2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import, unicode_literals
32 compat_get_terminal_size,
37 compat_tokenize_tokenize,
39 compat_urllib_request,
40 compat_urllib_request_DataHandler,
63 PerRequestProxyHandler,
68 register_socks_protocols,
78 UnavailableVideoError,
83 YoutubeDLCookieProcessor,
86 from .cache import Cache
87 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
88 from .downloader import get_suitable_downloader
89 from .downloader.rtmp import rtmpdump_version
90 from .postprocessor import (
93 FFmpegFixupStretchedPP,
98 from .version import __version__
100 if compat_os_name == 'nt':
104 class YoutubeDL(object):
107 YoutubeDL objects are the ones responsible of downloading the
108 actual video file and writing it to disk if the user has requested
109 it, among some other tasks. In most cases there should be one per
110 program. As, given a video URL, the downloader doesn't know how to
111 extract all the needed information, task that InfoExtractors do, it
112 has to pass the URL to one of them.
114 For this, YoutubeDL objects have a method that allows
115 InfoExtractors to be registered in a given order. When it is passed
116 a URL, the YoutubeDL object handles it to the first InfoExtractor it
117 finds that reports being able to handle it. The InfoExtractor extracts
118 all the information about the video or videos the URL refers to, and
119 YoutubeDL process the extracted information, possibly using a File
120 Downloader to download the video.
122 YoutubeDL objects accept a lot of parameters. In order not to saturate
123 the object constructor with arguments, it receives a dictionary of
124 options instead. These options are available through the params
125 attribute for the InfoExtractors to use. The YoutubeDL also
126 registers itself as the downloader in charge for the InfoExtractors
127 that are added to it, so this is a "mutual registration".
131 username: Username for authentication purposes.
132 password: Password for authentication purposes.
133 videopassword: Password for accessing a video.
134 usenetrc: Use netrc for authentication instead.
135 verbose: Print additional info to stdout.
136 quiet: Do not print messages to stdout.
137 no_warnings: Do not print out anything for warnings.
138 forceurl: Force printing final URL.
139 forcetitle: Force printing title.
140 forceid: Force printing ID.
141 forcethumbnail: Force printing thumbnail URL.
142 forcedescription: Force printing description.
143 forcefilename: Force printing final filename.
144 forceduration: Force printing duration.
145 forcejson: Force printing info_dict as JSON.
146 dump_single_json: Force printing the info_dict of the whole playlist
147 (or video) as a single JSON line.
148 simulate: Do not download the video files.
149 format: Video format code. See options.py for more information.
150 outtmpl: Template for output names.
151 restrictfilenames: Do not allow "&" and spaces in file names
152 ignoreerrors: Do not stop on download errors.
153 force_generic_extractor: Force downloader to use the generic extractor
154 nooverwrites: Prevent overwriting files.
155 playliststart: Playlist item to start at.
156 playlistend: Playlist item to end at.
157 playlist_items: Specific indices of playlist to download.
158 playlistreverse: Download playlist items in reverse order.
159 matchtitle: Download only matching titles.
160 rejecttitle: Reject downloads for matching titles.
161 logger: Log messages to a logging.Logger instance.
162 logtostderr: Log messages to stderr instead of stdout.
163 writedescription: Write the video description to a .description file
164 writeinfojson: Write the video description to a .info.json file
165 writeannotations: Write the video annotations to a .annotations.xml file
166 writethumbnail: Write the thumbnail image to a file
167 write_all_thumbnails: Write all thumbnail formats to files
168 writesubtitles: Write the video subtitles to a file
169 writeautomaticsub: Write the automatically generated subtitles to a file
170 allsubtitles: Downloads all the subtitles of the video
171 (requires writesubtitles or writeautomaticsub)
172 listsubtitles: Lists all available subtitles for the video
173 subtitlesformat: The format code for subtitles
174 subtitleslangs: List of languages of the subtitles to download
175 keepvideo: Keep the video file after post-processing
176 daterange: A DateRange object, download only if the upload_date is in the range.
177 skip_download: Skip the actual download of the video file
178 cachedir: Location of the cache files in the filesystem.
179 False to disable filesystem cache.
180 noplaylist: Download single video instead of a playlist if in doubt.
181 age_limit: An integer representing the user's age in years.
182 Unsuitable videos for the given age are skipped.
183 min_views: An integer representing the minimum view count the video
184 must have in order to not be skipped.
185 Videos without view count information are always
186 downloaded. None for no limit.
187 max_views: An integer representing the maximum view count.
188 Videos that are more popular than that are not
190 Videos without view count information are always
191 downloaded. None for no limit.
192 download_archive: File name of a file where all downloads are recorded.
193 Videos already present in the file are not downloaded
195 cookiefile: File name where cookies should be read from and dumped to.
196 nocheckcertificate:Do not verify SSL certificates
197 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
198 At the moment, this is only supported by YouTube.
199 proxy: URL of the proxy server to use
200 geo_verification_proxy: URL of the proxy to use for IP address verification
201 on geo-restricted sites. (Experimental)
202 socket_timeout: Time to wait for unresponsive hosts, in seconds
203 bidi_workaround: Work around buggy terminals without bidirectional text
204 support, using fridibi
205 debug_printtraffic:Print out sent and received HTTP traffic
206 include_ads: Download ads as well
207 default_search: Prepend this string if an input url is not valid.
208 'auto' for elaborate guessing
209 encoding: Use this encoding instead of the system-specified.
210 extract_flat: Do not resolve URLs, return the immediate result.
211 Pass in 'in_playlist' to only show this behavior for
213 postprocessors: A list of dictionaries, each with an entry
214 * key: The name of the postprocessor. See
215 youtube_dl/postprocessor/__init__.py for a list.
216 as well as any further keyword arguments for the
218 progress_hooks: A list of functions that get called on download
219 progress, with a dictionary with the entries
220 * status: One of "downloading", "error", or "finished".
221 Check this first and ignore unknown values.
223 If status is one of "downloading", or "finished", the
224 following properties may also be present:
225 * filename: The final filename (always present)
226 * tmpfilename: The filename we're currently writing to
227 * downloaded_bytes: Bytes on disk
228 * total_bytes: Size of the whole file, None if unknown
229 * total_bytes_estimate: Guess of the eventual file size,
231 * elapsed: The number of seconds since download started.
232 * eta: The estimated time in seconds, None if unknown
233 * speed: The download speed in bytes/second, None if
235 * fragment_index: The counter of the currently
236 downloaded video fragment.
237 * fragment_count: The number of fragments (= individual
238 files that will be merged)
240 Progress hooks are guaranteed to be called at least once
241 (with status "finished") if the download is successful.
242 merge_output_format: Extension to use when merging formats.
243 fixup: Automatically correct known faults of the file.
245 - "never": do nothing
246 - "warn": only emit a warning
247 - "detect_or_warn": check whether we can do anything
248 about it, warn otherwise (default)
249 source_address: (Experimental) Client-side IP address to bind to.
250 call_home: Boolean, true iff we are allowed to contact the
251 youtube-dl servers for debugging.
252 sleep_interval: Minimum number of seconds to sleep before each download.
253 Sleep will be for a random interval if --max-sleep-interval is also passed.
254 max_sleep_interval:Max number of seconds to sleep before each download.
255 Sleep will be for a random interval if passed along with --min-sleep-interval
256 or --sleep-interval, otherwise ignored.
257 listformats: Print an overview of available video formats and exit.
258 list_thumbnails: Print a table of all thumbnails and exit.
259 match_filter: A function that gets called with the info_dict of
261 If it returns a message, the video is ignored.
262 If it returns None, the video is downloaded.
263 match_filter_func in utils.py is one example for this.
264 no_color: Do not emit color codes in output.
266 The following options determine which downloader is picked:
267 external_downloader: Executable of the external downloader to call.
