 info_dict['title'] = info_dict['title'][:197] + '...'
1682 if 'format' not in info_dict:
1683 info_dict['format'] = info_dict['ext']
1685 reason = self._match_entry(info_dict, incomplete=False)
1686 if reason is not None:
1687 self.to_screen('[download] ' + reason)
1690 self._num_downloads += 1
1692 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1695 if self.params.get('forcetitle', False):
1696 self.to_stdout(info_dict['fulltitle'])
1697 if self.params.get('forceid', False):
1698 self.to_stdout(info_dict['id'])
1699 if self.params.get('forceurl', False):
1700 if info_dict.get('requested_formats') is not None:
1701 for f in info_dict['requested_formats']:
1702 self.to_stdout(f['url'] + f.get('play_path', ''))
1704 # For RTMP URLs, also include the playpath
1705 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1706 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1707 self.to_stdout(info_dict['thumbnail'])
1708 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1709 self.to_stdout(info_dict['description'])
1710 if self.params.get('forcefilename', False) and filename is not None:
1711 self.to_stdout(filename)
1712 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1713 self.to_stdout(formatSeconds(info_dict['duration']))
1714 if self.params.get('forceformat', False):
1715 self.to_stdout(info_dict['format'])
1716 if self.params.get('forcejson', False):
1717 self.to_stdout(json.dumps(info_dict))
1719 # Do nothing else if in simulate mode
1720 if self.params.get('simulate', False):
1723 if filename is None:
1726 def ensure_dir_exists(path):
1728 dn = os.path.dirname(path)
1729 if dn and not os.path.exists(dn):
1732 except (OSError, IOError) as err:
1733 self.report_error('unable to create directory ' + error_to_compat_str(err))
1736 if not ensure_dir_exists(sanitize_path(encodeFilename(filename))):
1739 if self.params.get('writedescription', False):
1740 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1741 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1742 self.to_screen('[info] Video description is already present')
1743 elif info_dict.get('description') is None:
1744 self.report_warning('There\'s no description to write.')
1747 self.to_screen('[info] Writing video description to: ' + descfn)
1748 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1749 descfile.write(info_dict['description'])
1750 except (OSError, IOError):
1751 self.report_error('Cannot write description file ' + descfn)
1754 if self.params.get('writeannotations', False):
1755 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1756 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1757 self.to_screen('[info] Video annotations are already present')
1760 self.to_screen('[info] Writing video annotations to: ' + annofn)
1761 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1762 annofile.write(info_dict['annotations'])
1763 except (KeyError, TypeError):
1764 self.report_warning('There are no annotations to write.')
1765 except (OSError, IOError):
1766 self.report_error('Cannot write annotations file: ' + annofn)
1769 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1770 self.params.get('writeautomaticsub')])
1772 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1773 # subtitles download errors are already managed as troubles in relevant IE
1774 # that way it will silently go on when used with unsupporting IE
1775 subtitles = info_dict['requested_subtitles']
1776 ie = self.get_info_extractor(info_dict['extractor_key'])
1777 for sub_lang, sub_info in subtitles.items():
1778 sub_format = sub_info['ext']
1779 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1780 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1781 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
1783 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1784 if sub_info.get('data') is not None:
1786 # Use newline='' to prevent conversion of newline characters
1787 # See https://github.com/rg3/youtube-dl/issues/10268
1788 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
1789 subfile.write(sub_info['data'])
1790 except (OSError, IOError):
1791 self.report_error('Cannot write subtitles file ' + sub_filename)
1795 sub_data = ie._request_webpage(
1796 sub_info['url'], info_dict['id'], note=False).read()
1797 with io.open(encodeFilename(sub_filename), 'wb') as subfile:
1798 subfile.write(sub_data)
1799 except (ExtractorError, IOError, OSError, ValueError) as err:
1800 self.report_warning('Unable to download subtitle for "%s": %s' %
1801 (sub_lang, error_to_compat_str(err)))
1804 if self.params.get('writeinfojson', False):
1805 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1806 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1807 self.to_screen('[info] Video description metadata is already present')
1809 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1811 write_json_file(self.filter_requested_info(info_dict), infofn)
1812 except (OSError, IOError):
1813 self.report_error('Cannot write metadata to JSON file ' + infofn)
1816 self._write_thumbnails(info_dict, filename)
1818 if not self.params.get('skip_download', False):
1821 fd = get_suitable_downloader(info, self.params)(self, self.params)
1822 for ph in self._progress_hooks:
1823 fd.add_progress_hook(ph)
1824 if self.params.get('verbose'):
1825 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1826 return fd.download(name, info)
1828 if info_dict.get('requested_formats') is not None:
1831 merger = FFmpegMergerPP(self)
1832 if not merger.available:
1834 self.report_warning('You have requested multiple '
1835 'formats but ffmpeg or avconv are not installed.'
