2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import, unicode_literals
34 compat_get_terminal_size,
38 compat_tokenize_tokenize,
40 compat_urllib_request,
41 compat_urllib_request_DataHandler,
62 PerRequestProxyHandler,
73 UnavailableVideoError,
78 YoutubeDLCookieProcessor,
85 from .cache import Cache
86 from .extractor import get_info_extractor, gen_extractors
87 from .downloader import get_suitable_downloader
88 from .downloader.rtmp import rtmpdump_version
89 from .postprocessor import (
91 FFmpegFixupStretchedPP,
96 from .version import __version__
99 class YoutubeDL(object):
102 YoutubeDL objects are the ones responsible of downloading the
103 actual video file and writing it to disk if the user has requested
104 it, among some other tasks. In most cases there should be one per
105 program. As, given a video URL, the downloader doesn't know how to
106 extract all the needed information, task that InfoExtractors do, it
107 has to pass the URL to one of them.
109 For this, YoutubeDL objects have a method that allows
110 InfoExtractors to be registered in a given order. When it is passed
111 a URL, the YoutubeDL object handles it to the first InfoExtractor it
112 finds that reports being able to handle it. The InfoExtractor extracts
113 all the information about the video or videos the URL refers to, and
114 YoutubeDL process the extracted information, possibly using a File
115 Downloader to download the video.
117 YoutubeDL objects accept a lot of parameters. In order not to saturate
118 the object constructor with arguments, it receives a dictionary of
119 options instead. These options are available through the params
120 attribute for the InfoExtractors to use. The YoutubeDL also
121 registers itself as the downloader in charge for the InfoExtractors
122 that are added to it, so this is a "mutual registration".
126 username: Username for authentication purposes.
127 password: Password for authentication purposes.
128 videopassword: Password for accessing a video.
129 usenetrc: Use netrc for authentication instead.
130 verbose: Print additional info to stdout.
131 quiet: Do not print messages to stdout.
132 no_warnings: Do not print out anything for warnings.
133 forceurl: Force printing final URL.
134 forcetitle: Force printing title.
135 forceid: Force printing ID.
136 forcethumbnail: Force printing thumbnail URL.
137 forcedescription: Force printing description.
138 forcefilename: Force printing final filename.
139 forceduration: Force printing duration.
140 forcejson: Force printing info_dict as JSON.
141 dump_single_json: Force printing the info_dict of the whole playlist
142 (or video) as a single JSON line.
143 simulate: Do not download the video files.
144 format: Video format code. See options.py for more information.
145 outtmpl: Template for output names.
146 restrictfilenames: Do not allow "&" and spaces in file names
147 ignoreerrors: Do not stop on download errors.
148 force_generic_extractor: Force downloader to use the generic extractor
149 nooverwrites: Prevent overwriting files.
150 playliststart: Playlist item to start at.
151 playlistend: Playlist item to end at.
152 playlist_items: Specific indices of playlist to download.
153 playlistreverse: Download playlist items in reverse order.
154 matchtitle: Download only matching titles.
155 rejecttitle: Reject downloads for matching titles.
156 logger: Log messages to a logging.Logger instance.
157 logtostderr: Log messages to stderr instead of stdout.
158 writedescription: Write the video description to a .description file
159 writeinfojson: Write the video description to a .info.json file
160 writeannotations: Write the video annotations to a .annotations.xml file
161 writethumbnail: Write the thumbnail image to a file
162 write_all_thumbnails: Write all thumbnail formats to files
163 writesubtitles: Write the video subtitles to a file
164 writeautomaticsub: Write the automatically generated subtitles to a file
165 allsubtitles: Downloads all the subtitles of the video
166 (requires writesubtitles or writeautomaticsub)
167 listsubtitles: Lists all available subtitles for the video
168 subtitlesformat: The format code for subtitles
169 subtitleslangs: List of languages of the subtitles to download
170 keepvideo: Keep the video file after post-processing
171 daterange: A DateRange object, download only if the upload_date is in the range.
