4 from __future__ import absolute_import, unicode_literals
33 compat_get_terminal_size,
39 compat_tokenize_tokenize,
41 compat_urllib_request,
42 compat_urllib_request_DataHandler,
67 PerRequestProxyHandler,
72 register_socks_protocols,
82 UnavailableVideoError,
87 YoutubeDLCookieProcessor,
90 from .cache import Cache
91 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
92 from .downloader import get_suitable_downloader
93 from .downloader.rtmp import rtmpdump_version
94 from .postprocessor import (
97 FFmpegFixupStretchedPP,
102 from .version import __version__
104 if compat_os_name == 'nt':
108 class YoutubeDL(object):
111 YoutubeDL objects are the ones responsible of downloading the
112 actual video file and writing it to disk if the user has requested
113 it, among some other tasks. In most cases there should be one per
114 program. As, given a video URL, the downloader doesn't know how to
115 extract all the needed information, task that InfoExtractors do, it
116 has to pass the URL to one of them.
118 For this, YoutubeDL objects have a method that allows
119 InfoExtractors to be registered in a given order. When it is passed
120 a URL, the YoutubeDL object handles it to the first InfoExtractor it
121 finds that reports being able to handle it. The InfoExtractor extracts
122 all the information about the video or videos the URL refers to, and
123 YoutubeDL process the extracted information, possibly using a File
124 Downloader to download the video.
126 YoutubeDL objects accept a lot of parameters. In order not to saturate
127 the object constructor with arguments, it receives a dictionary of
128 options instead. These options are available through the params
129 attribute for the InfoExtractors to use. The YoutubeDL also
130 registers itself as the downloader in charge for the InfoExtractors
131 that are added to it, so this is a "mutual registration".
135 username: Username for authentication purposes.
136 password: Password for authentication purposes.
137 videopassword: Password for accessing a video.
138 ap_mso: Adobe Pass multiple-system operator identifier.
139 ap_username: Multiple-system operator account username.
140 ap_password: Multiple-system operator account password.
141 usenetrc: Use netrc for authentication instead.
142 verbose: Print additional info to stdout.
143 quiet: Do not print messages to stdout.
144 no_warnings: Do not print out anything for warnings.
145 forceurl: Force printing final URL.
146 forcetitle: Force printing title.
147 forceid: Force printing ID.
148 forcethumbnail: Force printing thumbnail URL.
149 forcedescription: Force printing description.
150 forcefilename: Force printing final filename.
151 forceduration: Force printing duration.
152 forcejson: Force printing info_dict as JSON.
153 dump_single_json: Force printing the info_dict of the whole playlist
154 (or video) as a single JSON line.
155 simulate: Do not download the video files.
156 format: Video format code. See options.py for more information.
157 outtmpl: Template for output names.
158 restrictfilenames: Do not allow "&" and spaces in file names
159 ignoreerrors: Do not stop on download errors.
160 force_generic_extractor: Force downloader to use the generic extractor
161 nooverwrites: Prevent overwriting files.
162 playliststart: Playlist item to start at.
163 playlistend: Playlist item to end at.
164 playlist_items: Specific indices of playlist to download.
165 playlistreverse: Download playlist items in reverse order.
166 playlistrandom: Download playlist items in random order.
167 matchtitle: Download only matching titles.
168 rejecttitle: Reject downloads for matching titles.
169 logger: Log messages to a logging.Logger instance.
170 logtostderr: Log messages to stderr instead of stdout.
171 writedescription: Write the video description to a .description file
172 writeinfojson: Write the video description to a .info.json file
173 writeannotations: Write the video annotations to a .annotations.xml file
174 writethumbnail: Write the thumbnail image to a file
175 write_all_thumbnails: Write all thumbnail formats to files
176 writesubtitles: Write the video subtitles to a file
177 writeautomaticsub: Write the automatically generated subtitles to a file
178 allsubtitles: Downloads all the subtitles of the video
179 (requires writesubtitles or writeautomaticsub)
180 listsubtitles: Lists all available subtitles for the video
181 subtitlesformat: The format code for subtitles
182 subtitleslangs: List of languages of the subtitles to download
183 keepvideo: Keep the video file after post-processing
184 daterange: A DateRange object, download only if the upload_date is in the range.
185 skip_download: Skip the actual download of the video file
186 cachedir: Location of the cache files in the filesystem.
187 False to disable filesystem cache.
188 noplaylist: Download single video instead of a playlist if in doubt.
189 age_limit: An integer representing the user's age in years.
190 Unsuitable videos for the given age are skipped.
191 min_views: An integer representing the minimum view count the video
192 must have in order to not be skipped.
193 Videos without view count information are always
194 downloaded. None for no limit.
195 max_views: An integer representing the maximum view count.
196 Videos that are more popular than that are not
198 Videos without view count information are always
199 downloaded. None for no limit.
200 download_archive: File name of a file where all downloads are recorded.
201 Videos already present in the file are not downloaded
203 cookiefile: File name where cookies should be read from and dumped to.
204 nocheckcertificate:Do not verify SSL certificates
205 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
206 At the moment, this is only supported by YouTube.
207 proxy: URL of the proxy server to use
208 geo_verification_proxy: URL of the proxy to use for IP address verification
209 on geo-restricted sites. (Experimental)
210 socket_timeout: Time to wait for unresponsive hosts, in seconds
211 bidi_workaround: Work around buggy terminals without bidirectional text
212 support, using fridibi
213 debug_printtraffic:Print out sent and received HTTP traffic
214 include_ads: Download ads as well
215 default_search: Prepend this string if an input url is not valid.
216 'auto' for elaborate guessing
217 encoding: Use this encoding instead of the system-specified.
218 extract_flat: Do not resolve URLs, return the immediate result.
219 Pass in 'in_playlist' to only show this behavior for
221 postprocessors: A list of dictionaries, each with an entry
222 * key: The name of the postprocessor. See
223 youtube_dl/postprocessor/__init__.py for a list.
224 as well as any further keyword arguments for the
226 progress_hooks: A list of functions that get called on download
227 progress, with a dictionary with the entries
228 * status: One of "downloading", "error", or "finished".
229 Check this first and ignore unknown values.
231 If status is one of "downloading", or "finished", the
232 following properties may also be present:
233 * filename: The final filename (always present)
234 * tmpfilename: The filename we're currently writing to
235 * downloaded_bytes: Bytes on disk
236 * total_bytes: Size of the whole file, None if unknown
237 * total_bytes_estimate: Guess of the eventual file size,
239 * elapsed: The number of seconds since download started.
240 * eta: The estimated time in seconds, None if unknown
241 * speed: The download speed in bytes/second, None if
243 * fragment_index: The counter of the currently
244 downloaded video fragment.
245 * fragment_count: The number of fragments (= individual
246 files that will be merged)
248 Progress hooks are guaranteed to be called at least once
249 (with status "finished") if the download is successful.
250 merge_output_format: Extension to use when merging formats.
251 fixup: Automatically correct known faults of the file.
253 - "never": do nothing
254 - "warn": only emit a warning
255 - "detect_or_warn": check whether we can do anything
256 about it, warn otherwise (default)
257 source_address: (Experimental) Client-side IP address to bind to.
258 call_home: Boolean, true iff we are allowed to contact the
259 youtube-dl servers for debugging.
260 sleep_interval: Number of seconds to sleep before each download when
261 used alone or a lower bound of a range for randomized
262 sleep before each download (minimum possible number
263 of seconds to sleep) when used along with
265 max_sleep_interval:Upper bound of a range for randomized sleep before each
266 download (maximum possible number of seconds to sleep).
267 Must only be used along with sleep_interval.
268 Actual sleep time will be a random float from range
269 [sleep_interval; max_sleep_interval].
270 listformats: Print an overview of available video formats and exit.
271 list_thumbnails: Print a table of all thumbnails and exit.
272 match_filter: A function that gets called with the info_dict of
274 If it returns a message, the video is ignored.
275 If it returns None, the video is downloaded.
276 match_filter_func in utils.py is one example for this.
277 no_color: Do not emit color codes in output.
