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(itertools.islice(
952 ie_entries, 0, max(playlistitems))))
954 entries = list(itertools.islice(
955 ie_entries, playliststart, playlistend))
956 n_entries = len(entries)
957 report_download(n_entries)
959 if self.params.get('playlistreverse', False):
960 entries = entries[::-1]
962 if self.params.get('playlistrandom', False):
963 random.shuffle(entries)
965 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
967 for i, entry in enumerate(entries, 1):
968 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
969 # This __x_forwarded_for_ip thing is a bit ugly but requires
972 entry['__x_forwarded_for_ip'] = x_forwarded_for
974 'n_entries': n_entries,
975 'playlist': playlist,
976 'playlist_id': ie_result.get('id'),
977 'playlist_title': ie_result.get('title'),
978 'playlist_index': i + playliststart,
979 'extractor': ie_result['extractor'],
980 'webpage_url': ie_result['webpage_url'],
981 'webpage_url_basename': url_basename(ie_result['webpage_url']),
982 'extractor_key': ie_result['extractor_key'],
985 reason = self._match_entry(entry, incomplete=True)
986 if reason is not None:
987 self.to_screen('[download] ' + reason)
990 entry_result = self.process_ie_result(entry,
993 playlist_results.append(entry_result)
994 ie_result['entries'] = playlist_results
995 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
997 elif result_type == 'compat_list':
999 'Extractor %s returned a compat_list result. '
1000 'It needs to be updated.' % ie_result.get('extractor'))
1003 self.add_extra_info(
1006 'extractor': ie_result['extractor'],
1007 'webpage_url': ie_result['webpage_url'],
1008 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1009 'extractor_key': ie_result['extractor_key'],
1013 ie_result['entries'] = [
1014 self.process_ie_result(_fixup(r), download, extra_info)
1015 for r in ie_result['entries']
1019 raise Exception('Invalid result type: %s' % result_type)
1021 def _build_format_filter(self, filter_spec):
1022 " Returns a function to filter the formats according to the filter_spec "
1032 operator_rex = re.compile(r'''(?x)\s*
1033 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
1034 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1035 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
1037 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1038 m = operator_rex.search(filter_spec)
1041 comparison_value = int(m.group('value'))
1043 comparison_value = parse_filesize(m.group('value'))
1044 if comparison_value is None:
1045 comparison_value = parse_filesize(m.group('value') + 'B')
1046 if comparison_value is None:
1048 'Invalid value %r in format specification %r' % (
1049 m.group('value'), filter_spec))
1050 op = OPERATORS[m.group('op')]
1056 '^=': lambda attr, value: attr.startswith(value),
1057 '$=': lambda attr, value: attr.endswith(value),
1058 '*=': lambda attr, value: value in attr,
1060 str_operator_rex = re.compile(r'''(?x)
1061 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
1062 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
1063 \s*(?P<value>[a-zA-Z0-9._-]+)
1065 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1066 m = str_operator_rex.search(filter_spec)
1068 comparison_value = m.group('value')
1069 op = STR_OPERATORS[m.group('op')]
1072 raise ValueError('Invalid filter specification %r' % filter_spec)
1075 actual_value = f.get(m.group('key'))
1076 if actual_value is None:
1077 return m.group('none_inclusive')
1078 return op(actual_value, comparison_value)
1081 def _default_format_spec(self, info_dict, download=True):
1084 merger = FFmpegMergerPP(self)
1085 return merger.available and merger.can_merge()
1088 if self.params.get('simulate', False):
1092 if self.params.get('outtmpl', DEFAULT_OUTTMPL) == '-':
1094 if info_dict.get('is_live'):
1100 req_format_list = ['bestvideo+bestaudio', 'best']
1102 req_format_list.reverse()
1103 return '/'.join(req_format_list)
1105 def build_format_selector(self, format_spec):
1106 def syntax_error(note, start):
1108 'Invalid format specification: '
1109 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1110 return SyntaxError(message)
1112 PICKFIRST = 'PICKFIRST'
1116 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1118 def _parse_filter(tokens):
1120 for type, string, start, _, _ in tokens:
1121 if type == tokenize.OP and string == ']':
1122 return ''.join(filter_parts)
1124 filter_parts.append(string)
1126 def _remove_unused_ops(tokens):
1127 # Remove operators that we don't use and join them with the surrounding strings
1128 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1129 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1130 last_string, last_start, last_end, last_line = None, None, None, None
1131 for type, string, start, end, line in tokens:
1132 if type == tokenize.OP and string == '[':
1134 yield tokenize.NAME, last_string, last_start, last_end, last_line
1136 yield type, string, start, end, line
1137 # everything inside brackets will be handled by _parse_filter
1138 for type, string, start, end, line in tokens:
1139 yield type, string, start, end, line
1140 if type == tokenize.OP and string == ']':
1142 elif type == tokenize.OP and string in ALLOWED_OPS:
1144 yield tokenize.NAME, last_string, last_start, last_end, last_line
1146 yield type, string, start, end, line