1836 ' The formats won\'t be merged.')
1838 postprocessors = [merger]
1840 def compatible_formats(formats):
1841 video, audio = formats
1843 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1844 if video_ext and audio_ext:
1846 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'),
1849 for exts in COMPATIBLE_EXTS:
1850 if video_ext in exts and audio_ext in exts:
1852 # TODO: Check acodec/vcodec
1855 filename_real_ext = os.path.splitext(filename)[1][1:]
1857 os.path.splitext(filename)[0]
1858 if filename_real_ext == info_dict['ext']
1860 requested_formats = info_dict['requested_formats']
1861 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1862 info_dict['ext'] = 'mkv'
1863 self.report_warning(
1864 'Requested formats are incompatible for merge and will be merged into mkv.')
1865 # Ensure filename always has a correct extension for successful merge
1866 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1867 if os.path.exists(encodeFilename(filename)):
1869 '[download] %s has already been downloaded and '
1870 'merged' % filename)
1872 for f in requested_formats:
1873 new_info = dict(info_dict)
1875 fname = prepend_extension(
1876 self.prepare_filename(new_info),
1877 'f%s' % f['format_id'], new_info['ext'])
1878 if not ensure_dir_exists(fname):
1880 downloaded.append(fname)
1881 partial_success = dl(fname, new_info)
1882 success = success and partial_success
1883 info_dict['__postprocessors'] = postprocessors
1884 info_dict['__files_to_merge'] = downloaded
1886 # Just a single file
1887 success = dl(filename, info_dict)
1888 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1889 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1891 except (OSError, IOError) as err:
1892 raise UnavailableVideoError(err)
1893 except (ContentTooShortError, ) as err:
1894 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1897 if success and filename != '-':
1899 fixup_policy = self.params.get('fixup')
1900 if fixup_policy is None:
1901 fixup_policy = 'detect_or_warn'
1903 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1905 stretched_ratio = info_dict.get('stretched_ratio')
1906 if stretched_ratio is not None and stretched_ratio != 1:
1907 if fixup_policy == 'warn':
1908 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1909 info_dict['id'], stretched_ratio))
1910 elif fixup_policy == 'detect_or_warn':
1911 stretched_pp = FFmpegFixupStretchedPP(self)
1912 if stretched_pp.available:
1913 info_dict.setdefault('__postprocessors', [])
1914 info_dict['__postprocessors'].append(stretched_pp)
1916 self.report_warning(
1917 '%s: Non-uniform pixel ratio (%s). %s'
1918 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1920 assert fixup_policy in ('ignore', 'never')
1922 if (info_dict.get('requested_formats') is None and
1923 info_dict.get('container') == 'm4a_dash'):
1924 if fixup_policy == 'warn':
1925 self.report_warning(
1926 '%s: writing DASH m4a. '
1927 'Only some players support this container.'