278 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
279 HTTP header (experimental)
281 Two-letter ISO 3166-2 country code that will be used for
282 explicit geographic restriction bypassing via faking
283 X-Forwarded-For HTTP header (experimental)
285 The following options determine which downloader is picked:
286 external_downloader: Executable of the external downloader to call.
287 None or unset for standard (built-in) downloader.
288 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
289 if True, otherwise use ffmpeg/avconv if False, otherwise
290 use downloader suggested by extractor if None.
292 The following parameters are not used by YoutubeDL itself, they are used by
293 the downloader (see youtube_dl/downloader/common.py):
294 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
295 noresizebuffer, retries, continuedl, noprogress, consoletitle,
296 xattr_set_filesize, external_downloader_args, hls_use_mpegts.
298 The following options are used by the post processors:
299 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
300 otherwise prefer avconv.
301 postprocessor_args: A list of additional command-line arguments for the
308 _download_retcode = None
309 _num_downloads = None
312 def __init__(self, params=None, auto_init=True):
313 """Create a FileDownloader object with the given options."""
317 self._ies_instances = {}
319 self._progress_hooks = []
320 self._download_retcode = 0
321 self._num_downloads = 0
322 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
323 self._err_file = sys.stderr
326 'nocheckcertificate': False,
328 self.params.update(params)
329 self.cache = Cache(self)
331 if self.params.get('cn_verification_proxy') is not None:
332 self.report_warning('--cn-verification-proxy is deprecated. Use --geo-verification-proxy instead.')
333 if self.params.get('geo_verification_proxy') is None:
334 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
336 if params.get('bidi_workaround', False):
339 master, slave = pty.openpty()
340 width = compat_get_terminal_size().columns
344 width_args = ['-w', str(width)]
346 stdin=subprocess.PIPE,
348 stderr=self._err_file)
350 self._output_process = subprocess.Popen(
351 ['bidiv'] + width_args, **sp_kwargs
354 self._output_process = subprocess.Popen(
355 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
356 self._output_channel = os.fdopen(master, 'rb')
357 except OSError as ose:
358 if ose.errno == errno.ENOENT:
359 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 $PATH.')
363 if (sys.version_info >= (3,) and sys.platform != 'win32' and
364 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
365 not params.get('restrictfilenames', False)):
366 # On Python 3, the Unicode filesystem API will throw errors (#1474)
368 'Assuming --restrict-filenames since file system encoding '
369 'cannot encode all characters. '
370 'Set the LC_ALL environment variable to fix this.')
371 self.params['restrictfilenames'] = True
373 if isinstance(params.get('outtmpl'), bytes):
375 'Parameter outtmpl is bytes, but should be a unicode string. '
376 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
381 self.print_debug_header()
382 self.add_default_info_extractors()
384 for pp_def_raw in self.params.get('postprocessors', []):
385 pp_class = get_postprocessor(pp_def_raw['key'])
386 pp_def = dict(pp_def_raw)
388 pp = pp_class(self, **compat_kwargs(pp_def))
389 self.add_post_processor(pp)
391 for ph in self.params.get('progress_hooks', []):
392 self.add_progress_hook(ph)
394 register_socks_protocols()
396 def warn_if_short_id(self, argv):
397 # short YouTube ID starting with dash?
399 i for i, a in enumerate(argv)
400 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
404 [a for i, a in enumerate(argv) if i not in idxs] +
405 ['--'] + [argv[i] for i in idxs]
408 'Long argument string detected. '
409 'Use -- to separate parameters and URLs, like this:\n%s\n' %
410 args_to_str(correct_argv))
412 def add_info_extractor(self, ie):
413 """Add an InfoExtractor object to the end of the list."""
415 if not isinstance(ie, type):
416 self._ies_instances[ie.ie_key()] = ie
417 ie.set_downloader(self)
419 def get_info_extractor(self, ie_key):
421 Get an instance of an IE with name ie_key, it will try to get one from
422 the _ies list, if there's no instance it will create a new one and add
423 it to the extractor list.
425 ie = self._ies_instances.get(ie_key)
427 ie = get_info_extractor(ie_key)()
428 self.add_info_extractor(ie)
431 def add_default_info_extractors(self):
433 Add the InfoExtractors returned by gen_extractors to the end of the list
435 for ie in gen_extractor_classes():
436 self.add_info_extractor(ie)
438 def add_post_processor(self, pp):
439 """Add a PostProcessor object to the end of the chain."""
441 pp.set_downloader(self)
443 def add_progress_hook(self, ph):
444 """Add the progress hook (currently only for the file downloader)"""
445 self._progress_hooks.append(ph)
447 def _bidi_workaround(self, message):
448 if not hasattr(self, '_output_channel'):
451 assert hasattr(self, '_output_process')
452 assert isinstance(message, compat_str)
453 line_count = message.count('\n') + 1
454 self._output_process.stdin.write((message + '\n').encode('utf-8'))
455 self._output_process.stdin.flush()
456 res = ''.join(self._output_channel.readline().decode('utf-8')
457 for _ in range(line_count))
458 return res[:-len('\n')]
460 def to_screen(self, message, skip_eol=False):
461 """Print message to stdout if not in quiet mode."""
462 return self.to_stdout(message, skip_eol, check_quiet=True)
464 def _write_string(self, s, out=None):
465 write_string(s, out=out, encoding=self.params.get('encoding'))
467 def to_stdout(self, message, skip_eol=False, check_quiet=False):
468 """Print message to stdout if not in quiet mode."""
469 if self.params.get('logger'):
470 self.params['logger'].debug(message)
471 elif not check_quiet or not self.params.get('quiet', False):
472 message = self._bidi_workaround(message)
473 terminator = ['\n', ''][skip_eol]
474 output = message + terminator
476 self._write_string(output, self._screen_file)
478 def to_stderr(self, message):
479 """Print message to stderr."""
480 assert isinstance(message, compat_str)
481 if self.params.get('logger'):
482 self.params['logger'].error(message)
484 message = self._bidi_workaround(message)
485 output = message + '\n'
486 self._write_string(output, self._err_file)
488 def to_console_title(self, message):
489 if not self.params.get('consoletitle', False):
491 if compat_os_name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
492 # c_wchar_p() might not be necessary if `message` is
493 # already of type unicode()
494 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
495 elif 'TERM' in os.environ:
496 self._write_string('\033]0;%s\007' % message, self._screen_file)
498 def save_console_title(self):
499 if not self.params.get('consoletitle', False):
501 if 'TERM' in os.environ:
502 # Save the title on stack
503 self._write_string('\033[22;0t', self._screen_file)
505 def restore_console_title(self):
506 if not self.params.get('consoletitle', False):
508 if 'TERM' in os.environ:
509 # Restore the title from stack
510 self._write_string('\033[23;0t', self._screen_file)
513 self.save_console_title()
516 def __exit__(self, *args):
517 self.restore_console_title()
519 if self.params.get('cookiefile') is not None:
520 self.cookiejar.save()
522 def trouble(self, message=None, tb=None):
523 """Determine action to take when a download problem appears.
525 Depending on if the downloader has been configured to ignore
526 download errors or not, this method may throw an exception or
527 not when errors are found, after printing the message.
529 tb, if given, is additional traceback information.
531 if message is not None:
532 self.to_stderr(message)
533 if self.params.get('verbose'):
535 if sys.exc_info()[0]: # if .trouble has been called from an except block
537 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
538 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
539 tb += encode_compat_str(traceback.format_exc())
541 tb_data = traceback.format_list(traceback.extract_stack())
542 tb = ''.join(tb_data)
544 if not self.params.get('ignoreerrors', False):
545 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
546 exc_info = sys.exc_info()[1].exc_info
548 exc_info = sys.exc_info()
549 raise DownloadError(message, exc_info)
550 self._download_retcode = 1
552 def report_warning(self, message):
554 Print the message to stderr, it will be prefixed with 'WARNING:'
555 If stderr is a tty file the 'WARNING:' will be colored
557 if self.params.get('logger') is not None:
558 self.params['logger'].warning(message)
560 if self.params.get('no_warnings'):
562 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
563 _msg_header = '\033[0;33mWARNING:\033[0m'
565 _msg_header = 'WARNING:'
566 warning_message = '%s %s' % (_msg_header, message)
567 self.to_stderr(warning_message)
569 def report_error(self, message, tb=None):
571 Do the same as trouble, but prefixes the message with 'ERROR:', colored
572 in red if stderr is a tty file.