172 skip_download: Skip the actual download of the video file
173 cachedir: Location of the cache files in the filesystem.
174 False to disable filesystem cache.
175 noplaylist: Download single video instead of a playlist if in doubt.
176 age_limit: An integer representing the user's age in years.
177 Unsuitable videos for the given age are skipped.
178 min_views: An integer representing the minimum view count the video
179 must have in order to not be skipped.
180 Videos without view count information are always
181 downloaded. None for no limit.
182 max_views: An integer representing the maximum view count.
183 Videos that are more popular than that are not
185 Videos without view count information are always
186 downloaded. None for no limit.
187 download_archive: File name of a file where all downloads are recorded.
188 Videos already present in the file are not downloaded
190 cookiefile: File name where cookies should be read from and dumped to.
191 nocheckcertificate:Do not verify SSL certificates
192 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
193 At the moment, this is only supported by YouTube.
194 proxy: URL of the proxy server to use
195 cn_verification_proxy: URL of the proxy to use for IP address verification
196 on Chinese sites. (Experimental)
197 socket_timeout: Time to wait for unresponsive hosts, in seconds
198 bidi_workaround: Work around buggy terminals without bidirectional text
199 support, using fridibi
200 debug_printtraffic:Print out sent and received HTTP traffic
201 include_ads: Download ads as well
202 default_search: Prepend this string if an input url is not valid.
203 'auto' for elaborate guessing
204 encoding: Use this encoding instead of the system-specified.
205 extract_flat: Do not resolve URLs, return the immediate result.
206 Pass in 'in_playlist' to only show this behavior for
208 postprocessors: A list of dictionaries, each with an entry
209 * key: The name of the postprocessor. See
210 youtube_dl/postprocessor/__init__.py for a list.
211 as well as any further keyword arguments for the
213 progress_hooks: A list of functions that get called on download
214 progress, with a dictionary with the entries
215 * status: One of "downloading", "error", or "finished".
216 Check this first and ignore unknown values.
218 If status is one of "downloading", or "finished", the
219 following properties may also be present:
220 * filename: The final filename (always present)
221 * tmpfilename: The filename we're currently writing to
222 * downloaded_bytes: Bytes on disk
223 * total_bytes: Size of the whole file, None if unknown
224 * total_bytes_estimate: Guess of the eventual file size,
226 * elapsed: The number of seconds since download started.
227 * eta: The estimated time in seconds, None if unknown
228 * speed: The download speed in bytes/second, None if
230 * fragment_index: The counter of the currently
231 downloaded video fragment.
232 * fragment_count: The number of fragments (= individual
233 files that will be merged)
235 Progress hooks are guaranteed to be called at least once
236 (with status "finished") if the download is successful.
237 merge_output_format: Extension to use when merging formats.
238 fixup: Automatically correct known faults of the file.
240 - "never": do nothing
241 - "warn": only emit a warning
242 - "detect_or_warn": check whether we can do anything
243 about it, warn otherwise (default)
244 source_address: (Experimental) Client-side IP address to bind to.
245 call_home: Boolean, true iff we are allowed to contact the
246 youtube-dl servers for debugging.
247 sleep_interval: Number of seconds to sleep before each download.
248 listformats: Print an overview of available video formats and exit.
249 list_thumbnails: Print a table of all thumbnails and exit.
250 match_filter: A function that gets called with the info_dict of
252 If it returns a message, the video is ignored.
253 If it returns None, the video is downloaded.
254 match_filter_func in utils.py is one example for this.
255 no_color: Do not emit color codes in output.
257 The following options determine which downloader is picked:
258 external_downloader: Executable of the external downloader to call.
259 None or unset for standard (built-in) downloader.
260 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv.
262 The following parameters are not used by YoutubeDL itself, they are used by
263 the downloader (see youtube_dl/downloader/common.py):
264 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
265 noresizebuffer, retries, continuedl, noprogress, consoletitle,
266 xattr_set_filesize, external_downloader_args.
268 The following options are used by the post processors:
269 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
270 otherwise prefer avconv.