1929 elif fixup_policy == 'detect_or_warn':
1930 fixup_pp = FFmpegFixupM4aPP(self)
1931 if fixup_pp.available:
1932 info_dict.setdefault('__postprocessors', [])
1933 info_dict['__postprocessors'].append(fixup_pp)
1935 self.report_warning(
1936 '%s: writing DASH m4a. '
1937 'Only some players support this container. %s'
1938 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1940 assert fixup_policy in ('ignore', 'never')
1942 if (info_dict.get('protocol') == 'm3u8_native' or
1943 info_dict.get('protocol') == 'm3u8' and
1944 self.params.get('hls_prefer_native')):
1945 if fixup_policy == 'warn':
1946 self.report_warning('%s: malformed AAC bitstream detected.' % (
1948 elif fixup_policy == 'detect_or_warn':
1949 fixup_pp = FFmpegFixupM3u8PP(self)
1950 if fixup_pp.available:
1951 info_dict.setdefault('__postprocessors', [])
1952 info_dict['__postprocessors'].append(fixup_pp)
1954 self.report_warning(
1955 '%s: malformed AAC bitstream detected. %s'
1956 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1958 assert fixup_policy in ('ignore', 'never')
1961 self.post_process(filename, info_dict)
1962 except (PostProcessingError) as err:
1963 self.report_error('postprocessing: %s' % str(err))
1965 self.record_download_archive(info_dict)
1967 def download(self, url_list):
1968 """Download a given list of URLs."""
1969 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1970 if (len(url_list) > 1 and
1972 '%' not in outtmpl and
1973 self.params.get('max_downloads') != 1):
1974 raise SameFileError(outtmpl)
1976 for url in url_list:
1978 # It also downloads the videos
1979 res = self.extract_info(
1980 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1981 except UnavailableVideoError:
1982 self.report_error('unable to download video')
1983 except MaxDownloadsReached:
1984 self.to_screen('[info] Maximum number of downloaded files reached.')
1987 if self.params.get('dump_single_json', False):
1988 self.to_stdout(json.dumps(res))
1990 return self._download_retcode
1992 def download_with_info_file(self, info_filename):
1993 with contextlib.closing(fileinput.FileInput(
1994 [info_filename], mode='r',
1995 openhook=fileinput.hook_encoded('utf-8'))) as f:
1996 # FileInput doesn't have a read method, we can't call json.load
1997 info = self.filter_requested_info(json.loads('\n'.join(f)))
1999 self.process_ie_result(info, download=True)
2000 except DownloadError:
2001 webpage_url = info.get('webpage_url')
2002 if webpage_url is not None:
2003 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2004 return self.download([webpage_url])
2007 return self._download_retcode
2010 def filter_requested_info(info_dict):
2012 (k, v) for k, v in info_dict.items()
2013 if k not in ['requested_formats', 'requested_subtitles'])
2015 def post_process(self, filename, ie_info):
2016 """Run all the postprocessors on the given file."""
2017 info = dict(ie_info)
2018 info['filepath'] = filename
2020 if ie_info.get('__postprocessors') is not None:
2021 pps_chain.extend(ie_info['__postprocessors'])
2022 pps_chain.extend(self._pps)
2023 for pp in pps_chain:
2024 files_to_delete = []
2026 files_to_delete, info = pp.run(info)
2027 except PostProcessingError as e:
2028 self.report_error(e.msg)
2029 if files_to_delete and not self.params.get('keepvideo', False):
2030 for old_filename in files_to_delete:
2031 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2033 os.remove(encodeFilename(old_filename))
2034 except (IOError, OSError):
2035 self.report_warning('Unable to remove downloaded original file')
2037 def _make_archive_id(self, info_dict):
2038 # Future-proof against any change in case
2039 # and backwards compatibility with prior versions
2040 extractor = info_dict.get('extractor_key')
2041 if extractor is None:
2042 if 'id' in info_dict:
2043 extractor = info_dict.get('ie_key') # key in a playlist
2044 if extractor is None:
2045 return None # Incomplete video information
2046 return extractor.lower() + ' ' + info_dict['id']
2048 def in_download_archive(self, info_dict):
2049 fn = self.params.get('download_archive')
2053 vid_id = self._make_archive_id(info_dict)
2055 return False # Incomplete video information
2058 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
2059 for line in archive_file:
2060 if line.strip() == vid_id:
2062 except IOError as ioe:
2063 if ioe.errno != errno.ENOENT:
2067 def record_download_archive(self, info_dict):
2068 fn = self.params.get('download_archive')