574 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
575 _msg_header = '\033[0;31mERROR:\033[0m'
577 _msg_header = 'ERROR:'
578 error_message = '%s %s' % (_msg_header, message)
579 self.trouble(error_message, tb)
581 def report_file_already_downloaded(self, file_name):
582 """Report file has already been fully downloaded."""
584 self.to_screen('[download] %s has already been downloaded' % file_name)
585 except UnicodeEncodeError:
586 self.to_screen('[download] The file has already been downloaded')
588 def prepare_filename(self, info_dict):
589 """Generate the output filename."""
591 template_dict = dict(info_dict)
593 template_dict['epoch'] = int(time.time())
594 autonumber_size = self.params.get('autonumber_size')
595 if autonumber_size is None:
597 template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
598 if template_dict.get('resolution') is None:
599 if template_dict.get('width') and template_dict.get('height'):
600 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
601 elif template_dict.get('height'):
602 template_dict['resolution'] = '%sp' % template_dict['height']
603 elif template_dict.get('width'):
604 template_dict['resolution'] = '%dx?' % template_dict['width']
606 sanitize = lambda k, v: sanitize_filename(
608 restricted=self.params.get('restrictfilenames'),
610 template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
611 for k, v in template_dict.items()
612 if v is not None and not isinstance(v, (list, tuple, dict)))
613 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
615 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
617 # For fields playlist_index and autonumber convert all occurrences
618 # of %(field)s to %(field)0Nd for backward compatibility
619 field_size_compat_map = {
620 'playlist_index': len(str(template_dict['n_entries'])),
621 'autonumber': autonumber_size,
623 FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
624 mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
627 FIELD_SIZE_COMPAT_RE,
628 r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
631 NUMERIC_FIELDS = set((
632 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
633 'upload_year', 'upload_month', 'upload_day',
634 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
635 'average_rating', 'comment_count', 'age_limit',
636 'start_time', 'end_time',
637 'chapter_number', 'season_number', 'episode_number',
638 'track_number', 'disc_number', 'release_year',
642 # Missing numeric fields used together with integer presentation types
643 # in format specification will break the argument substitution since
644 # string 'NA' is returned for missing fields. We will patch output
645 # template for missing fields to meet string presentation type.
646 for numeric_field in NUMERIC_FIELDS:
647 if numeric_field not in template_dict:
648 # As of [1] format syntax is:
649 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
650 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
654 \({0}\) # mapping key
655 (?:[#0\-+ ]+)? # conversion flags (optional)
656 (?:\d+)? # minimum field width (optional)
657 (?:\.\d+)? # precision (optional)
658 [hlL]? # length modifier (optional)
659 [diouxXeEfFgGcrs%] # conversion type
662 FORMAT_RE.format(numeric_field),
663 r'%({0})s'.format(numeric_field), outtmpl)
665 tmpl = compat_expanduser(outtmpl)
666 filename = tmpl % template_dict
667 # Temporary fix for #4787
668 # 'Treat' all problem characters by passing filename through preferredencoding
669 # to workaround encoding issues with subprocess on python2 @ Windows
670 if sys.version_info < (3, 0) and sys.platform == 'win32':
671 filename = encodeFilename(filename, True).decode(preferredencoding())
672 return sanitize_path(filename)
673 except ValueError as err:
674 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
677 def _match_entry(self, info_dict, incomplete):
678 """ Returns None iff the file should be downloaded """
680 video_title = info_dict.get('title', info_dict.get('id', 'video'))
681 if 'title' in info_dict:
682 # This can happen when we're just evaluating the playlist
683 title = info_dict['title']
684 matchtitle = self.params.get('matchtitle', False)
686 if not re.search(matchtitle, title, re.IGNORECASE):
687 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
688 rejecttitle = self.params.get('rejecttitle', False)
690 if re.search(rejecttitle, title, re.IGNORECASE):
691 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
692 date = info_dict.get('upload_date')
694 dateRange = self.params.get('daterange', DateRange())
695 if date not in dateRange:
696 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
697 view_count = info_dict.get('view_count')
698 if view_count is not None:
699 min_views = self.params.get('min_views')
700 if min_views is not None and view_count < min_views:
701 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
702 max_views = self.params.get('max_views')
703 if max_views is not None and view_count > max_views:
704 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
705 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
706 return 'Skipping "%s" because it is age restricted' % video_title
707 if self.in_download_archive(info_dict):
708 return '%s has already been recorded in archive' % video_title
711 match_filter = self.params.get('match_filter')
712 if match_filter is not None:
713 ret = match_filter(info_dict)
720 def add_extra_info(info_dict, extra_info):
721 '''Set the keys from extra_info in info dict if they are missing'''
722 for key, value in extra_info.items():
723 info_dict.setdefault(key, value)
725 def extract_info(self, url, download=True, ie_key=None, extra_info={},
726 process=True, force_generic_extractor=False):
728 Returns a list with a dictionary for each video we find.
729 If 'download', also downloads the videos.
730 extra_info is a dict containing the extra values to add to each result
733 if not ie_key and force_generic_extractor:
737 ies = [self.get_info_extractor(ie_key)]
742 if not ie.suitable(url):
745 ie = self.get_info_extractor(ie.ie_key())
747 self.report_warning('The program functionality for this site has been marked as broken, '
748 'and will probably not work.')
751 ie_result = ie.extract(url)
752 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
754 if isinstance(ie_result, list):
755 # Backwards compatibility: old IE result format
757 '_type': 'compat_list',
758 'entries': ie_result,
760 self.add_default_extra_info(ie_result, ie, url)
762 return self.process_ie_result(ie_result, download, extra_info)
765 except GeoRestrictedError as e:
768 msg += '\nThis video is available in %s.' % ', '.join(
769 map(ISO3166Utils.short2full, e.countries))
770 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
771 self.report_error(msg)
773 except ExtractorError as e: # An error we somewhat expected
774 self.report_error(compat_str(e), e.format_traceback())
776 except MaxDownloadsReached:
778 except Exception as e:
779 if self.params.get('ignoreerrors', False):
780 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
785 self.report_error('no suitable InfoExtractor for URL %s' % url)
787 def add_default_extra_info(self, ie_result, ie, url):
788 self.add_extra_info(ie_result, {
789 'extractor': ie.IE_NAME,
791 'webpage_url_basename': url_basename(url),
792 'extractor_key': ie.ie_key(),
795 def process_ie_result(self, ie_result, download=True, extra_info={}):
797 Take the result of the ie(may be modified) and resolve all unresolved
798 references (URLs, playlist items).