271 postprocessor_args: A list of additional command-line arguments for the
278 _download_retcode = None
279 _num_downloads = None
282 def __init__(self, params=None, auto_init=True):
283 """Create a FileDownloader object with the given options."""
287 self._ies_instances = {}
289 self._progress_hooks = []
290 self._download_retcode = 0
291 self._num_downloads = 0
292 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
293 self._err_file = sys.stderr
296 'nocheckcertificate': False,
298 self.params.update(params)
299 self.cache = Cache(self)
301 if params.get('bidi_workaround', False):
304 master, slave = pty.openpty()
305 width = compat_get_terminal_size().columns
309 width_args = ['-w', str(width)]
311 stdin=subprocess.PIPE,
313 stderr=self._err_file)
315 self._output_process = subprocess.Popen(
316 ['bidiv'] + width_args, **sp_kwargs
319 self._output_process = subprocess.Popen(
320 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
321 self._output_channel = os.fdopen(master, 'rb')
322 except OSError as ose:
324 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 $PATH.')
328 if (sys.version_info >= (3,) and sys.platform != 'win32' and
329 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
330 not params.get('restrictfilenames', False)):
331 # On Python 3, the Unicode filesystem API will throw errors (#1474)
333 'Assuming --restrict-filenames since file system encoding '
334 'cannot encode all characters. '
335 'Set the LC_ALL environment variable to fix this.')
336 self.params['restrictfilenames'] = True
338 if isinstance(params.get('outtmpl'), bytes):
340 'Parameter outtmpl is bytes, but should be a unicode string. '
341 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
346 self.print_debug_header()
347 self.add_default_info_extractors()
349 for pp_def_raw in self.params.get('postprocessors', []):
350 pp_class = get_postprocessor(pp_def_raw['key'])
351 pp_def = dict(pp_def_raw)
353 pp = pp_class(self, **compat_kwargs(pp_def))
354 self.add_post_processor(pp)
356 for ph in self.params.get('progress_hooks', []):
357 self.add_progress_hook(ph)
359 def warn_if_short_id(self, argv):
360 # short YouTube ID starting with dash?
362 i for i, a in enumerate(argv)
363 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
367 [a for i, a in enumerate(argv) if i not in idxs] +
368 ['--'] + [argv[i] for i in idxs]
371 'Long argument string detected. '
372 'Use -- to separate parameters and URLs, like this:\n%s\n' %
373 args_to_str(correct_argv))
375 def add_info_extractor(self, ie):
376 """Add an InfoExtractor object to the end of the list."""
378 self._ies_instances[ie.ie_key()] = ie
379 ie.set_downloader(self)
381 def get_info_extractor(self, ie_key):
383 Get an instance of an IE with name ie_key, it will try to get one from
384 the _ies list, if there's no instance it will create a new one and add
385 it to the extractor list.
387 ie = self._ies_instances.get(ie_key)
389 ie = get_info_extractor(ie_key)()
390 self.add_info_extractor(ie)
393 def add_default_info_extractors(self):
395 Add the InfoExtractors returned by gen_extractors to the end of the list
397 for ie in gen_extractors():
398 self.add_info_extractor(ie)
400 def add_post_processor(self, pp):
401 """Add a PostProcessor object to the end of the chain."""
403 pp.set_downloader(self)
405 def add_progress_hook(self, ph):
406 """Add the progress hook (currently only for the file downloader)"""
407 self._progress_hooks.append(ph)
409 def _bidi_workaround(self, message):
410 if not hasattr(self, '_output_channel'):
413 assert hasattr(self, '_output_process')
414 assert isinstance(message, compat_str)
415 line_count = message.count('\n') + 1
416 self._output_process.stdin.write((message + '\n').encode('utf-8'))
417 self._output_process.stdin.flush()
418 res = ''.join(self._output_channel.readline().decode('utf-8')
419 for _ in range(line_count))
420 return res[:-len('\n')]
422 def to_screen(self, message, skip_eol=False):
423 """Print message to stdout if not in quiet mode."""