2071 vid_id = self._make_archive_id(info_dict)
2073 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2074 archive_file.write(vid_id + '\n')
2077 def format_resolution(format, default='unknown'):
2078 if format.get('vcodec') == 'none':
2080 if format.get('resolution') is not None:
2081 return format['resolution']
2082 if format.get('height') is not None:
2083 if format.get('width') is not None:
2084 res = '%sx%s' % (format['width'], format['height'])
2086 res = '%sp' % format['height']
2087 elif format.get('width') is not None:
2088 res = '%dx?' % format['width']
2093 def _format_note(self, fdict):
2095 if fdict.get('ext') in ['f4f', 'f4m']:
2096 res += '(unsupported) '
2097 if fdict.get('language'):
2100 res += '[%s] ' % fdict['language']
2101 if fdict.get('format_note') is not None:
2102 res += fdict['format_note'] + ' '
2103 if fdict.get('tbr') is not None:
2104 res += '%4dk ' % fdict['tbr']
2105 if fdict.get('container') is not None:
2108 res += '%s container' % fdict['container']
2109 if (fdict.get('vcodec') is not None and
2110 fdict.get('vcodec') != 'none'):
2113 res += fdict['vcodec']
2114 if fdict.get('vbr') is not None:
2116 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2118 if fdict.get('vbr') is not None:
2119 res += '%4dk' % fdict['vbr']
2120 if fdict.get('fps') is not None:
2123 res += '%sfps' % fdict['fps']
2124 if fdict.get('acodec') is not None:
2127 if fdict['acodec'] == 'none':
2130 res += '%-5s' % fdict['acodec']
2131 elif fdict.get('abr') is not None:
2135 if fdict.get('abr') is not None:
2136 res += '@%3dk' % fdict['abr']
2137 if fdict.get('asr') is not None:
2138 res += ' (%5dHz)' % fdict['asr']
2139 if fdict.get('filesize') is not None:
2142 res += format_bytes(fdict['filesize'])
2143 elif fdict.get('filesize_approx') is not None:
2146 res += '~' + format_bytes(fdict['filesize_approx'])
2149 def list_formats(self, info_dict):
2150 formats = info_dict.get('formats', [info_dict])
2152 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
2154 if f.get('preference') is None or f['preference'] >= -1000]
2155 if len(formats) > 1:
2156 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
2158 header_line = ['format code', 'extension', 'resolution', 'note']
2160 '[info] Available formats for %s:\n%s' %
2161 (info_dict['id'], render_table(header_line, table)))
2163 def list_thumbnails(self, info_dict):
2164 thumbnails = info_dict.get('thumbnails')
2166 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
2170 '[info] Thumbnails for %s:' % info_dict['id'])
2171 self.to_screen(render_table(
2172 ['ID', 'width', 'height', 'URL'],
2173 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
2175 def list_subtitles(self, video_id, subtitles, name='subtitles'):
2177 self.to_screen('%s has no %s' % (video_id, name))
2180 'Available %s for %s:' % (name, video_id))
2181 self.to_screen(render_table(
2182 ['Language', 'formats'],
2183 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
2184 for lang, formats in subtitles.items()]))
2186 def urlopen(self, req):
2187 """ Start an HTTP download """
2188 if isinstance(req, compat_basestring):
2189 req = sanitized_Request(req)
2190 return self._opener.open(req, timeout=self._socket_timeout)
2192 def print_debug_header(self):
2193 if not self.params.get('verbose'):
2196 if type('') is not compat_str:
2197 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
2198 self.report_warning(
2199 'Your Python is broken! Update to a newer and supported version')
2201 stdout_encoding = getattr(
2202 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2204 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2205 locale.getpreferredencoding(),
2206 sys.getfilesystemencoding(),
2208 self.get_encoding()))
2209 write_string(encoding_str, encoding=None)
2211 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
2213 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2215 sp = subprocess.Popen(
2216 ['git', 'rev-parse', '--short', 'HEAD'],
2217 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2218 cwd=os.path.dirname(os.path.abspath(__file__)))
2219 out, err = sp.communicate()
2220 out = out.decode().strip()
2221 if re.match('[0-9a-f]+', out):
2222 self._write_string('[debug] Git HEAD: ' + out + '\n')
2228 self._write_string('[debug] Python version %s - %s\n' % (
2229 platform.python_version(), platform_name()))
2231 exe_versions = FFmpegPostProcessor.get_versions(self)
2232 exe_versions['rtmpdump'] = rtmpdump_version()