1147 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1149 last_string = string
1153 last_string += string
1155 yield tokenize.NAME, last_string, last_start, last_end, last_line
1157 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1159 current_selector = None
1160 for type, string, start, _, _ in tokens:
1161 # ENCODING is only defined in python 3.x
1162 if type == getattr(tokenize, 'ENCODING', None):
1164 elif type in [tokenize.NAME, tokenize.NUMBER]:
1165 current_selector = FormatSelector(SINGLE, string, [])
1166 elif type == tokenize.OP:
1168 if not inside_group:
1169 # ')' will be handled by the parentheses group
1170 tokens.restore_last_token()
1172 elif inside_merge and string in ['/', ',']:
1173 tokens.restore_last_token()
1175 elif inside_choice and string == ',':
1176 tokens.restore_last_token()
1179 if not current_selector:
1180 raise syntax_error('"," must follow a format selector', start)
1181 selectors.append(current_selector)
1182 current_selector = None
1184 if not current_selector:
1185 raise syntax_error('"/" must follow a format selector', start)
1186 first_choice = current_selector
1187 second_choice = _parse_format_selection(tokens, inside_choice=True)
1188 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1190 if not current_selector:
1191 current_selector = FormatSelector(SINGLE, 'best', [])
1192 format_filter = _parse_filter(tokens)
1193 current_selector.filters.append(format_filter)
1195 if current_selector:
1196 raise syntax_error('Unexpected "("', start)
1197 group = _parse_format_selection(tokens, inside_group=True)
1198 current_selector = FormatSelector(GROUP, group, [])
1200 video_selector = current_selector
1201 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1202 if not video_selector or not audio_selector:
1203 raise syntax_error('"+" must be between two format selectors', start)
1204 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1206 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1207 elif type == tokenize.ENDMARKER:
1209 if current_selector:
1210 selectors.append(current_selector)
1213 def _build_selector_function(selector):
1214 if isinstance(selector, list):
1215 fs = [_build_selector_function(s) for s in selector]
1217 def selector_function(ctx):
1219 for format in f(ctx):
1221 return selector_function
1222 elif selector.type == GROUP:
1223 selector_function = _build_selector_function(selector.selector)
1224 elif selector.type == PICKFIRST:
1225 fs = [_build_selector_function(s) for s in selector.selector]
1227 def selector_function(ctx):
1229 picked_formats = list(f(ctx))
1231 return picked_formats
1233 elif selector.type == SINGLE:
1234 format_spec = selector.selector
1236 def selector_function(ctx):
1237 formats = list(ctx['formats'])
1240 if format_spec == 'all':
1243 elif format_spec in ['best', 'worst', None]:
1244 format_idx = 0 if format_spec == 'worst' else -1
1245 audiovideo_formats = [
1247 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1248 if audiovideo_formats:
1249 yield audiovideo_formats[format_idx]
1250 # for extractors with incomplete formats (audio only (soundcloud)
1251 # or video only (imgur)) we will fallback to best/worst
1252 # {video,audio}-only format
1253 elif ctx['incomplete_formats']:
1254 yield formats[format_idx]
1255 elif format_spec == 'bestaudio':
1258 if f.get('vcodec') == 'none']
1260 yield audio_formats[-1]
1261 elif format_spec == 'worstaudio':
1264 if f.get('vcodec') == 'none']
1266 yield audio_formats[0]
1267 elif format_spec == 'bestvideo':
1270 if f.get('acodec') == 'none']
1272 yield video_formats[-1]
1273 elif format_spec == 'worstvideo':
1276 if f.get('acodec') == 'none']
1278 yield video_formats[0]
1280 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1281 if format_spec in extensions:
1282 filter_f = lambda f: f['ext'] == format_spec
1284 filter_f = lambda f: f['format_id'] == format_spec
1285 matches = list(filter(filter_f, formats))
1288 elif selector.type == MERGE:
1289 def _merge(formats_info):
1290 format_1, format_2 = [f['format_id'] for f in formats_info]
1291 # The first format must contain the video and the
1293 if formats_info[0].get('vcodec') == 'none':
1294 self.report_error('The first format must '
1295 'contain the video, try using '
1296 '"-f %s+%s"' % (format_2, format_1))
1298 # Formats must be opposite (video+audio)
1299 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1301 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1302 % (format_1, format_2))
1305 formats_info[0]['ext']
1306 if self.params.get('merge_output_format') is None
1307 else self.params['merge_output_format'])
1309 'requested_formats': formats_info,
1310 'format': '%s+%s' % (formats_info[0].get('format'),
1311 formats_info[1].get('format')),
1312 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1313 formats_info[1].get('format_id')),
1314 'width': formats_info[0].get('width'),
1315 'height': formats_info[0].get('height'),
1316 'resolution': formats_info[0].get('resolution'),
1317 'fps': formats_info[0].get('fps'),
1318 'vcodec': formats_info[0].get('vcodec'),
1319 'vbr': formats_info[0].get('vbr'),
1320 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1321 'acodec': formats_info[1].get('acodec'),
1322 'abr': formats_info[1].get('abr'),
1325 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1327 def selector_function(ctx):
1328 for pair in itertools.product(