424 return self.to_stdout(message, skip_eol, check_quiet=True)
426 def _write_string(self, s, out=None):
427 write_string(s, out=out, encoding=self.params.get('encoding'))
429 def to_stdout(self, message, skip_eol=False, check_quiet=False):
430 """Print message to stdout if not in quiet mode."""
431 if self.params.get('logger'):
432 self.params['logger'].debug(message)
433 elif not check_quiet or not self.params.get('quiet', False):
434 message = self._bidi_workaround(message)
435 terminator = ['\n', ''][skip_eol]
436 output = message + terminator
438 self._write_string(output, self._screen_file)
440 def to_stderr(self, message):
441 """Print message to stderr."""
442 assert isinstance(message, compat_str)
443 if self.params.get('logger'):
444 self.params['logger'].error(message)
446 message = self._bidi_workaround(message)
447 output = message + '\n'
448 self._write_string(output, self._err_file)
450 def to_console_title(self, message):
451 if not self.params.get('consoletitle', False):
453 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
454 # c_wchar_p() might not be necessary if `message` is
455 # already of type unicode()
456 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
457 elif 'TERM' in os.environ:
458 self._write_string('\033]0;%s\007' % message, self._screen_file)
460 def save_console_title(self):
461 if not self.params.get('consoletitle', False):
463 if 'TERM' in os.environ:
464 # Save the title on stack
465 self._write_string('\033[22;0t', self._screen_file)
467 def restore_console_title(self):
468 if not self.params.get('consoletitle', False):
470 if 'TERM' in os.environ:
471 # Restore the title from stack
472 self._write_string('\033[23;0t', self._screen_file)
475 self.save_console_title()
478 def __exit__(self, *args):
479 self.restore_console_title()
481 if self.params.get('cookiefile') is not None:
482 self.cookiejar.save()
484 def trouble(self, message=None, tb=None):
485 """Determine action to take when a download problem appears.
487 Depending on if the downloader has been configured to ignore
488 download errors or not, this method may throw an exception or
489 not when errors are found, after printing the message.
491 tb, if given, is additional traceback information.
493 if message is not None:
494 self.to_stderr(message)
495 if self.params.get('verbose'):
497 if sys.exc_info()[0]: # if .trouble has been called from an except block
499 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
500 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
501 tb += encode_compat_str(traceback.format_exc())
503 tb_data = traceback.format_list(traceback.extract_stack())
504 tb = ''.join(tb_data)
506 if not self.params.get('ignoreerrors', False):
507 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
508 exc_info = sys.exc_info()[1].exc_info
510 exc_info = sys.exc_info()
511 raise DownloadError(message, exc_info)
512 self._download_retcode = 1
514 def report_warning(self, message):
516 Print the message to stderr, it will be prefixed with 'WARNING:'
517 If stderr is a tty file the 'WARNING:' will be colored
519 if self.params.get('logger') is not None:
520 self.params['logger'].warning(message)
522 if self.params.get('no_warnings'):
524 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
525 _msg_header = '\033[0;33mWARNING:\033[0m'
527 _msg_header = 'WARNING:'
528 warning_message = '%s %s' % (_msg_header, message)
529 self.to_stderr(warning_message)
531 def report_error(self, message, tb=None):
533 Do the same as trouble, but prefixes the message with 'ERROR:', colored
534 in red if stderr is a tty file.
536 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
537 _msg_header = '\033[0;31mERROR:\033[0m'
539 _msg_header = 'ERROR:'
540 error_message = '%s %s' % (_msg_header, message)
541 self.trouble(error_message, tb)
543 def report_file_already_downloaded(self, file_name):
544 """Report file has already been fully downloaded."""
546 self.to_screen('[download] %s has already been downloaded' % file_name)
547 except UnicodeEncodeError:
548 self.to_screen('[download] The file has already been downloaded')
550 def prepare_filename(self, info_dict):
551 """Generate the output filename."""