800 It will also download the videos if 'download'.
801 Returns the resolved ie_result.
803 result_type = ie_result.get('_type', 'video')
805 if result_type in ('url', 'url_transparent'):
806 ie_result['url'] = sanitize_url(ie_result['url'])
807 extract_flat = self.params.get('extract_flat', False)
808 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
809 extract_flat is True):
810 if self.params.get('forcejson', False):
811 self.to_stdout(json.dumps(ie_result))
814 if result_type == 'video':
815 self.add_extra_info(ie_result, extra_info)
816 return self.process_video_result(ie_result, download=download)
817 elif result_type == 'url':
818 # We have to add extra_info to the results because it may be
819 # contained in a playlist
820 return self.extract_info(ie_result['url'],
822 ie_key=ie_result.get('ie_key'),
823 extra_info=extra_info)
824 elif result_type == 'url_transparent':
825 # Use the information from the embedding page
826 info = self.extract_info(
827 ie_result['url'], ie_key=ie_result.get('ie_key'),
828 extra_info=extra_info, download=False, process=False)
830 force_properties = dict(
831 (k, v) for k, v in ie_result.items() if v is not None)
832 for f in ('_type', 'url', 'ie_key'):
833 if f in force_properties:
834 del force_properties[f]
835 new_result = info.copy()
836 new_result.update(force_properties)
838 assert new_result.get('_type') != 'url_transparent'
840 return self.process_ie_result(
841 new_result, download=download, extra_info=extra_info)
842 elif result_type == 'playlist' or result_type == 'multi_video':
843 # We process each entry in the playlist
844 playlist = ie_result.get('title') or ie_result.get('id')
845 self.to_screen('[download] Downloading playlist: %s' % playlist)
847 playlist_results = []
849 playliststart = self.params.get('playliststart', 1) - 1
850 playlistend = self.params.get('playlistend')
851 # For backwards compatibility, interpret -1 as whole list
852 if playlistend == -1:
855 playlistitems_str = self.params.get('playlist_items')
857 if playlistitems_str is not None:
858 def iter_playlistitems(format):
859 for string_segment in format.split(','):
860 if '-' in string_segment:
861 start, end = string_segment.split('-')
862 for item in range(int(start), int(end) + 1):
865 yield int(string_segment)
866 playlistitems = iter_playlistitems(playlistitems_str)
868 ie_entries = ie_result['entries']
869 if isinstance(ie_entries, list):
870 n_all_entries = len(ie_entries)
873 ie_entries[i - 1] for i in playlistitems
874 if -n_all_entries <= i - 1 < n_all_entries]
876 entries = ie_entries[playliststart:playlistend]
877 n_entries = len(entries)
879 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
880 (ie_result['extractor'], playlist, n_all_entries, n_entries))
881 elif isinstance(ie_entries, PagedList):
884 for item in playlistitems:
885 entries.extend(ie_entries.getslice(
889 entries = ie_entries.getslice(
890 playliststart, playlistend)
891 n_entries = len(entries)
893 '[%s] playlist %s: Downloading %d videos' %
894 (ie_result['extractor'], playlist, n_entries))
897 entry_list = list(ie_entries)
898 entries = [entry_list[i - 1] for i in playlistitems]
900 entries = list(itertools.islice(
901 ie_entries, playliststart, playlistend))
902 n_entries = len(entries)
904 '[%s] playlist %s: Downloading %d videos' %
905 (ie_result['extractor'], playlist, n_entries))
907 if self.params.get('playlistreverse', False):
908 entries = entries[::-1]
910 if self.params.get('playlistrandom', False):
911 random.shuffle(entries)
913 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
915 for i, entry in enumerate(entries, 1):
916 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
917 # This __x_forwarded_for_ip thing is a bit ugly but requires
920 entry['__x_forwarded_for_ip'] = x_forwarded_for
922 'n_entries': n_entries,
923 'playlist': playlist,
924 'playlist_id': ie_result.get('id'),
925 'playlist_title': ie_result.get('title'),
926 'playlist_index': i + playliststart,
927 'extractor': ie_result['extractor'],
928 'webpage_url': ie_result['webpage_url'],
929 'webpage_url_basename': url_basename(ie_result['webpage_url']),
930 'extractor_key': ie_result['extractor_key'],
933 reason = self._match_entry(entry, incomplete=True)
934 if reason is not None:
935 self.to_screen('[download] ' + reason)
938 entry_result = self.process_ie_result(entry,
941 playlist_results.append(entry_result)
942 ie_result['entries'] = playlist_results
943 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
945 elif result_type == 'compat_list':
947 'Extractor %s returned a compat_list result. '
948 'It needs to be updated.' % ie_result.get('extractor'))
954 'extractor': ie_result['extractor'],
955 'webpage_url': ie_result['webpage_url'],
956 'webpage_url_basename': url_basename(ie_result['webpage_url']),
957 'extractor_key': ie_result['extractor_key'],
961 ie_result['entries'] = [
962 self.process_ie_result(_fixup(r), download, extra_info)
963 for r in ie_result['entries']
967 raise Exception('Invalid result type: %s' % result_type)
969 def _build_format_filter(self, filter_spec):
970 " Returns a function to filter the formats according to the filter_spec "
980 operator_rex = re.compile(r'''(?x)\s*
981 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
982 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
983 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
985 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
986 m = operator_rex.search(filter_spec)
989 comparison_value = int(m.group('value'))
991 comparison_value = parse_filesize(m.group('value'))
992 if comparison_value is None:
993 comparison_value = parse_filesize(m.group('value') + 'B')
994 if comparison_value is None:
996 'Invalid value %r in format specification %r' % (
997 m.group('value'), filter_spec))
998 op = OPERATORS[m.group('op')]
1004 '^=': lambda attr, value: attr.startswith(value),
1005 '$=': lambda attr, value: attr.endswith(value),
1006 '*=': lambda attr, value: value in attr,
1008 str_operator_rex = re.compile(r'''(?x)
1009 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
1010 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
1011 \s*(?P<value>[a-zA-Z0-9._-]+)
1013 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1014 m = str_operator_rex.search(filter_spec)
1016 comparison_value = m.group('value')
1017 op = STR_OPERATORS[m.group('op')]
1020 raise ValueError('Invalid filter specification %r' % filter_spec)
1023 actual_value = f.get(m.group('key'))
1024 if actual_value is None:
1025 return m.group('none_inclusive')
1026 return op(actual_value, comparison_value)
1029 def build_format_selector(self, format_spec):
1030 def syntax_error(note, start):
1032 'Invalid format specification: '
1033 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1034 return SyntaxError(message)
1036 PICKFIRST = 'PICKFIRST'
1040 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1042 def _parse_filter(tokens):
1044 for type, string, start, _, _ in tokens:
1045 if type == tokenize.OP and string == ']':
1046 return ''.join(filter_parts)
1048 filter_parts.append(string)
1050 def _remove_unused_ops(tokens):
1051 # Remove operators that we don't use and join them with the surrounding strings
1052 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1053 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1054 last_string, last_start, last_end, last_line = None, None, None, None
1055 for type, string, start, end, line in tokens:
1056 if type == tokenize.OP and string == '[':
1058 yield tokenize.NAME, last_string, last_start, last_end, last_line
1060 yield type, string, start, end, line
1061 # everything inside brackets will be handled by _parse_filter
1062 for type, string, start, end, line in tokens:
1063 yield type, string, start, end, line
1064 if type == tokenize.OP and string == ']':
1066 elif type == tokenize.OP and string in ALLOWED_OPS:
1068 yield tokenize.NAME, last_string, last_start, last_end, last_line
1070 yield type, string, start, end, line
1071 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1073 last_string = string
1077 last_string += string
1079 yield tokenize.NAME, last_string, last_start, last_end, last_line
1081 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1083 current_selector = None
1084 for type, string, start, _, _ in tokens:
1085 # ENCODING is only defined in python 3.x
1086 if type == getattr(tokenize, 'ENCODING', None):
1088 elif type in [tokenize.NAME, tokenize.NUMBER]:
1089 current_selector = FormatSelector(SINGLE, string, [])
1090 elif type == tokenize.OP:
1092 if not inside_group:
1093 # ')' will be handled by the parentheses group
1094 tokens.restore_last_token()
1096 elif inside_merge and string in ['/', ',']:
1097 tokens.restore_last_token()
1099 elif inside_choice and string == ',':
1100 tokens.restore_last_token()
1103 if not current_selector:
1104 raise syntax_error('"," must follow a format selector', start)
1105 selectors.append(current_selector)
1106 current_selector = None
1108 if not current_selector:
1109 raise syntax_error('"/" must follow a format selector', start)