1329 video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
1332 filters = [self._build_format_filter(f) for f in selector.filters]
1334 def final_selector(ctx):
1335 ctx_copy = copy.deepcopy(ctx)
1336 for _filter in filters:
1337 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1338 return selector_function(ctx_copy)
1339 return final_selector
1341 stream = io.BytesIO(format_spec.encode('utf-8'))
1343 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1344 except tokenize.TokenError:
1345 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1347 class TokenIterator(object):
1348 def __init__(self, tokens):
1349 self.tokens = tokens
1356 if self.counter >= len(self.tokens):
1357 raise StopIteration()
1358 value = self.tokens[self.counter]
1364 def restore_last_token(self):
1367 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1368 return _build_selector_function(parsed_selector)
1370 def _calc_headers(self, info_dict):
1371 res = std_headers.copy()
1373 add_headers = info_dict.get('http_headers')
1375 res.update(add_headers)
1377 cookies = self._calc_cookies(info_dict)
1379 res['Cookie'] = cookies
1381 if 'X-Forwarded-For' not in res:
1382 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1383 if x_forwarded_for_ip:
1384 res['X-Forwarded-For'] = x_forwarded_for_ip
1388 def _calc_cookies(self, info_dict):
1389 pr = sanitized_Request(info_dict['url'])
1390 self.cookiejar.add_cookie_header(pr)
1391 return pr.get_header('Cookie')
1393 def process_video_result(self, info_dict, download=True):
1394 assert info_dict.get('_type', 'video') == 'video'
1396 if 'id' not in info_dict:
1397 raise ExtractorError('Missing "id" field in extractor result')
1398 if 'title' not in info_dict:
1399 raise ExtractorError('Missing "title" field in extractor result')
1401 def report_force_conversion(field, field_not, conversion):
1402 self.report_warning(
1403 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
1404 % (field, field_not, conversion))
1406 def sanitize_string_field(info, string_field):
1407 field = info.get(string_field)
1408 if field is None or isinstance(field, compat_str):
1410 report_force_conversion(string_field, 'a string', 'string')
1411 info[string_field] = compat_str(field)
1413 def sanitize_numeric_fields(info):
1414 for numeric_field in self._NUMERIC_FIELDS:
1415 field = info.get(numeric_field)
1416 if field is None or isinstance(field, compat_numeric_types):
1418 report_force_conversion(numeric_field, 'numeric', 'int')
1419 info[numeric_field] = int_or_none(field)
1421 sanitize_string_field(info_dict, 'id')
1422 sanitize_numeric_fields(info_dict)
1424 if 'playlist' not in info_dict:
1425 # It isn't part of a playlist
1426 info_dict['playlist'] = None
1427 info_dict['playlist_index'] = None
1429 thumbnails = info_dict.get('thumbnails')
1430 if thumbnails is None:
1431 thumbnail = info_dict.get('thumbnail')
1433 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1435 thumbnails.sort(key=lambda t: (
1436 t.get('preference') if t.get('preference') is not None else -1,
1437 t.get('width') if t.get('width') is not None else -1,
1438 t.get('height') if t.get('height') is not None else -1,
1439 t.get('id') if t.get('id') is not None else '', t.get('url')))
1440 for i, t in enumerate(thumbnails):
1441 t['url'] = sanitize_url(t['url'])
1442 if t.get('width') and t.get('height'):
1443 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1444 if t.get('id') is None:
1447 if self.params.get('list_thumbnails'):
1448 self.list_thumbnails(info_dict)
1451 thumbnail = info_dict.get('thumbnail')
1453 info_dict['thumbnail'] = sanitize_url(thumbnail)
1455 info_dict['thumbnail'] = thumbnails[-1]['url']
1457 if 'display_id' not in info_dict and 'id' in info_dict:
1458 info_dict['display_id'] = info_dict['id']
1460 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1461 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1462 # see http://bugs.python.org/issue1646728)
1464 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1465 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1466 except (ValueError, OverflowError, OSError):
1469 # Auto generate title fields corresponding to the *_number fields when missing
1470 # in order to always have clean titles. This is very common for TV series.
1471 for field in ('chapter', 'season', 'episode'):
1472 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1473 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1475 subtitles = info_dict.get('subtitles')
1477 for _, subtitle in subtitles.items():
1478 for subtitle_format in subtitle:
1479 if subtitle_format.get('url'):
1480 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1481 if subtitle_format.get('ext') is None:
1482 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1484 if self.params.get('listsubtitles', False):
1485 if 'automatic_captions' in info_dict:
1486 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1487 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1489 info_dict['requested_subtitles'] = self.process_subtitles(
1490 info_dict['id'], subtitles,
1491 info_dict.get('automatic_captions'))
1493 # We now pick which formats have to be downloaded
1494 if info_dict.get('formats') is None:
1495 # There's only one format available
1496 formats = [info_dict]
1498 formats = info_dict['formats']
1501 raise ExtractorError('No video formats found!')