268 None or unset for standard (built-in) downloader.
269 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
270 if True, otherwise use ffmpeg/avconv if False, otherwise
271 use downloader suggested by extractor if None.
273 The following parameters are not used by YoutubeDL itself, they are used by
274 the downloader (see youtube_dl/downloader/common.py):
275 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
276 noresizebuffer, retries, continuedl, noprogress, consoletitle,
277 xattr_set_filesize, external_downloader_args, hls_use_mpegts.
279 The following options are used by the post processors:
280 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
281 otherwise prefer avconv.
282 postprocessor_args: A list of additional command-line arguments for the
289 _download_retcode = None
290 _num_downloads = None
293 def __init__(self, params=None, auto_init=True):
294 """Create a FileDownloader object with the given options."""
298 self._ies_instances = {}
300 self._progress_hooks = []
301 self._download_retcode = 0
302 self._num_downloads = 0
303 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
304 self._err_file = sys.stderr
307 'nocheckcertificate': False,
309 self.params.update(params)
310 self.cache = Cache(self)
312 if self.params.get('cn_verification_proxy') is not None:
313 self.report_warning('--cn-verification-proxy is deprecated. Use --geo-verification-proxy instead.')
314 if self.params.get('geo_verification_proxy') is None:
315 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
317 if params.get('bidi_workaround', False):
320 master, slave = pty.openpty()
321 width = compat_get_terminal_size().columns
325 width_args = ['-w', str(width)]
327 stdin=subprocess.PIPE,
329 stderr=self._err_file)
331 self._output_process = subprocess.Popen(
332 ['bidiv'] + width_args, **sp_kwargs
335 self._output_process = subprocess.Popen(
336 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
337 self._output_channel = os.fdopen(master, 'rb')
338 except OSError as ose:
339 if ose.errno == errno.ENOENT:
340 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 $PATH.')
344 if (sys.version_info >= (3,) and sys.platform != 'win32' and
345 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
346 not params.get('restrictfilenames', False)):
347 # On Python 3, the Unicode filesystem API will throw errors (#1474)
349 'Assuming --restrict-filenames since file system encoding '
350 'cannot encode all characters. '
351 'Set the LC_ALL environment variable to fix this.')
352 self.params['restrictfilenames'] = True
354 if isinstance(params.get('outtmpl'), bytes):
356 'Parameter outtmpl is bytes, but should be a unicode string. '
357 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
362 self.print_debug_header()
363 self.add_default_info_extractors()
365 for pp_def_raw in self.params.get('postprocessors', []):
366 pp_class = get_postprocessor(pp_def_raw['key'])
367 pp_def = dict(pp_def_raw)
369 pp = pp_class(self, **compat_kwargs(pp_def))
370 self.add_post_processor(pp)
372 for ph in self.params.get('progress_hooks', []):
373 self.add_progress_hook(ph)
375 register_socks_protocols()
377 def warn_if_short_id(self, argv):
378 # short YouTube ID starting with dash?
380 i for i, a in enumerate(argv)
381 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
385 [a for i, a in enumerate(argv) if i not in idxs] +
386 ['--'] + [argv[i] for i in idxs]
389 'Long argument string detected. '
390 'Use -- to separate parameters and URLs, like this:\n%s\n' %
391 args_to_str(correct_argv))
393 def add_info_extractor(self, ie):
394 """Add an InfoExtractor object to the end of the list."""
396 if not isinstance(ie, type):
397 self._ies_instances[ie.ie_key()] = ie
398 ie.set_downloader(self)
400 def get_info_extractor(self, ie_key):
402 Get an instance of an IE with name ie_key, it will try to get one from
403 the _ies list, if there's no instance it will create a new one and add
404 it to the extractor list.
406 ie = self._ies_instances.get(ie_key)
408 ie = get_info_extractor(ie_key)()
409 self.add_info_extractor(ie)
412 def add_default_info_extractors(self):
414 Add the InfoExtractors returned by gen_extractors to the end of the list
416 for ie in gen_extractor_classes():
417 self.add_info_extractor(ie)
419 def add_post_processor(self, pp):
420 """Add a PostProcessor object to the end of the chain."""
422 pp.set_downloader(self)
424 def add_progress_hook(self, ph):
425 """Add the progress hook (currently only for the file downloader)"""
426 self._progress_hooks.append(ph)
428 def _bidi_workaround(self, message):
429 if not hasattr(self, '_output_channel'):
432 assert hasattr(self, '_output_process')
433 assert isinstance(message, compat_str)
434 line_count = message.count('\n') + 1
435 self._output_process.stdin.write((message + '\n').encode('utf-8'))
436 self._output_process.stdin.flush()
437 res = ''.join(self._output_channel.readline().decode('utf-8')
438 for _ in range(line_count))
439 return res[:-len('\n')]
441 def to_screen(self, message, skip_eol=False):
442 """Print message to stdout if not in quiet mode."""
443 return self.to_stdout(message, skip_eol, check_quiet=True)
445 def _write_string(self, s, out=None):
446 write_string(s, out=out, encoding=self.params.get('encoding'))
448 def to_stdout(self, message, skip_eol=False, check_quiet=False):
449 """Print message to stdout if not in quiet mode."""
450 if self.params.get('logger'):
451 self.params['logger'].debug(message)
452 elif not check_quiet or not self.params.get('quiet', False):
453 message = self._bidi_workaround(message)
454 terminator = ['\n', ''][skip_eol]
455 output = message + terminator
457 self._write_string(output, self._screen_file)
459 def to_stderr(self, message):
460 """Print message to stderr."""
461 assert isinstance(message, compat_str)
462 if self.params.get('logger'):
463 self.params['logger'].error(message)
465 message = self._bidi_workaround(message)
466 output = message + '\n'
467 self._write_string(output, self._err_file)
469 def to_console_title(self, message):
470 if not self.params.get('consoletitle', False):
472 if compat_os_name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
473 # c_wchar_p() might not be necessary if `message` is
474 # already of type unicode()
475 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
476 elif 'TERM' in os.environ:
477 self._write_string('\033]0;%s\007' % message, self._screen_file)
479 def save_console_title(self):
480 if not self.params.get('consoletitle', False):
482 if 'TERM' in os.environ:
483 # Save the title on stack
484 self._write_string('\033[22;0t', self._screen_file)
486 def restore_console_title(self):
487 if not self.params.get('consoletitle', False):
489 if 'TERM' in os.environ:
490 # Restore the title from stack
491 self._write_string('\033[23;0t', self._screen_file)
494 self.save_console_title()
497 def __exit__(self, *args):
498 self.restore_console_title()
500 if self.params.get('cookiefile') is not None:
501 self.cookiejar.save()
503 def trouble(self, message=None, tb=None):
504 """Determine action to take when a download problem appears.
506 Depending on if the downloader has been configured to ignore
507 download errors or not, this method may throw an exception or
508 not when errors are found, after printing the message.
510 tb, if given, is additional traceback information.
512 if message is not None:
513 self.to_stderr(message)
514 if self.params.get('verbose'):
516 if sys.exc_info()[0]: # if .trouble has been called from an except block
518 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
519 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
520 tb += encode_compat_str(traceback.format_exc())
522 tb_data = traceback.format_list(traceback.extract_stack())
523 tb = ''.join(tb_data)
525 if not self.params.get('ignoreerrors', False):
526 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
527 exc_info = sys.exc_info()[1].exc_info
529 exc_info = sys.exc_info()
530 raise DownloadError(message, exc_info)
531 self._download_retcode = 1
533 def report_warning(self, message):
535 Print the message to stderr, it will be prefixed with 'WARNING:'
536 If stderr is a tty file the 'WARNING:' will be colored
538 if self.params.get('logger') is not None:
539 self.params['logger'].warning(message)
541 if self.params.get('no_warnings'):
543 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
544 _msg_header = '\033[0;33mWARNING:\033[0m'
546 _msg_header = 'WARNING:'
547 warning_message = '%s %s' % (_msg_header, message)
548 self.to_stderr(warning_message)
550 def report_error(self, message, tb=None):
552 Do the same as trouble, but prefixes the message with 'ERROR:', colored
553 in red if stderr is a tty file.