1110 first_choice = current_selector
1111 second_choice = _parse_format_selection(tokens, inside_choice=True)
1112 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1114 if not current_selector:
1115 current_selector = FormatSelector(SINGLE, 'best', [])
1116 format_filter = _parse_filter(tokens)
1117 current_selector.filters.append(format_filter)
1119 if current_selector:
1120 raise syntax_error('Unexpected "("', start)
1121 group = _parse_format_selection(tokens, inside_group=True)
1122 current_selector = FormatSelector(GROUP, group, [])
1124 video_selector = current_selector
1125 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1126 if not video_selector or not audio_selector:
1127 raise syntax_error('"+" must be between two format selectors', start)
1128 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1130 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1131 elif type == tokenize.ENDMARKER:
1133 if current_selector:
1134 selectors.append(current_selector)
1137 def _build_selector_function(selector):
1138 if isinstance(selector, list):
1139 fs = [_build_selector_function(s) for s in selector]
1141 def selector_function(ctx):
1143 for format in f(ctx):
1145 return selector_function
1146 elif selector.type == GROUP:
1147 selector_function = _build_selector_function(selector.selector)
1148 elif selector.type == PICKFIRST:
1149 fs = [_build_selector_function(s) for s in selector.selector]
1151 def selector_function(ctx):
1153 picked_formats = list(f(ctx))
1155 return picked_formats
1157 elif selector.type == SINGLE:
1158 format_spec = selector.selector
1160 def selector_function(ctx):
1161 formats = list(ctx['formats'])
1164 if format_spec == 'all':
1167 elif format_spec in ['best', 'worst', None]:
1168 format_idx = 0 if format_spec == 'worst' else -1
1169 audiovideo_formats = [
1171 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1172 if audiovideo_formats:
1173 yield audiovideo_formats[format_idx]
1174 # for extractors with incomplete formats (audio only (soundcloud)
1175 # or video only (imgur)) we will fallback to best/worst
1176 # {video,audio}-only format
1177 elif ctx['incomplete_formats']:
1178 yield formats[format_idx]
1179 elif format_spec == 'bestaudio':
1182 if f.get('vcodec') == 'none']
1184 yield audio_formats[-1]
1185 elif format_spec == 'worstaudio':
1188 if f.get('vcodec') == 'none']
1190 yield audio_formats[0]
1191 elif format_spec == 'bestvideo':
1194 if f.get('acodec') == 'none']
1196 yield video_formats[-1]
1197 elif format_spec == 'worstvideo':
1200 if f.get('acodec') == 'none']
1202 yield video_formats[0]
1204 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1205 if format_spec in extensions:
1206 filter_f = lambda f: f['ext'] == format_spec
1208 filter_f = lambda f: f['format_id'] == format_spec
1209 matches = list(filter(filter_f, formats))
1212 elif selector.type == MERGE:
1213 def _merge(formats_info):
1214 format_1, format_2 = [f['format_id'] for f in formats_info]
1215 # The first format must contain the video and the
1217 if formats_info[0].get('vcodec') == 'none':
1218 self.report_error('The first format must '
1219 'contain the video, try using '
1220 '"-f %s+%s"' % (format_2, format_1))
1222 # Formats must be opposite (video+audio)
1223 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1225 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1226 % (format_1, format_2))
1229 formats_info[0]['ext']
1230 if self.params.get('merge_output_format') is None
1231 else self.params['merge_output_format'])
1233 'requested_formats': formats_info,
1234 'format': '%s+%s' % (formats_info[0].get('format'),
1235 formats_info[1].get('format')),
1236 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1237 formats_info[1].get('format_id')),
1238 'width': formats_info[0].get('width'),
1239 'height': formats_info[0].get('height'),
1240 'resolution': formats_info[0].get('resolution'),
1241 'fps': formats_info[0].get('fps'),
1242 'vcodec': formats_info[0].get('vcodec'),
1243 'vbr': formats_info[0].get('vbr'),
1244 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1245 'acodec': formats_info[1].get('acodec'),
1246 'abr': formats_info[1].get('abr'),
1249 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1251 def selector_function(ctx):
1252 for pair in itertools.product(
1253 video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
1256 filters = [self._build_format_filter(f) for f in selector.filters]
1258 def final_selector(ctx):
1259 ctx_copy = copy.deepcopy(ctx)
1260 for _filter in filters:
1261 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1262 return selector_function(ctx_copy)
1263 return final_selector
1265 stream = io.BytesIO(format_spec.encode('utf-8'))
1267 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1268 except tokenize.TokenError:
1269 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1271 class TokenIterator(object):
1272 def __init__(self, tokens):
1273 self.tokens = tokens
1280 if self.counter >= len(self.tokens):
1281 raise StopIteration()
1282 value = self.tokens[self.counter]
1288 def restore_last_token(self):
1291 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1292 return _build_selector_function(parsed_selector)
1294 def _calc_headers(self, info_dict):
1295 res = std_headers.copy()
1297 add_headers = info_dict.get('http_headers')
1299 res.update(add_headers)
1301 cookies = self._calc_cookies(info_dict)
1303 res['Cookie'] = cookies
1305 if 'X-Forwarded-For' not in res:
1306 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1307 if x_forwarded_for_ip:
1308 res['X-Forwarded-For'] = x_forwarded_for_ip
1312 def _calc_cookies(self, info_dict):
1313 pr = sanitized_Request(info_dict['url'])
1314 self.cookiejar.add_cookie_header(pr)
1315 return pr.get_header('Cookie')
1317 def process_video_result(self, info_dict, download=True):
1318 assert info_dict.get('_type', 'video') == 'video'
1320 if 'id' not in info_dict:
1321 raise ExtractorError('Missing "id" field in extractor result')
1322 if 'title' not in info_dict:
1323 raise ExtractorError('Missing "title" field in extractor result')
1325 if not isinstance(info_dict['id'], compat_str):
1326 self.report_warning('"id" field is not a string - forcing string conversion')
1327 info_dict['id'] = compat_str(info_dict['id'])
1329 if 'playlist' not in info_dict:
1330 # It isn't part of a playlist
1331 info_dict['playlist'] = None
1332 info_dict['playlist_index'] = None
1334 thumbnails = info_dict.get('thumbnails')
1335 if thumbnails is None:
1336 thumbnail = info_dict.get('thumbnail')
1338 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1340 thumbnails.sort(key=lambda t: (
1341 t.get('preference') if t.get('preference') is not None else -1,
1342 t.get('width') if t.get('width') is not None else -1,
1343 t.get('height') if t.get('height') is not None else -1,
1344 t.get('id') if t.get('id') is not None else '', t.get('url')))
1345 for i, t in enumerate(thumbnails):
1346 t['url'] = sanitize_url(t['url'])
1347 if t.get('width') and t.get('height'):
1348 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1349 if t.get('id') is None:
1352 if self.params.get('list_thumbnails'):
1353 self.list_thumbnails(info_dict)
1356 thumbnail = info_dict.get('thumbnail')
1358 info_dict['thumbnail'] = sanitize_url(thumbnail)
1360 info_dict['thumbnail'] = thumbnails[-1]['url']
1362 if 'display_id' not in info_dict and 'id' in info_dict:
1363 info_dict['display_id'] = info_dict['id']
1365 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1366 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1367 # see http://bugs.python.org/issue1646728)
1369 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1370 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1371 except (ValueError, OverflowError, OSError):
1374 # Auto generate title fields corresponding to the *_number fields when missing
1375 # in order to always have clean titles. This is very common for TV series.
1376 for field in ('chapter', 'season', 'episode'):
1377 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1378 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1380 subtitles = info_dict.get('subtitles')
1382 for _, subtitle in subtitles.items():
1383 for subtitle_format in subtitle:
1384 if subtitle_format.get('url'):
1385 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1386 if subtitle_format.get('ext') is None:
1387 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1389 if self.params.get('listsubtitles', False):
1390 if 'automatic_captions' in info_dict:
1391 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1392 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1394 info_dict['requested_subtitles'] = self.process_subtitles(
1395 info_dict['id'], subtitles,
1396 info_dict.get('automatic_captions'))
1398 # We now pick which formats have to be downloaded
1399 if info_dict.get('formats') is None:
1400 # There's only one format available
1401 formats = [info_dict]
1403 formats = info_dict['formats']
1406 raise ExtractorError('No video formats found!')