1503 def is_wellformed(f):
1506 self.report_warning(
1507 '"url" field is missing or empty - skipping format, '
1508 'there is an error in extractor')
1510 if isinstance(url, bytes):
1511 sanitize_string_field(f, 'url')
1514 # Filter out malformed formats for better extraction robustness
1515 formats = list(filter(is_wellformed, formats))
1519 # We check that all the formats have the format and format_id fields
1520 for i, format in enumerate(formats):
1521 sanitize_string_field(format, 'format_id')
1522 sanitize_numeric_fields(format)
1523 format['url'] = sanitize_url(format['url'])
1524 if not format.get('format_id'):
1525 format['format_id'] = compat_str(i)
1527 # Sanitize format_id from characters used in format selector expression
1528 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
1529 format_id = format['format_id']
1530 if format_id not in formats_dict:
1531 formats_dict[format_id] = []
1532 formats_dict[format_id].append(format)
1534 # Make sure all formats have unique format_id
1535 for format_id, ambiguous_formats in formats_dict.items():
1536 if len(ambiguous_formats) > 1:
1537 for i, format in enumerate(ambiguous_formats):
1538 format['format_id'] = '%s-%d' % (format_id, i)
1540 for i, format in enumerate(formats):
1541 if format.get('format') is None:
1542 format['format'] = '{id} - {res}{note}'.format(
1543 id=format['format_id'],
1544 res=self.format_resolution(format),
1545 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1547 # Automatically determine file extension if missing
1548 if format.get('ext') is None:
1549 format['ext'] = determine_ext(format['url']).lower()
1550 # Automatically determine protocol if missing (useful for format
1551 # selection purposes)
1552 if format.get('protocol') is None:
1553 format['protocol'] = determine_protocol(format)
1554 # Add HTTP headers, so that external programs can use them from the
1556 full_format_info = info_dict.copy()
1557 full_format_info.update(format)
1558 format['http_headers'] = self._calc_headers(full_format_info)
1559 # Remove private housekeeping stuff
1560 if '__x_forwarded_for_ip' in info_dict:
1561 del info_dict['__x_forwarded_for_ip']
1563 # TODO Central sorting goes here
1565 if formats[0] is not info_dict:
1566 # only set the 'formats' fields if the original info_dict list them
1567 # otherwise we end up with a circular reference, the first (and unique)
1568 # element in the 'formats' field in info_dict is info_dict itself,
1569 # which can't be exported to json
1570 info_dict['formats'] = formats
1571 if self.params.get('listformats'):
1572 self.list_formats(info_dict)
1575 req_format = self.params.get('format')
1576 if req_format is None:
1577 req_format = self._default_format_spec(info_dict, download=download)
1578 if self.params.get('verbose'):
1579 self.to_stdout('[debug] Default format spec: %s' % req_format)
1581 format_selector = self.build_format_selector(req_format)
1583 # While in format selection we may need to have an access to the original
1584 # format set in order to calculate some metrics or do some processing.
1585 # For now we need to be able to guess whether original formats provided
1586 # by extractor are incomplete or not (i.e. whether extractor provides only
1587 # video-only or audio-only formats) for proper formats selection for
1588 # extractors with such incomplete formats (see
1589 # https://github.com/rg3/youtube-dl/pull/5556).
1590 # Since formats may be filtered during format selection and may not match
1591 # the original formats the results may be incorrect. Thus original formats
1592 # or pre-calculated metrics should be passed to format selection routines
1594 # We will pass a context object containing all necessary additional data
1595 # instead of just formats.
1596 # This fixes incorrect format selection issue (see
1597 # https://github.com/rg3/youtube-dl/issues/10083).
1598 incomplete_formats = (
1599 # All formats are video-only or
1600 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) or
1601 # all formats are audio-only
1602 all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1606 'incomplete_formats': incomplete_formats,
1609 formats_to_download = list(format_selector(ctx))
1610 if not formats_to_download:
1611 raise ExtractorError('requested format not available',
1615 if len(formats_to_download) > 1:
1616 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1617 for format in formats_to_download:
1618 new_info = dict(info_dict)
1619 new_info.update(format)
1620 self.process_info(new_info)
1621 # We update the info dict with the best quality format (backwards compatibility)
1622 info_dict.update(formats_to_download[-1])
1625 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1626 """Select the requested subtitles and their format"""
1628 if normal_subtitles and self.params.get('writesubtitles'):
1629 available_subs.update(normal_subtitles)
1630 if automatic_captions and self.params.get('writeautomaticsub'):
1631 for lang, cap_info in automatic_captions.items():
1632 if lang not in available_subs:
1633 available_subs[lang] = cap_info
1635 if (not self.params.get('writesubtitles') and not
1636 self.params.get('writeautomaticsub') or not
1640 if self.params.get('allsubtitles', False):
1641 requested_langs = available_subs.keys()
1643 if self.params.get('subtitleslangs', False):
1644 requested_langs = self.params.get('subtitleslangs')
1645 elif 'en' in available_subs:
1646 requested_langs = ['en']
1648 requested_langs = [list(available_subs.keys())[0]]
1650 formats_query = self.params.get('subtitlesformat', 'best')
1651 formats_preference = formats_query.split('/') if formats_query else []
1653 for lang in requested_langs:
1654 formats = available_subs.get(lang)
1656 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1658 for ext in formats_preference:
1662 matches = list(filter(lambda f: f['ext'] == ext, formats))
1668 self.report_warning(
1669 'No subtitle format found matching "%s" for language %s, '
1670 'using %s' % (formats_query, lang, f['ext']))
1674 def process_info(self, info_dict):
1675 """Process a single resolved IE result."""