555 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
556 _msg_header = '\033[0;31mERROR:\033[0m'
558 _msg_header = 'ERROR:'
559 error_message = '%s %s' % (_msg_header, message)
560 self.trouble(error_message, tb)
562 def report_file_already_downloaded(self, file_name):
563 """Report file has already been fully downloaded."""
565 self.to_screen('[download] %s has already been downloaded' % file_name)
566 except UnicodeEncodeError:
567 self.to_screen('[download] The file has already been downloaded')
569 def prepare_filename(self, info_dict):
570 """Generate the output filename."""
572 template_dict = dict(info_dict)
574 template_dict['epoch'] = int(time.time())
575 autonumber_size = self.params.get('autonumber_size')
576 if autonumber_size is None:
578 autonumber_templ = '%0' + str(autonumber_size) + 'd'
579 template_dict['autonumber'] = autonumber_templ % self._num_downloads
580 if template_dict.get('playlist_index') is not None:
581 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
582 if template_dict.get('resolution') is None:
583 if template_dict.get('width') and template_dict.get('height'):
584 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
585 elif template_dict.get('height'):
586 template_dict['resolution'] = '%sp' % template_dict['height']
587 elif template_dict.get('width'):
588 template_dict['resolution'] = '%dx?' % template_dict['width']
590 sanitize = lambda k, v: sanitize_filename(
592 restricted=self.params.get('restrictfilenames'),
594 template_dict = dict((k, sanitize(k, v))
595 for k, v in template_dict.items()
596 if v is not None and not isinstance(v, (list, tuple, dict)))
597 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
599 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
600 tmpl = compat_expanduser(outtmpl)
601 filename = tmpl % template_dict
602 # Temporary fix for #4787
603 # 'Treat' all problem characters by passing filename through preferredencoding
604 # to workaround encoding issues with subprocess on python2 @ Windows
605 if sys.version_info < (3, 0) and sys.platform == 'win32':
606 filename = encodeFilename(filename, True).decode(preferredencoding())
607 return sanitize_path(filename)
608 except ValueError as err:
609 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
612 def _match_entry(self, info_dict, incomplete):
613 """ Returns None iff the file should be downloaded """
615 video_title = info_dict.get('title', info_dict.get('id', 'video'))
616 if 'title' in info_dict:
617 # This can happen when we're just evaluating the playlist
618 title = info_dict['title']
619 matchtitle = self.params.get('matchtitle', False)
621 if not re.search(matchtitle, title, re.IGNORECASE):
622 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
623 rejecttitle = self.params.get('rejecttitle', False)
625 if re.search(rejecttitle, title, re.IGNORECASE):
626 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
627 date = info_dict.get('upload_date')
629 dateRange = self.params.get('daterange', DateRange())
630 if date not in dateRange:
631 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
632 view_count = info_dict.get('view_count')
633 if view_count is not None:
634 min_views = self.params.get('min_views')
635 if min_views is not None and view_count < min_views:
636 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
637 max_views = self.params.get('max_views')
638 if max_views is not None and view_count > max_views:
639 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
640 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
641 return 'Skipping "%s" because it is age restricted' % video_title
642 if self.in_download_archive(info_dict):
643 return '%s has already been recorded in archive' % video_title
646 match_filter = self.params.get('match_filter')
647 if match_filter is not None:
648 ret = match_filter(info_dict)
655 def add_extra_info(info_dict, extra_info):
656 '''Set the keys from extra_info in info dict if they are missing'''
657 for key, value in extra_info.items():
658 info_dict.setdefault(key, value)
660 def extract_info(self, url, download=True, ie_key=None, extra_info={},
661 process=True, force_generic_extractor=False):
663 Returns a list with a dictionary for each video we find.
664 If 'download', also downloads the videos.
665 extra_info is a dict containing the extra values to add to each result
668 if not ie_key and force_generic_extractor:
672 ies = [self.get_info_extractor(ie_key)]
677 if not ie.suitable(url):
680 ie = self.get_info_extractor(ie.ie_key())
682 self.report_warning('The program functionality for this site has been marked as broken, '
683 'and will probably not work.')
686 ie_result = ie.extract(url)
687 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
689 if isinstance(ie_result, list):
690 # Backwards compatibility: old IE result format
692 '_type': 'compat_list',
693 'entries': ie_result,
695 self.add_default_extra_info(ie_result, ie, url)
697 return self.process_ie_result(ie_result, download, extra_info)
700 except ExtractorError as e: # An error we somewhat expected
701 self.report_error(compat_str(e), e.format_traceback())
703 except MaxDownloadsReached:
705 except Exception as e:
706 if self.params.get('ignoreerrors', False):
707 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
712 self.report_error('no suitable InfoExtractor for URL %s' % url)
714 def add_default_extra_info(self, ie_result, ie, url):
715 self.add_extra_info(ie_result, {
716 'extractor': ie.IE_NAME,
718 'webpage_url_basename': url_basename(url),
719 'extractor_key': ie.ie_key(),
722 def process_ie_result(self, ie_result, download=True, extra_info={}):
724 Take the result of the ie(may be modified) and resolve all unresolved
725 references (URLs, playlist items).