553 template_dict = dict(info_dict)
555 template_dict['epoch'] = int(time.time())
556 autonumber_size = self.params.get('autonumber_size')
557 if autonumber_size is None:
559 autonumber_templ = '%0' + str(autonumber_size) + 'd'
560 template_dict['autonumber'] = autonumber_templ % self._num_downloads
561 if template_dict.get('playlist_index') is not None:
562 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
563 if template_dict.get('resolution') is None:
564 if template_dict.get('width') and template_dict.get('height'):
565 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
566 elif template_dict.get('height'):
567 template_dict['resolution'] = '%sp' % template_dict['height']
568 elif template_dict.get('width'):
569 template_dict['resolution'] = '?x%d' % template_dict['width']
571 sanitize = lambda k, v: sanitize_filename(
573 restricted=self.params.get('restrictfilenames'),
575 template_dict = dict((k, sanitize(k, v))
576 for k, v in template_dict.items()
578 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
580 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
581 tmpl = compat_expanduser(outtmpl)
582 filename = tmpl % template_dict
583 # Temporary fix for #4787
584 # 'Treat' all problem characters by passing filename through preferredencoding
585 # to workaround encoding issues with subprocess on python2 @ Windows
586 if sys.version_info < (3, 0) and sys.platform == 'win32':
587 filename = encodeFilename(filename, True).decode(preferredencoding())
588 return sanitize_path(filename)
589 except ValueError as err:
590 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
593 def _match_entry(self, info_dict, incomplete):
594 """ Returns None iff the file should be downloaded """
596 video_title = info_dict.get('title', info_dict.get('id', 'video'))
597 if 'title' in info_dict:
598 # This can happen when we're just evaluating the playlist
599 title = info_dict['title']
600 matchtitle = self.params.get('matchtitle', False)
602 if not re.search(matchtitle, title, re.IGNORECASE):
603 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
604 rejecttitle = self.params.get('rejecttitle', False)
606 if re.search(rejecttitle, title, re.IGNORECASE):
607 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
608 date = info_dict.get('upload_date', None)
610 dateRange = self.params.get('daterange', DateRange())
611 if date not in dateRange:
612 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
613 view_count = info_dict.get('view_count', None)
614 if view_count is not None:
615 min_views = self.params.get('min_views')
616 if min_views is not None and view_count < min_views:
617 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
618 max_views = self.params.get('max_views')
619 if max_views is not None and view_count > max_views:
620 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
621 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
622 return 'Skipping "%s" because it is age restricted' % video_title
623 if self.in_download_archive(info_dict):
624 return '%s has already been recorded in archive' % video_title
627 match_filter = self.params.get('match_filter')
628 if match_filter is not None:
629 ret = match_filter(info_dict)
636 def add_extra_info(info_dict, extra_info):
637 '''Set the keys from extra_info in info dict if they are missing'''
638 for key, value in extra_info.items():
639 info_dict.setdefault(key, value)
641 def extract_info(self, url, download=True, ie_key=None, extra_info={},
642 process=True, force_generic_extractor=False):
644 Returns a list with a dictionary for each video we find.
645 If 'download', also downloads the videos.
646 extra_info is a dict containing the extra values to add to each result
649 if not ie_key and force_generic_extractor:
653 ies = [self.get_info_extractor(ie_key)]
658 if not ie.suitable(url):
662 self.report_warning('The program functionality for this site has been marked as broken, '
663 'and will probably not work.')
666 ie_result = ie.extract(url)
667 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
669 if isinstance(ie_result, list):
670 # Backwards compatibility: old IE result format
672 '_type': 'compat_list',
673 'entries': ie_result,
675 self.add_default_extra_info(ie_result, ie, url)
677 return self.process_ie_result(ie_result, download, extra_info)
680 except ExtractorError as e: # An error we somewhat expected
681 self.report_error(compat_str(e), e.format_traceback())
683 except MaxDownloadsReached:
685 except Exception as e:
686 if self.params.get('ignoreerrors', False):
687 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
692 self.report_error('no suitable InfoExtractor for URL %s' % url)
694 def add_default_extra_info(self, ie_result, ie, url):
695 self.add_extra_info(ie_result, {
696 'extractor': ie.IE_NAME,
698 'webpage_url_basename': url_basename(url),
699 'extractor_key': ie.ie_key(),
702 def process_ie_result(self, ie_result, download=True, extra_info={}):
704 Take the result of the ie(may be modified) and resolve all unresolved
705 references (URLs, playlist items).