1677 assert info_dict.get('_type', 'video') == 'video'
1679 max_downloads = self.params.get('max_downloads')
1680 if max_downloads is not None:
1681 if self._num_downloads >= int(max_downloads):
1682 raise MaxDownloadsReached()
1684 info_dict['fulltitle'] = info_dict['title']
1685 if len(info_dict['title']) > 200:
1686 info_dict['title'] = info_dict['title'][:197] + '...'
1688 if 'format' not in info_dict:
1689 info_dict['format'] = info_dict['ext']
1691 reason = self._match_entry(info_dict, incomplete=False)
1692 if reason is not None:
1693 self.to_screen('[download] ' + reason)
1696 self._num_downloads += 1
1698 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1701 if self.params.get('forcetitle', False):
1702 self.to_stdout(info_dict['fulltitle'])
1703 if self.params.get('forceid', False):
1704 self.to_stdout(info_dict['id'])
1705 if self.params.get('forceurl', False):
1706 if info_dict.get('requested_formats') is not None:
1707 for f in info_dict['requested_formats']:
1708 self.to_stdout(f['url'] + f.get('play_path', ''))
1710 # For RTMP URLs, also include the playpath
1711 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1712 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1713 self.to_stdout(info_dict['thumbnail'])
1714 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1715 self.to_stdout(info_dict['description'])
1716 if self.params.get('forcefilename', False) and filename is not None:
1717 self.to_stdout(filename)
1718 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1719 self.to_stdout(formatSeconds(info_dict['duration']))
1720 if self.params.get('forceformat', False):
1721 self.to_stdout(info_dict['format'])
1722 if self.params.get('forcejson', False):
1723 self.to_stdout(json.dumps(info_dict))
1725 # Do nothing else if in simulate mode
1726 if self.params.get('simulate', False):
1729 if filename is None:
1732 def ensure_dir_exists(path):
1734 dn = os.path.dirname(path)
1735 if dn and not os.path.exists(dn):
1738 except (OSError, IOError) as err:
1739 self.report_error('unable to create directory ' + error_to_compat_str(err))
1742 if not ensure_dir_exists(sanitize_path(encodeFilename(filename))):
1745 if self.params.get('writedescription', False):
1746 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1747 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1748 self.to_screen('[info] Video description is already present')
1749 elif info_dict.get('description') is None:
1750 self.report_warning('There\'s no description to write.')
1753 self.to_screen('[info] Writing video description to: ' + descfn)
1754 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1755 descfile.write(info_dict['description'])
1756 except (OSError, IOError):
1757 self.report_error('Cannot write description file ' + descfn)
1760 if self.params.get('writeannotations', False):
1761 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1762 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1763 self.to_screen('[info] Video annotations are already present')
1766 self.to_screen('[info] Writing video annotations to: ' + annofn)
1767 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1768 annofile.write(info_dict['annotations'])
1769 except (KeyError, TypeError):
1770 self.report_warning('There are no annotations to write.')
1771 except (OSError, IOError):
1772 self.report_error('Cannot write annotations file: ' + annofn)
1775 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1776 self.params.get('writeautomaticsub')])
1778 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1779 # subtitles download errors are already managed as troubles in relevant IE
1780 # that way it will silently go on when used with unsupporting IE
1781 subtitles = info_dict['requested_subtitles']
1782 ie = self.get_info_extractor(info_dict['extractor_key'])
1783 for sub_lang, sub_info in subtitles.items():
1784 sub_format = sub_info['ext']
1785 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1786 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1787 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
1789 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1790 if sub_info.get('data') is not None:
1792 # Use newline='' to prevent conversion of newline characters
1793 # See https://github.com/rg3/youtube-dl/issues/10268
1794 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
1795 subfile.write(sub_info['data'])
1796 except (OSError, IOError):
1797 self.report_error('Cannot write subtitles file ' + sub_filename)
1801 sub_data = ie._request_webpage(
1802 sub_info['url'], info_dict['id'], note=False).read()
1803 with io.open(encodeFilename(sub_filename), 'wb') as subfile:
1804 subfile.write(sub_data)
1805 except (ExtractorError, IOError, OSError, ValueError) as err:
1806 self.report_warning('Unable to download subtitle for "%s": %s' %
1807 (sub_lang, error_to_compat_str(err)))
1810 if self.params.get('writeinfojson', False):
1811 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1812 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1813 self.to_screen('[info] Video description metadata is already present')
1815 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1817 write_json_file(self.filter_requested_info(info_dict), infofn)
1818 except (OSError, IOError):
1819 self.report_error('Cannot write metadata to JSON file ' + infofn)
1822 self._write_thumbnails(info_dict, filename)
1824 if not self.params.get('skip_download', False):
1827 fd = get_suitable_downloader(info, self.params)(self, self.params)
1828 for ph in self._progress_hooks:
1829 fd.add_progress_hook(ph)
1830 if self.params.get('verbose'):
1831 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1832 return fd.download(name, info)
1834 if info_dict.get('requested_formats') is not None:
1837 merger = FFmpegMergerPP(self)
1838 if not merger.available:
1840 self.report_warning('You have requested multiple '
1841 'formats but ffmpeg or avconv are not installed.'