1410 # We check that all the formats have the format and format_id fields
1411 for i, format in enumerate(formats):
1412 if 'url' not in format:
1413 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1415 format['url'] = sanitize_url(format['url'])
1417 if format.get('format_id') is None:
1418 format['format_id'] = compat_str(i)
1420 # Sanitize format_id from characters used in format selector expression
1421 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
1422 format_id = format['format_id']
1423 if format_id not in formats_dict:
1424 formats_dict[format_id] = []
1425 formats_dict[format_id].append(format)
1427 # Make sure all formats have unique format_id
1428 for format_id, ambiguous_formats in formats_dict.items():
1429 if len(ambiguous_formats) > 1:
1430 for i, format in enumerate(ambiguous_formats):
1431 format['format_id'] = '%s-%d' % (format_id, i)
1433 for i, format in enumerate(formats):
1434 if format.get('format') is None:
1435 format['format'] = '{id} - {res}{note}'.format(
1436 id=format['format_id'],
1437 res=self.format_resolution(format),
1438 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1440 # Automatically determine file extension if missing
1441 if format.get('ext') is None:
1442 format['ext'] = determine_ext(format['url']).lower()
1443 # Automatically determine protocol if missing (useful for format
1444 # selection purposes)
1445 if format.get('protocol') is None:
1446 format['protocol'] = determine_protocol(format)
1447 # Add HTTP headers, so that external programs can use them from the
1449 full_format_info = info_dict.copy()
1450 full_format_info.update(format)
1451 format['http_headers'] = self._calc_headers(full_format_info)
1452 # Remove private housekeeping stuff
1453 if '__x_forwarded_for_ip' in info_dict:
1454 del info_dict['__x_forwarded_for_ip']
1456 # TODO Central sorting goes here
1458 if formats[0] is not info_dict:
1459 # only set the 'formats' fields if the original info_dict list them
1460 # otherwise we end up with a circular reference, the first (and unique)
1461 # element in the 'formats' field in info_dict is info_dict itself,
1462 # which can't be exported to json
1463 info_dict['formats'] = formats
1464 if self.params.get('listformats'):
1465 self.list_formats(info_dict)
1468 req_format = self.params.get('format')
1469 if req_format is None:
1470 req_format_list = []
1471 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1472 not info_dict.get('is_live')):
1473 merger = FFmpegMergerPP(self)
1474 if merger.available and merger.can_merge():
1475 req_format_list.append('bestvideo+bestaudio')
1476 req_format_list.append('best')
1477 req_format = '/'.join(req_format_list)
1478 format_selector = self.build_format_selector(req_format)
1480 # While in format selection we may need to have an access to the original
1481 # format set in order to calculate some metrics or do some processing.
1482 # For now we need to be able to guess whether original formats provided
1483 # by extractor are incomplete or not (i.e. whether extractor provides only
1484 # video-only or audio-only formats) for proper formats selection for
1485 # extractors with such incomplete formats (see
1486 # https://github.com/rg3/youtube-dl/pull/5556).
1487 # Since formats may be filtered during format selection and may not match
1488 # the original formats the results may be incorrect. Thus original formats
1489 # or pre-calculated metrics should be passed to format selection routines
1491 # We will pass a context object containing all necessary additional data
1492 # instead of just formats.
1493 # This fixes incorrect format selection issue (see
1494 # https://github.com/rg3/youtube-dl/issues/10083).
1495 incomplete_formats = (
1496 # All formats are video-only or
1497 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) or
1498 # all formats are audio-only
1499 all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1503 'incomplete_formats': incomplete_formats,
1506 formats_to_download = list(format_selector(ctx))
1507 if not formats_to_download:
1508 raise ExtractorError('requested format not available',
1512 if len(formats_to_download) > 1:
1513 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1514 for format in formats_to_download:
1515 new_info = dict(info_dict)
1516 new_info.update(format)
1517 self.process_info(new_info)
1518 # We update the info dict with the best quality format (backwards compatibility)
1519 info_dict.update(formats_to_download[-1])
1522 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1523 """Select the requested subtitles and their format"""
1525 if normal_subtitles and self.params.get('writesubtitles'):
1526 available_subs.update(normal_subtitles)
1527 if automatic_captions and self.params.get('writeautomaticsub'):
1528 for lang, cap_info in automatic_captions.items():
1529 if lang not in available_subs:
1530 available_subs[lang] = cap_info
1532 if (not self.params.get('writesubtitles') and not
1533 self.params.get('writeautomaticsub') or not
1537 if self.params.get('allsubtitles', False):
1538 requested_langs = available_subs.keys()
1540 if self.params.get('subtitleslangs', False):
1541 requested_langs = self.params.get('subtitleslangs')
1542 elif 'en' in available_subs:
1543 requested_langs = ['en']
1545 requested_langs = [list(available_subs.keys())[0]]
1547 formats_query = self.params.get('subtitlesformat', 'best')
1548 formats_preference = formats_query.split('/') if formats_query else []
1550 for lang in requested_langs:
1551 formats = available_subs.get(lang)
1553 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1555 for ext in formats_preference:
1559 matches = list(filter(lambda f: f['ext'] == ext, formats))
1565 self.report_warning(
1566 'No subtitle format found matching "%s" for language %s, '
1567 'using %s' % (formats_query, lang, f['ext']))
1571 def process_info(self, info_dict):
1572 """Process a single resolved IE result."""
1574 assert info_dict.get('_type', 'video') == 'video'
1576 max_downloads = self.params.get('max_downloads')
1577 if max_downloads is not None:
1578 if self._num_downloads >= int(max_downloads):
1579 raise MaxDownloadsReached()
1581 info_dict['fulltitle'] = info_dict['title']
1582 if len(info_dict['title']) > 200:
1583 info_dict['title'] = info_dict['title'][:197] + '...'
1585 if 'format' not in info_dict:
1586 info_dict['format'] = info_dict['ext']
1588 reason = self._match_entry(info_dict, incomplete=False)
1589 if reason is not None:
1590 self.to_screen('[download] ' + reason)
1593 self._num_downloads += 1
1595 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1598 if self.params.get('forcetitle', False):
1599 self.to_stdout(info_dict['fulltitle'])
1600 if self.params.get('forceid', False):
1601 self.to_stdout(info_dict['id'])
1602 if self.params.get('forceurl', False):
1603 if info_dict.get('requested_formats') is not None:
1604 for f in info_dict['requested_formats']:
1605 self.to_stdout(f['url'] + f.get('play_path', ''))
1607 # For RTMP URLs, also include the playpath
1608 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1609 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1610 self.to_stdout(info_dict['thumbnail'])
1611 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1612 self.to_stdout(info_dict['description'])
1613 if self.params.get('forcefilename', False) and filename is not None:
1614 self.to_stdout(filename)
1615 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1616 self.to_stdout(formatSeconds(info_dict['duration']))
1617 if self.params.get('forceformat', False):
1618 self.to_stdout(info_dict['format'])
1619 if self.params.get('forcejson', False):
1620 self.to_stdout(json.dumps(info_dict))
1622 # Do nothing else if in simulate mode
1623 if self.params.get('simulate', False):
1626 if filename is None:
1630 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1631 if dn and not os.path.exists(dn):
1633 except (OSError, IOError) as err:
1634 self.report_error('unable to create directory ' + error_to_compat_str(err))
1637 if self.params.get('writedescription', False):
1638 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1639 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1640 self.to_screen('[info] Video description is already present')
1641 elif info_dict.get('description') is None:
1642 self.report_warning('There\'s no description to write.')
1645 self.to_screen('[info] Writing video description to: ' + descfn)
1646 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1647 descfile.write(info_dict['description'])
1648 except (OSError, IOError):
1649 self.report_error('Cannot write description file ' + descfn)
1652 if self.params.get('writeannotations', False):
1653 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1654 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1655 self.to_screen('[info] Video annotations are already present')
1658 self.to_screen('[info] Writing video annotations to: ' + annofn)
1659 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1660 annofile.write(info_dict['annotations'])
1661 except (KeyError, TypeError):
1662 self.report_warning('There are no annotations to write.')