727 It will also download the videos if 'download'.
728 Returns the resolved ie_result.
730 result_type = ie_result.get('_type', 'video')
732 if result_type in ('url', 'url_transparent'):
733 ie_result['url'] = sanitize_url(ie_result['url'])
734 extract_flat = self.params.get('extract_flat', False)
735 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
736 extract_flat is True):
737 if self.params.get('forcejson', False):
738 self.to_stdout(json.dumps(ie_result))
741 if result_type == 'video':
742 self.add_extra_info(ie_result, extra_info)
743 return self.process_video_result(ie_result, download=download)
744 elif result_type == 'url':
745 # We have to add extra_info to the results because it may be
746 # contained in a playlist
747 return self.extract_info(ie_result['url'],
749 ie_key=ie_result.get('ie_key'),
750 extra_info=extra_info)
751 elif result_type == 'url_transparent':
752 # Use the information from the embedding page
753 info = self.extract_info(
754 ie_result['url'], ie_key=ie_result.get('ie_key'),
755 extra_info=extra_info, download=False, process=False)
757 force_properties = dict(
758 (k, v) for k, v in ie_result.items() if v is not None)
759 for f in ('_type', 'url', 'ie_key'):
760 if f in force_properties:
761 del force_properties[f]
762 new_result = info.copy()
763 new_result.update(force_properties)
765 assert new_result.get('_type') != 'url_transparent'
767 return self.process_ie_result(
768 new_result, download=download, extra_info=extra_info)
769 elif result_type == 'playlist' or result_type == 'multi_video':
770 # We process each entry in the playlist
771 playlist = ie_result.get('title') or ie_result.get('id')
772 self.to_screen('[download] Downloading playlist: %s' % playlist)
774 playlist_results = []
776 playliststart = self.params.get('playliststart', 1) - 1
777 playlistend = self.params.get('playlistend')
778 # For backwards compatibility, interpret -1 as whole list
779 if playlistend == -1:
782 playlistitems_str = self.params.get('playlist_items')
784 if playlistitems_str is not None:
785 def iter_playlistitems(format):
786 for string_segment in format.split(','):
787 if '-' in string_segment:
788 start, end = string_segment.split('-')
789 for item in range(int(start), int(end) + 1):
792 yield int(string_segment)
793 playlistitems = iter_playlistitems(playlistitems_str)
795 ie_entries = ie_result['entries']
796 if isinstance(ie_entries, list):
797 n_all_entries = len(ie_entries)
800 ie_entries[i - 1] for i in playlistitems
801 if -n_all_entries <= i - 1 < n_all_entries]
803 entries = ie_entries[playliststart:playlistend]
804 n_entries = len(entries)
806 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
807 (ie_result['extractor'], playlist, n_all_entries, n_entries))
808 elif isinstance(ie_entries, PagedList):
811 for item in playlistitems:
812 entries.extend(ie_entries.getslice(
816 entries = ie_entries.getslice(
817 playliststart, playlistend)
818 n_entries = len(entries)
820 '[%s] playlist %s: Downloading %d videos' %
821 (ie_result['extractor'], playlist, n_entries))
824 entry_list = list(ie_entries)
825 entries = [entry_list[i - 1] for i in playlistitems]
827 entries = list(itertools.islice(
828 ie_entries, playliststart, playlistend))
829 n_entries = len(entries)
831 '[%s] playlist %s: Downloading %d videos' %
832 (ie_result['extractor'], playlist, n_entries))
834 if self.params.get('playlistreverse', False):
835 entries = entries[::-1]
837 for i, entry in enumerate(entries, 1):
838 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
840 'n_entries': n_entries,
841 'playlist': playlist,
842 'playlist_id': ie_result.get('id'),
843 'playlist_title': ie_result.get('title'),
844 'playlist_index': i + playliststart,
845 'extractor': ie_result['extractor'],
846 'webpage_url': ie_result['webpage_url'],
847 'webpage_url_basename': url_basename(ie_result['webpage_url']),
848 'extractor_key': ie_result['extractor_key'],
851 reason = self._match_entry(entry, incomplete=True)
852 if reason is not None:
853 self.to_screen('[download] ' + reason)
856 entry_result = self.process_ie_result(entry,
859 playlist_results.append(entry_result)
860 ie_result['entries'] = playlist_results
861 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
863 elif result_type == 'compat_list':
865 'Extractor %s returned a compat_list result. '
866 'It needs to be updated.' % ie_result.get('extractor'))
872 'extractor': ie_result['extractor'],
873 'webpage_url': ie_result['webpage_url'],
874 'webpage_url_basename': url_basename(ie_result['webpage_url']),
875 'extractor_key': ie_result['extractor_key'],
879 ie_result['entries'] = [
880 self.process_ie_result(_fixup(r), download, extra_info)
881 for r in ie_result['entries']
885 raise Exception('Invalid result type: %s' % result_type)
887 def _build_format_filter(self, filter_spec):
888 " Returns a function to filter the formats according to the filter_spec "
898 operator_rex = re.compile(r'''(?x)\s*
899 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
900 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
901 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
903 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
904 m = operator_rex.search(filter_spec)
907 comparison_value = int(m.group('value'))
909 comparison_value = parse_filesize(m.group('value'))
910 if comparison_value is None:
911 comparison_value = parse_filesize(m.group('value') + 'B')
912 if comparison_value is None:
914 'Invalid value %r in format specification %r' % (
915 m.group('value'), filter_spec))
916 op = OPERATORS[m.group('op')]
922 '^=': lambda attr, value: attr.startswith(value),
923 '$=': lambda attr, value: attr.endswith(value),
924 '*=': lambda attr, value: value in attr,
926 str_operator_rex = re.compile(r'''(?x)
927 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
928 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
929 \s*(?P<value>[a-zA-Z0-9._-]+)
931 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
932 m = str_operator_rex.search(filter_spec)
934 comparison_value = m.group('value')
935 op = STR_OPERATORS[m.group('op')]
938 raise ValueError('Invalid filter specification %r' % filter_spec)
941 actual_value = f.get(m.group('key'))
942 if actual_value is None:
943 return m.group('none_inclusive')
944 return op(actual_value, comparison_value)
947 def build_format_selector(self, format_spec):
948 def syntax_error(note, start):
950 'Invalid format specification: '
951 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
952 return SyntaxError(message)
954 PICKFIRST = 'PICKFIRST'
958 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
960 def _parse_filter(tokens):
962 for type, string, start, _, _ in tokens:
963 if type == tokenize.OP and string == ']':
964 return ''.join(filter_parts)
966 filter_parts.append(string)
968 def _remove_unused_ops(tokens):
969 # Remove operators that we don't use and join them with the surrounding strings
970 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
971 ALLOWED_OPS = ('/', '+', ',', '(', ')')
972 last_string, last_start, last_end, last_line = None, None, None, None
973 for type, string, start, end, line in tokens:
974 if type == tokenize.OP and string == '[':
976 yield tokenize.NAME, last_string, last_start, last_end, last_line
978 yield type, string, start, end, line
979 # everything inside brackets will be handled by _parse_filter
980 for type, string, start, end, line in tokens:
981 yield type, string, start, end, line
982 if type == tokenize.OP and string == ']':
984 elif type == tokenize.OP and string in ALLOWED_OPS:
986 yield tokenize.NAME, last_string, last_start, last_end, last_line
988 yield type, string, start, end, line
989 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
995 last_string += string
997 yield tokenize.NAME, last_string, last_start, last_end, last_line
999 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1001 current_selector = None
1002 for type, string, start, _, _ in tokens:
1003 # ENCODING is only defined in python 3.x
1004 if type == getattr(tokenize, 'ENCODING', None):
1006 elif type in [tokenize.NAME, tokenize.NUMBER]:
1007 current_selector = FormatSelector(SINGLE, string, [])
1008 elif type == tokenize.OP:
1010 if not inside_group:
1011 # ')' will be handled by the parentheses group
1012 tokens.restore_last_token()
1014 elif inside_merge and string in ['/', ',']:
1015 tokens.restore_last_token()
1017 elif inside_choice and string == ',':
1018 tokens.restore_last_token()
1021 if not current_selector:
1022 raise syntax_error('"," must follow a format selector', start)
1023 selectors.append(current_selector)
1024 current_selector = None
1026 if not current_selector:
1027 raise syntax_error('"/" must follow a format selector', start)
1028 first_choice = current_selector
1029 second_choice = _parse_format_selection(tokens, inside_choice=True)
1030 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1032 if not current_selector:
1033 current_selector = FormatSelector(SINGLE, 'best', [])
1034 format_filter = _parse_filter(tokens)
1035 current_selector.filters.append(format_filter)
1037 if current_selector:
1038 raise syntax_error('Unexpected "("', start)
1039 group = _parse_format_selection(tokens, inside_group=True)
1040 current_selector = FormatSelector(GROUP, group, [])
1042 video_selector = current_selector
1043 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1044 if not video_selector or not audio_selector:
1045 raise syntax_error('"+" must be between two format selectors', start)
1046 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1048 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1049 elif type == tokenize.ENDMARKER:
1051 if current_selector:
1052 selectors.append(current_selector)
1055 def _build_selector_function(selector):
1056 if isinstance(selector, list):
1057 fs = [_build_selector_function(s) for s in selector]
1059 def selector_function(ctx):
1061 for format in f(ctx):
1063 return selector_function
1064 elif selector.type == GROUP:
1065 selector_function = _build_selector_function(selector.selector)
1066 elif selector.type == PICKFIRST:
1067 fs = [_build_selector_function(s) for s in selector.selector]
1069 def selector_function(ctx):
1071 picked_formats = list(f(ctx))
1073 return picked_formats
1075 elif selector.type == SINGLE:
1076 format_spec = selector.selector
1078 def selector_function(ctx):
1079 formats = list(ctx['formats'])
1082 if format_spec == 'all':
1085 elif format_spec in ['best', 'worst', None]:
1086 format_idx = 0 if format_spec == 'worst' else -1
1087 audiovideo_formats = [
1089 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1090 if audiovideo_formats:
1091 yield audiovideo_formats[format_idx]
1092 # for extractors with incomplete formats (audio only (soundcloud)
1093 # or video only (imgur)) we will fallback to best/worst
1094 # {video,audio}-only format
1095 elif ctx['incomplete_formats']:
1096 yield formats[format_idx]
1097 elif format_spec == 'bestaudio':
1100 if f.get('vcodec') == 'none']
1102 yield audio_formats[-1]
1103 elif format_spec == 'worstaudio':
1106 if f.get('vcodec') == 'none']
1108 yield audio_formats[0]
1109 elif format_spec == 'bestvideo':
1112 if f.get('acodec') == 'none']
1114 yield video_formats[-1]
1115 elif format_spec == 'worstvideo':
1118 if f.get('acodec') == 'none']
1120 yield video_formats[0]
1122 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1123 if format_spec in extensions:
1124 filter_f = lambda f: f['ext'] == format_spec
1126 filter_f = lambda f: f['format_id'] == format_spec
1127 matches = list(filter(filter_f, formats))
1130 elif selector.type == MERGE:
1131 def _merge(formats_info):
1132 format_1, format_2 = [f['format_id'] for f in formats_info]
1133 # The first format must contain the video and the
1135 if formats_info[0].get('vcodec') == 'none':
1136 self.report_error('The first format must '
1137 'contain the video, try using '
1138 '"-f %s+%s"' % (format_2, format_1))
1140 # Formats must be opposite (video+audio)
1141 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1143 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1144 % (format_1, format_2))
1147 formats_info[0]['ext']
1148 if self.params.get('merge_output_format') is None
1149 else self.params['merge_output_format'])
1151 'requested_formats': formats_info,
1152 'format': '%s+%s' % (formats_info[0].get('format'),
1153 formats_info[1].get('format')),
1154 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1155 formats_info[1].get('format_id')),
1156 'width': formats_info[0].get('width'),
1157 'height': formats_info[0].get('height'),
1158 'resolution': formats_info[0].get('resolution'),
1159 'fps': formats_info[0].get('fps'),
1160 'vcodec': formats_info[0].get('vcodec'),
1161 'vbr': formats_info[0].get('vbr'),
1162 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1163 'acodec': formats_info[1].get('acodec'),
1164 'abr': formats_info[1].get('abr'),
1167 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1169 def selector_function(ctx):
1170 for pair in itertools.product(
1171 video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
1174 filters = [self._build_format_filter(f) for f in selector.filters]
1176 def final_selector(ctx):
1177 ctx_copy = copy.deepcopy(ctx)
1178 for _filter in filters:
1179 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1180 return selector_function(ctx_copy)
1181 return final_selector
1183 stream = io.BytesIO(format_spec.encode('utf-8'))
1185 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1186 except tokenize.TokenError:
1187 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1189 class TokenIterator(object):
1190 def __init__(self, tokens):
1191 self.tokens = tokens
1198 if self.counter >= len(self.tokens):
1199 raise StopIteration()
1200 value = self.tokens[self.counter]
1206 def restore_last_token(self):
1209 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1210 return _build_selector_function(parsed_selector)
1212 def _calc_headers(self, info_dict):
1213 res = std_headers.copy()
1215 add_headers = info_dict.get('http_headers')
1217 res.update(add_headers)
1219 cookies = self._calc_cookies(info_dict)
1221 res['Cookie'] = cookies
1225 def _calc_cookies(self, info_dict):
1226 pr = sanitized_Request(info_dict['url'])
1227 self.cookiejar.add_cookie_header(pr)
1228 return pr.get_header('Cookie')
1230 def process_video_result(self, info_dict, download=True):
1231 assert info_dict.get('_type', 'video') == 'video'
1233 if 'id' not in info_dict:
1234 raise ExtractorError('Missing "id" field in extractor result')
1235 if 'title' not in info_dict:
1236 raise ExtractorError('Missing "title" field in extractor result')
1238 if not isinstance(info_dict['id'], compat_str):
1239 self.report_warning('"id" field is not a string - forcing string conversion')
1240 info_dict['id'] = compat_str(info_dict['id'])
1242 if 'playlist' not in info_dict:
1243 # It isn't part of a playlist
1244 info_dict['playlist'] = None
1245 info_dict['playlist_index'] = None
1247 thumbnails = info_dict.get('thumbnails')
1248 if thumbnails is None:
1249 thumbnail = info_dict.get('thumbnail')
1251 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1253 thumbnails.sort(key=lambda t: (
1254 t.get('preference'), t.get('width'), t.get('height'),
1255 t.get('id'), t.get('url')))
1256 for i, t in enumerate(thumbnails):
1257 t['url'] = sanitize_url(t['url'])
1258 if t.get('width') and t.get('height'):
1259 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1260 if t.get('id') is None:
1263 if self.params.get('list_thumbnails'):
1264 self.list_thumbnails(info_dict)
1267 thumbnail = info_dict.get('thumbnail')
1269 info_dict['thumbnail'] = sanitize_url(thumbnail)
1271 info_dict['thumbnail'] = thumbnails[-1]['url']
1273 if 'display_id' not in info_dict and 'id' in info_dict:
1274 info_dict['display_id'] = info_dict['id']
1276 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1277 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1278 # see http://bugs.python.org/issue1646728)
1280 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1281 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1282 except (ValueError, OverflowError, OSError):
1285 # Auto generate title fields corresponding to the *_number fields when missing
1286 # in order to always have clean titles. This is very common for TV series.
1287 for field in ('chapter', 'season', 'episode'):
1288 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1289 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1291 subtitles = info_dict.get('subtitles')
1293 for _, subtitle in subtitles.items():
1294 for subtitle_format in subtitle:
1295 if subtitle_format.get('url'):
1296 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1297 if 'ext' not in subtitle_format:
1298 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1300 if self.params.get('listsubtitles', False):
1301 if 'automatic_captions' in info_dict:
1302 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1303 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1305 info_dict['requested_subtitles'] = self.process_subtitles(
1306 info_dict['id'], subtitles,
1307 info_dict.get('automatic_captions'))
1309 # We now pick which formats have to be downloaded
1310 if info_dict.get('formats') is None:
1311 # There's only one format available
1312 formats = [info_dict]
1314 formats = info_dict['formats']
1317 raise ExtractorError('No video formats found!')
1321 # We check that all the formats have the format and format_id fields
1322 for i, format in enumerate(formats):
1323 if 'url' not in format:
1324 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1326 format['url'] = sanitize_url(format['url'])
1328 if format.get('format_id') is None:
1329 format['format_id'] = compat_str(i)
1331 # Sanitize format_id from characters used in format selector expression
1332 format['format_id'] = re.sub('[\s,/+\[\]()]', '_', format['format_id'])
1333 format_id = format['format_id']
1334 if format_id not in formats_dict:
1335 formats_dict[format_id] = []
1336 formats_dict[format_id].append(format)
1338 # Make sure all formats have unique format_id
1339 for format_id, ambiguous_formats in formats_dict.items():
1340 if len(ambiguous_formats) > 1:
1341 for i, format in enumerate(ambiguous_formats):
1342 format['format_id'] = '%s-%d' % (format_id, i)
1344 for i, format in enumerate(formats):
1345 if format.get('format') is None:
1346 format['format'] = '{id} - {res}{note}'.format(
1347 id=format['format_id'],
1348 res=self.format_resolution(format),
1349 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1351 # Automatically determine file extension if missing
1352 if 'ext' not in format:
1353 format['ext'] = determine_ext(format['url']).lower()
1354 # Automatically determine protocol if missing (useful for format
1355 # selection purposes)
1356 if 'protocol' not in format:
1357 format['protocol'] = determine_protocol(format)
1358 # Add HTTP headers, so that external programs can use them from the
1360 full_format_info = info_dict.copy()
1361 full_format_info.update(format)
1362 format['http_headers'] = self._calc_headers(full_format_info)
1364 # TODO Central sorting goes here
1366 if formats[0] is not info_dict:
1367 # only set the 'formats' fields if the original info_dict list them
1368 # otherwise we end up with a circular reference, the first (and unique)
1369 # element in the 'formats' field in info_dict is info_dict itself,
1370 # which can't be exported to json
1371 info_dict['formats'] = formats
1372 if self.params.get('listformats'):
1373 self.list_formats(info_dict)
1376 req_format = self.params.get('format')
1377 if req_format is None:
1378 req_format_list = []
1379 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1380 not info_dict.get('is_live')):
1381 merger = FFmpegMergerPP(self)
1382 if merger.available and merger.can_merge():
1383 req_format_list.append('bestvideo+bestaudio')
1384 req_format_list.append('best')
1385 req_format = '/'.join(req_format_list)
1386 format_selector = self.build_format_selector(req_format)
1388 # While in format selection we may need to have an access to the original
1389 # format set in order to calculate some metrics or do some processing.