1842 ' The formats won\'t be merged.')
1844 postprocessors = [merger]
1846 def compatible_formats(formats):
1847 video, audio = formats
1849 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1850 if video_ext and audio_ext:
1852 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'),
1855 for exts in COMPATIBLE_EXTS:
1856 if video_ext in exts and audio_ext in exts:
1858 # TODO: Check acodec/vcodec
1861 filename_real_ext = os.path.splitext(filename)[1][1:]
1863 os.path.splitext(filename)[0]
1864 if filename_real_ext == info_dict['ext']
1866 requested_formats = info_dict['requested_formats']
1867 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1868 info_dict['ext'] = 'mkv'
1869 self.report_warning(
1870 'Requested formats are incompatible for merge and will be merged into mkv.')
1871 # Ensure filename always has a correct extension for successful merge
1872 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1873 if os.path.exists(encodeFilename(filename)):
1875 '[download] %s has already been downloaded and '
1876 'merged' % filename)
1878 for f in requested_formats:
1879 new_info = dict(info_dict)
1881 fname = prepend_extension(
1882 self.prepare_filename(new_info),
1883 'f%s' % f['format_id'], new_info['ext'])
1884 if not ensure_dir_exists(fname):
1886 downloaded.append(fname)
1887 partial_success = dl(fname, new_info)
1888 success = success and partial_success
1889 info_dict['__postprocessors'] = postprocessors
1890 info_dict['__files_to_merge'] = downloaded
1892 # Just a single file
1893 success = dl(filename, info_dict)
1894 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1895 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1897 except (OSError, IOError) as err:
1898 raise UnavailableVideoError(err)
1899 except (ContentTooShortError, ) as err:
1900 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1903 if success and filename != '-':
1905 fixup_policy = self.params.get('fixup')
1906 if fixup_policy is None:
1907 fixup_policy = 'detect_or_warn'
1909 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1911 stretched_ratio = info_dict.get('stretched_ratio')
1912 if stretched_ratio is not None and stretched_ratio != 1:
1913 if fixup_policy == 'warn':
1914 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1915 info_dict['id'], stretched_ratio))
1916 elif fixup_policy == 'detect_or_warn':
1917 stretched_pp = FFmpegFixupStretchedPP(self)
1918 if stretched_pp.available:
1919 info_dict.setdefault('__postprocessors', [])
1920 info_dict['__postprocessors'].append(stretched_pp)
1922 self.report_warning(
1923 '%s: Non-uniform pixel ratio (%s). %s'
1924 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1926 assert fixup_policy in ('ignore', 'never')
1928 if (info_dict.get('requested_formats') is None and
1929 info_dict.get('container') == 'm4a_dash'):
1930 if fixup_policy == 'warn':
1931 self.report_warning(
1932 '%s: writing DASH m4a. '
1933 'Only some players support this container.'
1935 elif fixup_policy == 'detect_or_warn':
1936 fixup_pp = FFmpegFixupM4aPP(self)
1937 if fixup_pp.available:
1938 info_dict.setdefault('__postprocessors', [])
1939 info_dict['__postprocessors'].append(fixup_pp)
1941 self.report_warning(
1942 '%s: writing DASH m4a. '
1943 'Only some players support this container. %s'
1944 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1946 assert fixup_policy in ('ignore', 'never')
1948 if (info_dict.get('protocol') == 'm3u8_native' or
1949 info_dict.get('protocol') == 'm3u8' and
1950 self.params.get('hls_prefer_native')):
1951 if fixup_policy == 'warn':
1952 self.report_warning('%s: malformed AAC bitstream detected.' % (
1954 elif fixup_policy == 'detect_or_warn':
1955 fixup_pp = FFmpegFixupM3u8PP(self)
1956 if fixup_pp.available:
1957 info_dict.setdefault('__postprocessors', [])
1958 info_dict['__postprocessors'].append(fixup_pp)
1960 self.report_warning(
1961 '%s: malformed AAC bitstream detected. %s'
1962 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1964 assert fixup_policy in ('ignore', 'never')
1967 self.post_process(filename, info_dict)
1968 except (PostProcessingError) as err:
1969 self.report_error('postprocessing: %s' % str(err))
1971 self.record_download_archive(info_dict)
1973 def download(self, url_list):
1974 """Download a given list of URLs."""
1975 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1976 if (len(url_list) > 1 and
1978 '%' not in outtmpl and
1979 self.params.get('max_downloads') != 1):
1980 raise SameFileError(outtmpl)
1982 for url in url_list:
1984 # It also downloads the videos
1985 res = self.extract_info(
1986 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1987 except UnavailableVideoError:
1988 self.report_error('unable to download video')
1989 except MaxDownloadsReached:
1990 self.to_screen('[info] Maximum number of downloaded files reached.')