2233 exe_versions['phantomjs'] = PhantomJSwrapper._version()
2234 exe_str = ', '.join(
2236 for exe, v in sorted(exe_versions.items())
2241 self._write_string('[debug] exe versions: %s\n' % exe_str)
2244 for handler in self._opener.handlers:
2245 if hasattr(handler, 'proxies'):
2246 proxy_map.update(handler.proxies)
2247 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2249 if self.params.get('call_home', False):
2250 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2251 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2252 latest_version = self.urlopen(
2253 'https://yt-dl.org/latest/version').read().decode('utf-8')
2254 if version_tuple(latest_version) > version_tuple(__version__):
2255 self.report_warning(
2256 'You are using an outdated version (newest version: %s)! '
2257 'See https://yt-dl.org/update if you need help updating.' %
2260 def _setup_opener(self):
2261 timeout_val = self.params.get('socket_timeout')
2262 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2264 opts_cookiefile = self.params.get('cookiefile')
2265 opts_proxy = self.params.get('proxy')
2267 if opts_cookiefile is None:
2268 self.cookiejar = compat_cookiejar.CookieJar()
2270 opts_cookiefile = expand_path(opts_cookiefile)
2271 self.cookiejar = compat_cookiejar.MozillaCookieJar(
2273 if os.access(opts_cookiefile, os.R_OK):
2274 self.cookiejar.load()
2276 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2277 if opts_proxy is not None:
2278 if opts_proxy == '':
2281 proxies = {'http': opts_proxy, 'https': opts_proxy}
2283 proxies = compat_urllib_request.getproxies()
2284 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2285 if 'http' in proxies and 'https' not in proxies:
2286 proxies['https'] = proxies['http']
2287 proxy_handler = PerRequestProxyHandler(proxies)
2289 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2290 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2291 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2292 data_handler = compat_urllib_request_DataHandler()
2294 # When passing our own FileHandler instance, build_opener won't add the
2295 # default FileHandler and allows us to disable the file protocol, which
2296 # can be used for malicious purposes (see
2297 # https://github.com/rg3/youtube-dl/issues/8227)
2298 file_handler = compat_urllib_request.FileHandler()
2300 def file_open(*args, **kwargs):
2301 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2302 file_handler.file_open = file_open
2304 opener = compat_urllib_request.build_opener(
2305 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2307 # Delete the default user-agent header, which would otherwise apply in
2308 # cases where our custom HTTP handler doesn't come into play
2309 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2310 opener.addheaders = []
2311 self._opener = opener
2313 def encode(self, s):
2314 if isinstance(s, bytes):
2315 return s # Already encoded
2318 return s.encode(self.get_encoding())
2319 except UnicodeEncodeError as err:
2320 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2323 def get_encoding(self):
2324 encoding = self.params.get('encoding')
2325 if encoding is None:
2326 encoding = preferredencoding()
2329 def _write_thumbnails(self, info_dict, filename):
2330 if self.params.get('writethumbnail', False):
2331 thumbnails = info_dict.get('thumbnails')
2333 thumbnails = [thumbnails[-1]]
2334 elif self.params.get('write_all_thumbnails', False):
2335 thumbnails = info_dict.get('thumbnails')
2340 # No thumbnails present, so return immediately
2343 for t in thumbnails:
2344 thumb_ext = determine_ext(t['url'], 'jpg')
2345 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2346 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2347 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2349 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2350 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2351 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2353 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2354 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2356 uf = self.urlopen(t['url'])
2357 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2358 shutil.copyfileobj(uf, thumbf)
2359 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2360 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2361 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2362 self.report_warning('Unable to download thumbnail "%s": %s' %
2363 (t['url'], error_to_compat_str(err)))