1390 # For now we need to be able to guess whether original formats provided
1391 # by extractor are incomplete or not (i.e. whether extractor provides only
1392 # video-only or audio-only formats) for proper formats selection for
1393 # extractors with such incomplete formats (see
1394 # https://github.com/rg3/youtube-dl/pull/5556).
1395 # Since formats may be filtered during format selection and may not match
1396 # the original formats the results may be incorrect. Thus original formats
1397 # or pre-calculated metrics should be passed to format selection routines
1399 # We will pass a context object containing all necessary additional data
1400 # instead of just formats.
1401 # This fixes incorrect format selection issue (see
1402 # https://github.com/rg3/youtube-dl/issues/10083).
1403 incomplete_formats = (
1404 # All formats are video-only or
1405 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) or
1406 # all formats are audio-only
1407 all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1411 'incomplete_formats': incomplete_formats,
1414 formats_to_download = list(format_selector(ctx))
1415 if not formats_to_download:
1416 raise ExtractorError('requested format not available',
1420 if len(formats_to_download) > 1:
1421 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1422 for format in formats_to_download:
1423 new_info = dict(info_dict)
1424 new_info.update(format)
1425 self.process_info(new_info)
1426 # We update the info dict with the best quality format (backwards compatibility)
1427 info_dict.update(formats_to_download[-1])
1430 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1431 """Select the requested subtitles and their format"""
1433 if normal_subtitles and self.params.get('writesubtitles'):
1434 available_subs.update(normal_subtitles)
1435 if automatic_captions and self.params.get('writeautomaticsub'):
1436 for lang, cap_info in automatic_captions.items():
1437 if lang not in available_subs:
1438 available_subs[lang] = cap_info
1440 if (not self.params.get('writesubtitles') and not
1441 self.params.get('writeautomaticsub') or not
1445 if self.params.get('allsubtitles', False):
1446 requested_langs = available_subs.keys()
1448 if self.params.get('subtitleslangs', False):
1449 requested_langs = self.params.get('subtitleslangs')
1450 elif 'en' in available_subs:
1451 requested_langs = ['en']
1453 requested_langs = [list(available_subs.keys())[0]]
1455 formats_query = self.params.get('subtitlesformat', 'best')
1456 formats_preference = formats_query.split('/') if formats_query else []
1458 for lang in requested_langs:
1459 formats = available_subs.get(lang)
1461 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1463 for ext in formats_preference:
1467 matches = list(filter(lambda f: f['ext'] == ext, formats))
1473 self.report_warning(
1474 'No subtitle format found matching "%s" for language %s, '
1475 'using %s' % (formats_query, lang, f['ext']))
1479 def process_info(self, info_dict):
1480 """Process a single resolved IE result."""
1482 assert info_dict.get('_type', 'video') == 'video'
1484 max_downloads = self.params.get('max_downloads')
1485 if max_downloads is not None:
1486 if self._num_downloads >= int(max_downloads):
1487 raise MaxDownloadsReached()
1489 info_dict['fulltitle'] = info_dict['title']
1490 if len(info_dict['title']) > 200:
1491 info_dict['title'] = info_dict['title'][:197] + '...'
1493 if 'format' not in info_dict:
1494 info_dict['format'] = info_dict['ext']
1496 reason = self._match_entry(info_dict, incomplete=False)
1497 if reason is not None:
1498 self.to_screen('[download] ' + reason)
1501 self._num_downloads += 1
1503 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1506 if self.params.get('forcetitle', False):
1507 self.to_stdout(info_dict['fulltitle'])
1508 if self.params.get('forceid', False):
1509 self.to_stdout(info_dict['id'])
1510 if self.params.get('forceurl', False):
1511 if info_dict.get('requested_formats') is not None:
1512 for f in info_dict['requested_formats']:
1513 self.to_stdout(f['url'] + f.get('play_path', ''))
1515 # For RTMP URLs, also include the playpath
1516 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1517 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1518 self.to_stdout(info_dict['thumbnail'])
1519 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1520 self.to_stdout(info_dict['description'])
1521 if self.params.get('forcefilename', False) and filename is not None:
1522 self.to_stdout(filename)
1523 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1524 self.to_stdout(formatSeconds(info_dict['duration']))
1525 if self.params.get('forceformat', False):
1526 self.to_stdout(info_dict['format'])
1527 if self.params.get('forcejson', False):
1528 self.to_stdout(json.dumps(info_dict))
1530 # Do nothing else if in simulate mode
1531 if self.params.get('simulate', False):
1534 if filename is None:
1538 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1539 if dn and not os.path.exists(dn):
1541 except (OSError, IOError) as err:
1542 self.report_error('unable to create directory ' + error_to_compat_str(err))
1545 if self.params.get('writedescription', False):
1546 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1547 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1548 self.to_screen('[info] Video description is already present')
1549 elif info_dict.get('description') is None:
1550 self.report_warning('There\'s no description to write.')
1553 self.to_screen('[info] Writing video description to: ' + descfn)
1554 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1555 descfile.write(info_dict['description'])
1556 except (OSError, IOError):
1557 self.report_error('Cannot write description file ' + descfn)
1560 if self.params.get('writeannotations', False):
1561 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1562 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1563 self.to_screen('[info] Video annotations are already present')
1566 self.to_screen('[info] Writing video annotations to: ' + annofn)
1567 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1568 annofile.write(info_dict['annotations'])
1569 except (KeyError, TypeError):
1570 self.report_warning('There are no annotations to write.')
1571 except (OSError, IOError):
1572 self.report_error('Cannot write annotations file: ' + annofn)
1575 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1576 self.params.get('writeautomaticsub')])
1578 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1579 # subtitles download errors are already managed as troubles in relevant IE
1580 # that way it will silently go on when used with unsupporting IE
1581 subtitles = info_dict['requested_subtitles']
1582 ie = self.get_info_extractor(info_dict['extractor_key'])
1583 for sub_lang, sub_info in subtitles.items():
1584 sub_format = sub_info['ext']
1585 if sub_info.get('data') is not None:
1586 sub_data = sub_info['data']
1589 sub_data = ie._download_webpage(
1590 sub_info['url'], info_dict['id'], note=False)
1591 except ExtractorError as err:
1592 self.report_warning('Unable to download subtitle for "%s": %s' %
1593 (sub_lang, error_to_compat_str(err.cause)))
1596 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1597 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1598 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1600 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1601 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1602 subfile.write(sub_data)
1603 except (OSError, IOError):
1604 self.report_error('Cannot write subtitles file ' + sub_filename)
1607 if self.params.get('writeinfojson', False):
1608 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1609 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1610 self.to_screen('[info] Video description metadata is already present')
1612 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1614 write_json_file(self.filter_requested_info(info_dict), infofn)
1615 except (OSError, IOError):
1616 self.report_error('Cannot write metadata to JSON file ' + infofn)
1619 self._write_thumbnails(info_dict, filename)
1621 if not self.params.get('skip_download', False):
1624 fd = get_suitable_downloader(info, self.params)(self, self.params)
1625 for ph in self._progress_hooks:
1626 fd.add_progress_hook(ph)
1627 if self.params.get('verbose'):
1628 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1629 return fd.download(name, info)
1631 if info_dict.get('requested_formats') is not None:
1634 merger = FFmpegMergerPP(self)
1635 if not merger.available:
1637 self.report_warning('You have requested multiple '
1638 'formats but ffmpeg or avconv are not installed.'