1993 if self.params.get('dump_single_json', False):
1994 self.to_stdout(json.dumps(res))
1996 return self._download_retcode
1998 def download_with_info_file(self, info_filename):
1999 with contextlib.closing(fileinput.FileInput(
2000 [info_filename], mode='r',
2001 openhook=fileinput.hook_encoded('utf-8'))) as f:
2002 # FileInput doesn't have a read method, we can't call json.load
2003 info = self.filter_requested_info(json.loads('\n'.join(f)))
2005 self.process_ie_result(info, download=True)
2006 except DownloadError:
2007 webpage_url = info.get('webpage_url')
2008 if webpage_url is not None:
2009 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2010 return self.download([webpage_url])
2013 return self._download_retcode
2016 def filter_requested_info(info_dict):
2018 (k, v) for k, v in info_dict.items()
2019 if k not in ['requested_formats', 'requested_subtitles'])
2021 def post_process(self, filename, ie_info):
2022 """Run all the postprocessors on the given file."""
2023 info = dict(ie_info)
2024 info['filepath'] = filename
2026 if ie_info.get('__postprocessors') is not None:
2027 pps_chain.extend(ie_info['__postprocessors'])
2028 pps_chain.extend(self._pps)
2029 for pp in pps_chain:
2030 files_to_delete = []
2032 files_to_delete, info = pp.run(info)
2033 except PostProcessingError as e:
2034 self.report_error(e.msg)
2035 if files_to_delete and not self.params.get('keepvideo', False):
2036 for old_filename in files_to_delete:
2037 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2039 os.remove(encodeFilename(old_filename))
2040 except (IOError, OSError):
2041 self.report_warning('Unable to remove downloaded original file')
2043 def _make_archive_id(self, info_dict):
2044 # Future-proof against any change in case
2045 # and backwards compatibility with prior versions
2046 extractor = info_dict.get('extractor_key')
2047 if extractor is None:
2048 if 'id' in info_dict:
2049 extractor = info_dict.get('ie_key') # key in a playlist
2050 if extractor is None:
2051 return None # Incomplete video information
2052 return extractor.lower() + ' ' + info_dict['id']
2054 def in_download_archive(self, info_dict):
2055 fn = self.params.get('download_archive')
2059 vid_id = self._make_archive_id(info_dict)
2061 return False # Incomplete video information
2064 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
2065 for line in archive_file:
2066 if line.strip() == vid_id:
2068 except IOError as ioe:
2069 if ioe.errno != errno.ENOENT:
2073 def record_download_archive(self, info_dict):
2074 fn = self.params.get('download_archive')
2077 vid_id = self._make_archive_id(info_dict)
2079 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2080 archive_file.write(vid_id + '\n')
2083 def format_resolution(format, default='unknown'):
2084 if format.get('vcodec') == 'none':
2086 if format.get('resolution') is not None:
2087 return format['resolution']
2088 if format.get('height') is not None:
2089 if format.get('width') is not None:
2090 res = '%sx%s' % (format['width'], format['height'])
2092 res = '%sp' % format['height']
2093 elif format.get('width') is not None:
2094 res = '%dx?' % format['width']
2099 def _format_note(self, fdict):
2101 if fdict.get('ext') in ['f4f', 'f4m']:
2102 res += '(unsupported) '
2103 if fdict.get('language'):
2106 res += '[%s] ' % fdict['language']
2107 if fdict.get('format_note') is not None:
2108 res += fdict['format_note'] + ' '
2109 if fdict.get('tbr') is not None:
2110 res += '%4dk ' % fdict['tbr']
2111 if fdict.get('container') is not None:
2114 res += '%s container' % fdict['container']
2115 if (fdict.get('vcodec') is not None and
2116 fdict.get('vcodec') != 'none'):
2119 res += fdict['vcodec']
2120 if fdict.get('vbr') is not None:
2122 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2124 if fdict.get('vbr') is not None:
2125 res += '%4dk' % fdict['vbr']
2126 if fdict.get('fps') is not None:
2129 res += '%sfps' % fdict['fps']
2130 if fdict.get('acodec') is not None:
2133 if fdict['acodec'] == 'none':
2136 res += '%-5s' % fdict['acodec']
2137 elif fdict.get('abr') is not None:
2141 if fdict.get('abr') is not None:
2142 res += '@%3dk' % fdict['abr']
2143 if fdict.get('asr') is not None:
2144 res += ' (%5dHz)' % fdict['asr']
2145 if fdict.get('filesize') is not None:
2148 res += format_bytes(fdict['filesize'])
2149 elif fdict.get('filesize_approx') is not None:
2152 res += '~' + format_bytes(fdict['filesize_approx'])
2155 def list_formats(self, info_dict):
2156 formats = info_dict.get('formats', [info_dict])
2158 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
2160 if f.get('preference') is None or f['preference'] >= -1000]
2161 if len(formats) > 1:
2162 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
2164 header_line = ['format code', 'extension', 'resolution', 'note']
2166 '[info] Available formats for %s:\n%s' %
2167 (info_dict['id'], render_table(header_line, table)))
2169 def list_thumbnails(self, info_dict):
2170 thumbnails = info_dict.get('thumbnails')
2172 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
2176 '[info] Thumbnails for %s:' % info_dict['id'])
2177 self.to_screen(render_table(
2178 ['ID', 'width', 'height', 'URL'],
2179 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
2181 def list_subtitles(self, video_id, subtitles, name='subtitles'):
2183 self.to_screen('%s has no %s' % (video_id, name))
2186 'Available %s for %s:' % (name, video_id))
2187 self.to_screen(render_table(
2188 ['Language', 'formats'],
2189 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
2190 for lang, formats in subtitles.items()]))
2192 def urlopen(self, req):
2193 """ Start an HTTP download """
2194 if isinstance(req, compat_basestring):
2195 req = sanitized_Request(req)
2196 return self._opener.open(req, timeout=self._socket_timeout)