1663 except (OSError, IOError):
1664 self.report_error('Cannot write annotations file: ' + annofn)
1667 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1668 self.params.get('writeautomaticsub')])
1670 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1671 # subtitles download errors are already managed as troubles in relevant IE
1672 # that way it will silently go on when used with unsupporting IE
1673 subtitles = info_dict['requested_subtitles']
1674 ie = self.get_info_extractor(info_dict['extractor_key'])
1675 for sub_lang, sub_info in subtitles.items():
1676 sub_format = sub_info['ext']
1677 if sub_info.get('data') is not None:
1678 sub_data = sub_info['data']
1681 sub_data = ie._download_webpage(
1682 sub_info['url'], info_dict['id'], note=False)
1683 except ExtractorError as err:
1684 self.report_warning('Unable to download subtitle for "%s": %s' %
1685 (sub_lang, error_to_compat_str(err.cause)))
1688 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1689 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1690 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1692 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1693 # Use newline='' to prevent conversion of newline characters
1694 # See https://github.com/rg3/youtube-dl/issues/10268
1695 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
1696 subfile.write(sub_data)
1697 except (OSError, IOError):
1698 self.report_error('Cannot write subtitles file ' + sub_filename)
1701 if self.params.get('writeinfojson', False):
1702 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1703 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1704 self.to_screen('[info] Video description metadata is already present')
1706 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1708 write_json_file(self.filter_requested_info(info_dict), infofn)
1709 except (OSError, IOError):
1710 self.report_error('Cannot write metadata to JSON file ' + infofn)
1713 self._write_thumbnails(info_dict, filename)
1715 if not self.params.get('skip_download', False):
1718 fd = get_suitable_downloader(info, self.params)(self, self.params)
1719 for ph in self._progress_hooks:
1720 fd.add_progress_hook(ph)
1721 if self.params.get('verbose'):
1722 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1723 return fd.download(name, info)
1725 if info_dict.get('requested_formats') is not None:
1728 merger = FFmpegMergerPP(self)
1729 if not merger.available:
1731 self.report_warning('You have requested multiple '
1732 'formats but ffmpeg or avconv are not installed.'
1733 ' The formats won\'t be merged.')
1735 postprocessors = [merger]
1737 def compatible_formats(formats):
1738 video, audio = formats
1740 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1741 if video_ext and audio_ext:
1743 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'),
1746 for exts in COMPATIBLE_EXTS:
1747 if video_ext in exts and audio_ext in exts:
1749 # TODO: Check acodec/vcodec
1752 filename_real_ext = os.path.splitext(filename)[1][1:]
1754 os.path.splitext(filename)[0]
1755 if filename_real_ext == info_dict['ext']
1757 requested_formats = info_dict['requested_formats']
1758 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1759 info_dict['ext'] = 'mkv'
1760 self.report_warning(
1761 'Requested formats are incompatible for merge and will be merged into mkv.')
1762 # Ensure filename always has a correct extension for successful merge
1763 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1764 if os.path.exists(encodeFilename(filename)):
1766 '[download] %s has already been downloaded and '
1767 'merged' % filename)
1769 for f in requested_formats:
1770 new_info = dict(info_dict)
1772 fname = self.prepare_filename(new_info)
1773 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1774 downloaded.append(fname)
1775 partial_success = dl(fname, new_info)
1776 success = success and partial_success
1777 info_dict['__postprocessors'] = postprocessors
1778 info_dict['__files_to_merge'] = downloaded
1780 # Just a single file
1781 success = dl(filename, info_dict)
1782 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1783 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1785 except (OSError, IOError) as err:
1786 raise UnavailableVideoError(err)
1787 except (ContentTooShortError, ) as err:
1788 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1791 if success and filename != '-':
1793 fixup_policy = self.params.get('fixup')
1794 if fixup_policy is None:
1795 fixup_policy = 'detect_or_warn'
1797 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1799 stretched_ratio = info_dict.get('stretched_ratio')
1800 if stretched_ratio is not None and stretched_ratio != 1:
1801 if fixup_policy == 'warn':
1802 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1803 info_dict['id'], stretched_ratio))
1804 elif fixup_policy == 'detect_or_warn':
1805 stretched_pp = FFmpegFixupStretchedPP(self)
1806 if stretched_pp.available:
1807 info_dict.setdefault('__postprocessors', [])
1808 info_dict['__postprocessors'].append(stretched_pp)
1810 self.report_warning(
1811 '%s: Non-uniform pixel ratio (%s). %s'
1812 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1814 assert fixup_policy in ('ignore', 'never')
1816 if (info_dict.get('requested_formats') is None and
1817 info_dict.get('container') == 'm4a_dash'):
1818 if fixup_policy == 'warn':
1819 self.report_warning(
1820 '%s: writing DASH m4a. '
1821 'Only some players support this container.'
1823 elif fixup_policy == 'detect_or_warn':
1824 fixup_pp = FFmpegFixupM4aPP(self)
1825 if fixup_pp.available:
1826 info_dict.setdefault('__postprocessors', [])
1827 info_dict['__postprocessors'].append(fixup_pp)
1829 self.report_warning(
1830 '%s: writing DASH m4a. '
1831 'Only some players support this container. %s'
1832 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1834 assert fixup_policy in ('ignore', 'never')
1836 if (info_dict.get('protocol') == 'm3u8_native' or
1837 info_dict.get('protocol') == 'm3u8' and
1838 self.params.get('hls_prefer_native')):
1839 if fixup_policy == 'warn':
1840 self.report_warning('%s: malformated aac bitstream.' % (
1842 elif fixup_policy == 'detect_or_warn':
1843 fixup_pp = FFmpegFixupM3u8PP(self)
1844 if fixup_pp.available:
1845 info_dict.setdefault('__postprocessors', [])
1846 info_dict['__postprocessors'].append(fixup_pp)
1848 self.report_warning(
1849 '%s: malformated aac bitstream. %s'
1850 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1852 assert fixup_policy in ('ignore', 'never')
1855 self.post_process(filename, info_dict)
1856 except (PostProcessingError) as err:
1857 self.report_error('postprocessing: %s' % str(err))
1859 self.record_download_archive(info_dict)
1861 def download(self, url_list):
1862 """Download a given list of URLs."""
1863 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1864 if (len(url_list) > 1 and
1865 '%' not in outtmpl and
1866 self.params.get('max_downloads') != 1):
1867 raise SameFileError(outtmpl)
1869 for url in url_list:
1871 # It also downloads the videos
1872 res = self.extract_info(
1873 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1874 except UnavailableVideoError:
1875 self.report_error('unable to download video')
1876 except MaxDownloadsReached:
1877 self.to_screen('[info] Maximum number of downloaded files reached.')
1880 if self.params.get('dump_single_json', False):
1881 self.to_stdout(json.dumps(res))
1883 return self._download_retcode
1885 def download_with_info_file(self, info_filename):
1886 with contextlib.closing(fileinput.FileInput(
1887 [info_filename], mode='r',
1888 openhook=fileinput.hook_encoded('utf-8'))) as f:
1889 # FileInput doesn't have a read method, we can't call json.load
1890 info = self.filter_requested_info(json.loads('\n'.join(f)))
1892 self.process_ie_result(info, download=True)
1893 except DownloadError:
1894 webpage_url = info.get('webpage_url')
1895 if webpage_url is not None:
1896 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1897 return self.download([webpage_url])
1900 return self._download_retcode
1903 def filter_requested_info(info_dict):
1905 (k, v) for k, v in info_dict.items()
1906 if k not in ['requested_formats', 'requested_subtitles'])
1908 def post_process(self, filename, ie_info):
1909 """Run all the postprocessors on the given file."""