1639 ' The formats won\'t be merged.')
1641 postprocessors = [merger]
1643 def compatible_formats(formats):
1644 video, audio = formats
1646 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1647 if video_ext and audio_ext:
1649 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1652 for exts in COMPATIBLE_EXTS:
1653 if video_ext in exts and audio_ext in exts:
1655 # TODO: Check acodec/vcodec
1658 filename_real_ext = os.path.splitext(filename)[1][1:]
1660 os.path.splitext(filename)[0]
1661 if filename_real_ext == info_dict['ext']
1663 requested_formats = info_dict['requested_formats']
1664 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1665 info_dict['ext'] = 'mkv'
1666 self.report_warning(
1667 'Requested formats are incompatible for merge and will be merged into mkv.')
1668 # Ensure filename always has a correct extension for successful merge
1669 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1670 if os.path.exists(encodeFilename(filename)):
1672 '[download] %s has already been downloaded and '
1673 'merged' % filename)
1675 for f in requested_formats:
1676 new_info = dict(info_dict)
1678 fname = self.prepare_filename(new_info)
1679 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1680 downloaded.append(fname)
1681 partial_success = dl(fname, new_info)
1682 success = success and partial_success
1683 info_dict['__postprocessors'] = postprocessors
1684 info_dict['__files_to_merge'] = downloaded
1686 # Just a single file
1687 success = dl(filename, info_dict)
1688 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1689 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1691 except (OSError, IOError) as err:
1692 raise UnavailableVideoError(err)
1693 except (ContentTooShortError, ) as err:
1694 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1697 if success and filename != '-':
1699 fixup_policy = self.params.get('fixup')
1700 if fixup_policy is None:
1701 fixup_policy = 'detect_or_warn'
1703 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1705 stretched_ratio = info_dict.get('stretched_ratio')
1706 if stretched_ratio is not None and stretched_ratio != 1:
1707 if fixup_policy == 'warn':
1708 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1709 info_dict['id'], stretched_ratio))
1710 elif fixup_policy == 'detect_or_warn':
1711 stretched_pp = FFmpegFixupStretchedPP(self)
1712 if stretched_pp.available:
1713 info_dict.setdefault('__postprocessors', [])
1714 info_dict['__postprocessors'].append(stretched_pp)
1716 self.report_warning(
1717 '%s: Non-uniform pixel ratio (%s). %s'
1718 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1720 assert fixup_policy in ('ignore', 'never')
1722 if (info_dict.get('requested_formats') is None and
1723 info_dict.get('container') == 'm4a_dash'):
1724 if fixup_policy == 'warn':
1725 self.report_warning(
1726 '%s: writing DASH m4a. '
1727 'Only some players support this container.'
1729 elif fixup_policy == 'detect_or_warn':
1730 fixup_pp = FFmpegFixupM4aPP(self)
1731 if fixup_pp.available:
1732 info_dict.setdefault('__postprocessors', [])
1733 info_dict['__postprocessors'].append(fixup_pp)
1735 self.report_warning(
1736 '%s: writing DASH m4a. '
1737 'Only some players support this container. %s'
1738 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1740 assert fixup_policy in ('ignore', 'never')
1742 if (info_dict.get('protocol') == 'm3u8_native' or
1743 info_dict.get('protocol') == 'm3u8' and
1744 self.params.get('hls_prefer_native')):
1745 if fixup_policy == 'warn':
1746 self.report_warning('%s: malformated aac bitstream.' % (
1748 elif fixup_policy == 'detect_or_warn':
1749 fixup_pp = FFmpegFixupM3u8PP(self)
1750 if fixup_pp.available:
1751 info_dict.setdefault('__postprocessors', [])
1752 info_dict['__postprocessors'].append(fixup_pp)
1754 self.report_warning(
1755 '%s: malformated aac bitstream. %s'
1756 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1758 assert fixup_policy in ('ignore', 'never')
1761 self.post_process(filename, info_dict)
1762 except (PostProcessingError) as err:
1763 self.report_error('postprocessing: %s' % str(err))
1765 self.record_download_archive(info_dict)
1767 def download(self, url_list):
1768 """Download a given list of URLs."""
1769 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1770 if (len(url_list) > 1 and
1771 '%' not in outtmpl and
1772 self.params.get('max_downloads') != 1):
1773 raise SameFileError(outtmpl)
1775 for url in url_list:
1777 # It also downloads the videos
1778 res = self.extract_info(
1779 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1780 except UnavailableVideoError:
1781 self.report_error('unable to download video')
1782 except MaxDownloadsReached:
1783 self.to_screen('[info] Maximum number of downloaded files reached.')
1786 if self.params.get('dump_single_json', False):
1787 self.to_stdout(json.dumps(res))
1789 return self._download_retcode
1791 def download_with_info_file(self, info_filename):
1792 with contextlib.closing(fileinput.FileInput(
1793 [info_filename], mode='r',
1794 openhook=fileinput.hook_encoded('utf-8'))) as f:
1795 # FileInput doesn't have a read method, we can't call json.load
1796 info = self.filter_requested_info(json.loads('\n'.join(f)))
1798 self.process_ie_result(info, download=True)
1799 except DownloadError:
1800 webpage_url = info.get('webpage_url')
1801 if webpage_url is not None:
1802 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1803 return self.download([webpage_url])
1806 return self._download_retcode
1809 def filter_requested_info(info_dict):
1811 (k, v) for k, v in info_dict.items()
1812 if k not in ['requested_formats', 'requested_subtitles'])
1814 def post_process(self, filename, ie_info):
1815 """Run all the postprocessors on the given file."""