2198 def print_debug_header(self):
2199 if not self.params.get('verbose'):
2202 if type('') is not compat_str:
2203 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
2204 self.report_warning(
2205 'Your Python is broken! Update to a newer and supported version')
2207 stdout_encoding = getattr(
2208 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2210 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2211 locale.getpreferredencoding(),
2212 sys.getfilesystemencoding(),
2214 self.get_encoding()))
2215 write_string(encoding_str, encoding=None)
2217 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
2219 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2221 sp = subprocess.Popen(
2222 ['git', 'rev-parse', '--short', 'HEAD'],
2223 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2224 cwd=os.path.dirname(os.path.abspath(__file__)))
2225 out, err = sp.communicate()
2226 out = out.decode().strip()
2227 if re.match('[0-9a-f]+', out):
2228 self._write_string('[debug] Git HEAD: ' + out + '\n')
2234 self._write_string('[debug] Python version %s - %s\n' % (
2235 platform.python_version(), platform_name()))
2237 exe_versions = FFmpegPostProcessor.get_versions(self)
2238 exe_versions['rtmpdump'] = rtmpdump_version()
2239 exe_versions['phantomjs'] = PhantomJSwrapper._version()
2240 exe_str = ', '.join(
2242 for exe, v in sorted(exe_versions.items())
2247 self._write_string('[debug] exe versions: %s\n' % exe_str)
2250 for handler in self._opener.handlers:
2251 if hasattr(handler, 'proxies'):
2252 proxy_map.update(handler.proxies)
2253 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2255 if self.params.get('call_home', False):
2256 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2257 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2258 latest_version = self.urlopen(
2259 'https://yt-dl.org/latest/version').read().decode('utf-8')
2260 if version_tuple(latest_version) > version_tuple(__version__):
2261 self.report_warning(
2262 'You are using an outdated version (newest version: %s)! '
2263 'See https://yt-dl.org/update if you need help updating.' %
2266 def _setup_opener(self):
2267 timeout_val = self.params.get('socket_timeout')
2268 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2270 opts_cookiefile = self.params.get('cookiefile')
2271 opts_proxy = self.params.get('proxy')
2273 if opts_cookiefile is None:
2274 self.cookiejar = compat_cookiejar.CookieJar()
2276 opts_cookiefile = expand_path(opts_cookiefile)
2277 self.cookiejar = compat_cookiejar.MozillaCookieJar(
2279 if os.access(opts_cookiefile, os.R_OK):
2280 self.cookiejar.load()
2282 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2283 if opts_proxy is not None:
2284 if opts_proxy == '':
2287 proxies = {'http': opts_proxy, 'https': opts_proxy}
2289 proxies = compat_urllib_request.getproxies()
2290 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2291 if 'http' in proxies and 'https' not in proxies:
2292 proxies['https'] = proxies['http']
2293 proxy_handler = PerRequestProxyHandler(proxies)
2295 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2296 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2297 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2298 data_handler = compat_urllib_request_DataHandler()
2300 # When passing our own FileHandler instance, build_opener won't add the
2301 # default FileHandler and allows us to disable the file protocol, which
2302 # can be used for malicious purposes (see
2303 # https://github.com/rg3/youtube-dl/issues/8227)
2304 file_handler = compat_urllib_request.FileHandler()
2306 def file_open(*args, **kwargs):
2307 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2308 file_handler.file_open = file_open
2310 opener = compat_urllib_request.build_opener(
2311 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2313 # Delete the default user-agent header, which would otherwise apply in
2314 # cases where our custom HTTP handler doesn't come into play
2315 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2316 opener.addheaders = []
2317 self._opener = opener
2319 def encode(self, s):
2320 if isinstance(s, bytes):
2321 return s # Already encoded
2324 return s.encode(self.get_encoding())
2325 except UnicodeEncodeError as err:
2326 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2329 def get_encoding(self):
2330 encoding = self.params.get('encoding')
2331 if encoding is None:
2332 encoding = preferredencoding()
2335 def _write_thumbnails(self, info_dict, filename):
2336 if self.params.get('writethumbnail', False):
2337 thumbnails = info_dict.get('thumbnails')
2339 thumbnails = [thumbnails[-1]]
2340 elif self.params.get('write_all_thumbnails', False):
2341 thumbnails = info_dict.get('thumbnails')
2346 # No thumbnails present, so return immediately
2349 for t in thumbnails:
2350 thumb_ext = determine_ext(t['url'], 'jpg')
2351 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2352 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2353 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2355 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2356 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2357 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2359 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2360 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2362 uf = self.urlopen(t['url'])
2363 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2364 shutil.copyfileobj(uf, thumbf)
2365 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2366 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2367 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2368 self.report_warning('Unable to download thumbnail "%s": %s' %
2369 (t['url'], error_to_compat_str(err)))