1910 info = dict(ie_info)
1911 info['filepath'] = filename
1913 if ie_info.get('__postprocessors') is not None:
1914 pps_chain.extend(ie_info['__postprocessors'])
1915 pps_chain.extend(self._pps)
1916 for pp in pps_chain:
1917 files_to_delete = []
1919 files_to_delete, info = pp.run(info)
1920 except PostProcessingError as e:
1921 self.report_error(e.msg)
1922 if files_to_delete and not self.params.get('keepvideo', False):
1923 for old_filename in files_to_delete:
1924 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1926 os.remove(encodeFilename(old_filename))
1927 except (IOError, OSError):
1928 self.report_warning('Unable to remove downloaded original file')
1930 def _make_archive_id(self, info_dict):
1931 # Future-proof against any change in case
1932 # and backwards compatibility with prior versions
1933 extractor = info_dict.get('extractor_key')
1934 if extractor is None:
1935 if 'id' in info_dict:
1936 extractor = info_dict.get('ie_key') # key in a playlist
1937 if extractor is None:
1938 return None # Incomplete video information
1939 return extractor.lower() + ' ' + info_dict['id']
1941 def in_download_archive(self, info_dict):
1942 fn = self.params.get('download_archive')
1946 vid_id = self._make_archive_id(info_dict)
1948 return False # Incomplete video information
1951 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1952 for line in archive_file:
1953 if line.strip() == vid_id:
1955 except IOError as ioe:
1956 if ioe.errno != errno.ENOENT:
1960 def record_download_archive(self, info_dict):
1961 fn = self.params.get('download_archive')
1964 vid_id = self._make_archive_id(info_dict)
1966 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1967 archive_file.write(vid_id + '\n')
1970 def format_resolution(format, default='unknown'):
1971 if format.get('vcodec') == 'none':
1973 if format.get('resolution') is not None:
1974 return format['resolution']
1975 if format.get('height') is not None:
1976 if format.get('width') is not None:
1977 res = '%sx%s' % (format['width'], format['height'])
1979 res = '%sp' % format['height']
1980 elif format.get('width') is not None:
1981 res = '%dx?' % format['width']
1986 def _format_note(self, fdict):
1988 if fdict.get('ext') in ['f4f', 'f4m']:
1989 res += '(unsupported) '
1990 if fdict.get('language'):
1993 res += '[%s] ' % fdict['language']
1994 if fdict.get('format_note') is not None:
1995 res += fdict['format_note'] + ' '
1996 if fdict.get('tbr') is not None:
1997 res += '%4dk ' % fdict['tbr']
1998 if fdict.get('container') is not None:
2001 res += '%s container' % fdict['container']
2002 if (fdict.get('vcodec') is not None and
2003 fdict.get('vcodec') != 'none'):
2006 res += fdict['vcodec']
2007 if fdict.get('vbr') is not None:
2009 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2011 if fdict.get('vbr') is not None:
2012 res += '%4dk' % fdict['vbr']
2013 if fdict.get('fps') is not None:
2016 res += '%sfps' % fdict['fps']
2017 if fdict.get('acodec') is not None:
2020 if fdict['acodec'] == 'none':
2023 res += '%-5s' % fdict['acodec']
2024 elif fdict.get('abr') is not None:
2028 if fdict.get('abr') is not None:
2029 res += '@%3dk' % fdict['abr']
2030 if fdict.get('asr') is not None:
2031 res += ' (%5dHz)' % fdict['asr']
2032 if fdict.get('filesize') is not None:
2035 res += format_bytes(fdict['filesize'])
2036 elif fdict.get('filesize_approx') is not None:
2039 res += '~' + format_bytes(fdict['filesize_approx'])
2042 def list_formats(self, info_dict):
2043 formats = info_dict.get('formats', [info_dict])
2045 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
2047 if f.get('preference') is None or f['preference'] >= -1000]
2048 if len(formats) > 1:
2049 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
2051 header_line = ['format code', 'extension', 'resolution', 'note']
2053 '[info] Available formats for %s:\n%s' %
2054 (info_dict['id'], render_table(header_line, table)))
2056 def list_thumbnails(self, info_dict):
2057 thumbnails = info_dict.get('thumbnails')
2059 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
2063 '[info] Thumbnails for %s:' % info_dict['id'])
2064 self.to_screen(render_table(
2065 ['ID', 'width', 'height', 'URL'],
2066 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
2068 def list_subtitles(self, video_id, subtitles, name='subtitles'):
2070 self.to_screen('%s has no %s' % (video_id, name))
2073 'Available %s for %s:' % (name, video_id))
2074 self.to_screen(render_table(
2075 ['Language', 'formats'],
2076 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
2077 for lang, formats in subtitles.items()]))
2079 def urlopen(self, req):
2080 """ Start an HTTP download """
2081 if isinstance(req, compat_basestring):
2082 req = sanitized_Request(req)
2083 return self._opener.open(req, timeout=self._socket_timeout)
2085 def print_debug_header(self):
2086 if not self.params.get('verbose'):
2089 if type('') is not compat_str:
2090 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
2091 self.report_warning(
2092 'Your Python is broken! Update to a newer and supported version')
2094 stdout_encoding = getattr(
2095 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2097 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2098 locale.getpreferredencoding(),
2099 sys.getfilesystemencoding(),
2101 self.get_encoding()))
2102 write_string(encoding_str, encoding=None)
2104 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
2106 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2108 sp = subprocess.Popen(
2109 ['git', 'rev-parse', '--short', 'HEAD'],
2110 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2111 cwd=os.path.dirname(os.path.abspath(__file__)))
2112 out, err = sp.communicate()
2113 out = out.decode().strip()
2114 if re.match('[0-9a-f]+', out):
2115 self._write_string('[debug] Git HEAD: ' + out + '\n')
2121 self._write_string('[debug] Python version %s - %s\n' % (
2122 platform.python_version(), platform_name()))
2124 exe_versions = FFmpegPostProcessor.get_versions(self)
2125 exe_versions['rtmpdump'] = rtmpdump_version()
2126 exe_str = ', '.join(
2128 for exe, v in sorted(exe_versions.items())
2133 self._write_string('[debug] exe versions: %s\n' % exe_str)
2136 for handler in self._opener.handlers:
2137 if hasattr(handler, 'proxies'):
2138 proxy_map.update(handler.proxies)
2139 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2141 if self.params.get('call_home', False):
2142 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2143 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2144 latest_version = self.urlopen(
2145 'https://yt-dl.org/latest/version').read().decode('utf-8')
2146 if version_tuple(latest_version) > version_tuple(__version__):
2147 self.report_warning(
2148 'You are using an outdated version (newest version: %s)! '
2149 'See https://yt-dl.org/update if you need help updating.' %
2152 def _setup_opener(self):
2153 timeout_val = self.params.get('socket_timeout')
2154 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2156 opts_cookiefile = self.params.get('cookiefile')
2157 opts_proxy = self.params.get('proxy')
2159 if opts_cookiefile is None:
2160 self.cookiejar = compat_cookiejar.CookieJar()
2162 opts_cookiefile = compat_expanduser(opts_cookiefile)
2163 self.cookiejar = compat_cookiejar.MozillaCookieJar(
2165 if os.access(opts_cookiefile, os.R_OK):
2166 self.cookiejar.load()
2168 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2169 if opts_proxy is not None:
2170 if opts_proxy == '':
2173 proxies = {'http': opts_proxy, 'https': opts_proxy}
2175 proxies = compat_urllib_request.getproxies()
2176 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2177 if 'http' in proxies and 'https' not in proxies:
2178 proxies['https'] = proxies['http']
2179 proxy_handler = PerRequestProxyHandler(proxies)
2181 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2182 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2183 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2184 data_handler = compat_urllib_request_DataHandler()
2186 # When passing our own FileHandler instance, build_opener won't add the
2187 # default FileHandler and allows us to disable the file protocol, which
2188 # can be used for malicious purposes (see
2189 # https://github.com/rg3/youtube-dl/issues/8227)
2190 file_handler = compat_urllib_request.FileHandler()
2192 def file_open(*args, **kwargs):
2193 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2194 file_handler.file_open = file_open
2196 opener = compat_urllib_request.build_opener(
2197 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2199 # Delete the default user-agent header, which would otherwise apply in
2200 # cases where our custom HTTP handler doesn't come into play
2201 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2202 opener.addheaders = []
2203 self._opener = opener
2205 def encode(self, s):
2206 if isinstance(s, bytes):
2207 return s # Already encoded
2210 return s.encode(self.get_encoding())
2211 except UnicodeEncodeError as err:
2212 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2215 def get_encoding(self):
2216 encoding = self.params.get('encoding')
2217 if encoding is None:
2218 encoding = preferredencoding()
2221 def _write_thumbnails(self, info_dict, filename):
2222 if self.params.get('writethumbnail', False):
2223 thumbnails = info_dict.get('thumbnails')
2225 thumbnails = [thumbnails[-1]]
2226 elif self.params.get('write_all_thumbnails', False):
2227 thumbnails = info_dict.get('thumbnails')
2232 # No thumbnails present, so return immediately
2235 for t in thumbnails:
2236 thumb_ext = determine_ext(t['url'], 'jpg')
2237 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2238 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2239 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2241 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2242 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2243 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2245 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2246 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2248 uf = self.urlopen(t['url'])
2249 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2250 shutil.copyfileobj(uf, thumbf)
2251 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2252 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2253 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2254 self.report_warning('Unable to download thumbnail "%s": %s' %
2255 (t['url'], error_to_compat_str(err)))