1816 info = dict(ie_info)
1817 info['filepath'] = filename
1819 if ie_info.get('__postprocessors') is not None:
1820 pps_chain.extend(ie_info['__postprocessors'])
1821 pps_chain.extend(self._pps)
1822 for pp in pps_chain:
1823 files_to_delete = []
1825 files_to_delete, info = pp.run(info)
1826 except PostProcessingError as e:
1827 self.report_error(e.msg)
1828 if files_to_delete and not self.params.get('keepvideo', False):
1829 for old_filename in files_to_delete:
1830 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1832 os.remove(encodeFilename(old_filename))
1833 except (IOError, OSError):
1834 self.report_warning('Unable to remove downloaded original file')
1836 def _make_archive_id(self, info_dict):
1837 # Future-proof against any change in case
1838 # and backwards compatibility with prior versions
1839 extractor = info_dict.get('extractor_key')
1840 if extractor is None:
1841 if 'id' in info_dict:
1842 extractor = info_dict.get('ie_key') # key in a playlist
1843 if extractor is None:
1844 return None # Incomplete video information
1845 return extractor.lower() + ' ' + info_dict['id']
1847 def in_download_archive(self, info_dict):
1848 fn = self.params.get('download_archive')
1852 vid_id = self._make_archive_id(info_dict)
1854 return False # Incomplete video information
1857 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1858 for line in archive_file:
1859 if line.strip() == vid_id:
1861 except IOError as ioe:
1862 if ioe.errno != errno.ENOENT:
1866 def record_download_archive(self, info_dict):
1867 fn = self.params.get('download_archive')
1870 vid_id = self._make_archive_id(info_dict)
1872 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1873 archive_file.write(vid_id + '\n')
1876 def format_resolution(format, default='unknown'):
1877 if format.get('vcodec') == 'none':
1879 if format.get('resolution') is not None:
1880 return format['resolution']
1881 if format.get('height') is not None:
1882 if format.get('width') is not None:
1883 res = '%sx%s' % (format['width'], format['height'])
1885 res = '%sp' % format['height']
1886 elif format.get('width') is not None:
1887 res = '%dx?' % format['width']
1892 def _format_note(self, fdict):
1894 if fdict.get('ext') in ['f4f', 'f4m']:
1895 res += '(unsupported) '
1896 if fdict.get('language'):
1899 res += '[%s] ' % fdict['language']
1900 if fdict.get('format_note') is not None:
1901 res += fdict['format_note'] + ' '
1902 if fdict.get('tbr') is not None:
1903 res += '%4dk ' % fdict['tbr']
1904 if fdict.get('container') is not None:
1907 res += '%s container' % fdict['container']
1908 if (fdict.get('vcodec') is not None and
1909 fdict.get('vcodec') != 'none'):
1912 res += fdict['vcodec']
1913 if fdict.get('vbr') is not None:
1915 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1917 if fdict.get('vbr') is not None:
1918 res += '%4dk' % fdict['vbr']
1919 if fdict.get('fps') is not None:
1922 res += '%sfps' % fdict['fps']
1923 if fdict.get('acodec') is not None:
1926 if fdict['acodec'] == 'none':
1929 res += '%-5s' % fdict['acodec']
1930 elif fdict.get('abr') is not None:
1934 if fdict.get('abr') is not None:
1935 res += '@%3dk' % fdict['abr']
1936 if fdict.get('asr') is not None:
1937 res += ' (%5dHz)' % fdict['asr']
1938 if fdict.get('filesize') is not None:
1941 res += format_bytes(fdict['filesize'])
1942 elif fdict.get('filesize_approx') is not None:
1945 res += '~' + format_bytes(fdict['filesize_approx'])
1948 def list_formats(self, info_dict):
1949 formats = info_dict.get('formats', [info_dict])
1951 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1953 if f.get('preference') is None or f['preference'] >= -1000]
1954 if len(formats) > 1:
1955 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1957 header_line = ['format code', 'extension', 'resolution', 'note']
1959 '[info] Available formats for %s:\n%s' %
1960 (info_dict['id'], render_table(header_line, table)))
1962 def list_thumbnails(self, info_dict):
1963 thumbnails = info_dict.get('thumbnails')
1965 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
1969 '[info] Thumbnails for %s:' % info_dict['id'])
1970 self.to_screen(render_table(
1971 ['ID', 'width', 'height', 'URL'],
1972 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1974 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1976 self.to_screen('%s has no %s' % (video_id, name))
1979 'Available %s for %s:' % (name, video_id))
1980 self.to_screen(render_table(
1981 ['Language', 'formats'],
1982 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1983 for lang, formats in subtitles.items()]))
1985 def urlopen(self, req):
1986 """ Start an HTTP download """
1987 if isinstance(req, compat_basestring):
1988 req = sanitized_Request(req)
1989 return self._opener.open(req, timeout=self._socket_timeout)
1991 def print_debug_header(self):
1992 if not self.params.get('verbose'):
1995 if type('') is not compat_str:
1996 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1997 self.report_warning(
1998 'Your Python is broken! Update to a newer and supported version')
2000 stdout_encoding = getattr(
2001 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2003 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2004 locale.getpreferredencoding(),
2005 sys.getfilesystemencoding(),
2007 self.get_encoding()))
2008 write_string(encoding_str, encoding=None)
2010 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
2012 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2014 sp = subprocess.Popen(
2015 ['git', 'rev-parse', '--short', 'HEAD'],
2016 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2017 cwd=os.path.dirname(os.path.abspath(__file__)))
2018 out, err = sp.communicate()
2019 out = out.decode().strip()
2020 if re.match('[0-9a-f]+', out):
2021 self._write_string('[debug] Git HEAD: ' + out + '\n')
2027 self._write_string('[debug] Python version %s - %s\n' % (
2028 platform.python_version(), platform_name()))
2030 exe_versions = FFmpegPostProcessor.get_versions(self)
2031 exe_versions['rtmpdump'] = rtmpdump_version()
2032 exe_str = ', '.join(
2034 for exe, v in sorted(exe_versions.items())
2039 self._write_string('[debug] exe versions: %s\n' % exe_str)
2042 for handler in self._opener.handlers:
2043 if hasattr(handler, 'proxies'):
2044 proxy_map.update(handler.proxies)
2045 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2047 if self.params.get('call_home', False):
2048 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2049 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2050 latest_version = self.urlopen(
2051 'https://yt-dl.org/latest/version').read().decode('utf-8')
2052 if version_tuple(latest_version) > version_tuple(__version__):
2053 self.report_warning(
2054 'You are using an outdated version (newest version: %s)! '
2055 'See https://yt-dl.org/update if you need help updating.' %
2058 def _setup_opener(self):
2059 timeout_val = self.params.get('socket_timeout')
2060 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2062 opts_cookiefile = self.params.get('cookiefile')
2063 opts_proxy = self.params.get('proxy')
2065 if opts_cookiefile is None:
2066 self.cookiejar = compat_cookiejar.CookieJar()
2068 opts_cookiefile = compat_expanduser(opts_cookiefile)
2069 self.cookiejar = compat_cookiejar.MozillaCookieJar(
2071 if os.access(opts_cookiefile, os.R_OK):
2072 self.cookiejar.load()
2074 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2075 if opts_proxy is not None:
2076 if opts_proxy == '':
2079 proxies = {'http': opts_proxy, 'https': opts_proxy}
2081 proxies = compat_urllib_request.getproxies()
2082 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2083 if 'http' in proxies and 'https' not in proxies:
2084 proxies['https'] = proxies['http']
2085 proxy_handler = PerRequestProxyHandler(proxies)
2087 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2088 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2089 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2090 data_handler = compat_urllib_request_DataHandler()
2092 # When passing our own FileHandler instance, build_opener won't add the
2093 # default FileHandler and allows us to disable the file protocol, which
2094 # can be used for malicious purposes (see
2095 # https://github.com/rg3/youtube-dl/issues/8227)
2096 file_handler = compat_urllib_request.FileHandler()
2098 def file_open(*args, **kwargs):
2099 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2100 file_handler.file_open = file_open
2102 opener = compat_urllib_request.build_opener(
2103 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2105 # Delete the default user-agent header, which would otherwise apply in
2106 # cases where our custom HTTP handler doesn't come into play
2107 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2108 opener.addheaders = []
2109 self._opener = opener
2111 def encode(self, s):
2112 if isinstance(s, bytes):
2113 return s # Already encoded
2116 return s.encode(self.get_encoding())
2117 except UnicodeEncodeError as err:
2118 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2121 def get_encoding(self):
2122 encoding = self.params.get('encoding')
2123 if encoding is None:
2124 encoding = preferredencoding()
2127 def _write_thumbnails(self, info_dict, filename):
2128 if self.params.get('writethumbnail', False):
2129 thumbnails = info_dict.get('thumbnails')
2131 thumbnails = [thumbnails[-1]]
2132 elif self.params.get('write_all_thumbnails', False):
2133 thumbnails = info_dict.get('thumbnails')
2138 # No thumbnails present, so return immediately
2141 for t in thumbnails:
2142 thumb_ext = determine_ext(t['url'], 'jpg')
2143 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2144 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2145 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2147 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2148 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2149 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2151 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2152 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2154 uf = self.urlopen(t['url'])
2155 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2156 shutil.copyfileobj(uf, thumbf)
2157 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2158 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2159 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2160 self.report_warning('Unable to download thumbnail "%s": %s' %
2161 (t['url'], error_to_compat_str(err)))