707 It will also download the videos if 'download'.
708 Returns the resolved ie_result.
711 result_type = ie_result.get('_type', 'video')
713 if result_type in ('url', 'url_transparent'):
714 extract_flat = self.params.get('extract_flat', False)
715 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
716 extract_flat is True):
717 if self.params.get('forcejson', False):
718 self.to_stdout(json.dumps(ie_result))
721 if result_type == 'video':
722 self.add_extra_info(ie_result, extra_info)
723 return self.process_video_result(ie_result, download=download)
724 elif result_type == 'url':
725 # We have to add extra_info to the results because it may be
726 # contained in a playlist
727 return self.extract_info(ie_result['url'],
729 ie_key=ie_result.get('ie_key'),
730 extra_info=extra_info)
731 elif result_type == 'url_transparent':
732 # Use the information from the embedding page
733 info = self.extract_info(
734 ie_result['url'], ie_key=ie_result.get('ie_key'),
735 extra_info=extra_info, download=False, process=False)
737 force_properties = dict(
738 (k, v) for k, v in ie_result.items() if v is not None)
739 for f in ('_type', 'url'):
740 if f in force_properties:
741 del force_properties[f]
742 new_result = info.copy()
743 new_result.update(force_properties)
745 assert new_result.get('_type') != 'url_transparent'
747 return self.process_ie_result(
748 new_result, download=download, extra_info=extra_info)
749 elif result_type == 'playlist' or result_type == 'multi_video':
750 # We process each entry in the playlist
751 playlist = ie_result.get('title', None) or ie_result.get('id', None)
752 self.to_screen('[download] Downloading playlist: %s' % playlist)
754 playlist_results = []
756 playliststart = self.params.get('playliststart', 1) - 1
757 playlistend = self.params.get('playlistend', None)
758 # For backwards compatibility, interpret -1 as whole list
759 if playlistend == -1:
762 playlistitems_str = self.params.get('playlist_items', None)
764 if playlistitems_str is not None:
765 def iter_playlistitems(format):
766 for string_segment in format.split(','):
767 if '-' in string_segment:
768 start, end = string_segment.split('-')
769 for item in range(int(start), int(end) + 1):
772 yield int(string_segment)
773 playlistitems = iter_playlistitems(playlistitems_str)
775 ie_entries = ie_result['entries']
776 if isinstance(ie_entries, list):
777 n_all_entries = len(ie_entries)
780 ie_entries[i - 1] for i in playlistitems
781 if -n_all_entries <= i - 1 < n_all_entries]
783 entries = ie_entries[playliststart:playlistend]
784 n_entries = len(entries)
786 "[%s] playlist %s: Collected %d video ids (downloading %d of them)" %
787 (ie_result['extractor'], playlist, n_all_entries, n_entries))
788 elif isinstance(ie_entries, PagedList):
791 for item in playlistitems:
792 entries.extend(ie_entries.getslice(
796 entries = ie_entries.getslice(
797 playliststart, playlistend)
798 n_entries = len(entries)
800 "[%s] playlist %s: Downloading %d videos" %
801 (ie_result['extractor'], playlist, n_entries))
804 entry_list = list(ie_entries)
805 entries = [entry_list[i - 1] for i in playlistitems]
807 entries = list(itertools.islice(
808 ie_entries, playliststart, playlistend))
809 n_entries = len(entries)
811 "[%s] playlist %s: Downloading %d videos" %
812 (ie_result['extractor'], playlist, n_entries))
814 if self.params.get('playlistreverse', False):
815 entries = entries[::-1]
817 for i, entry in enumerate(entries, 1):
818 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
820 'n_entries': n_entries,
821 'playlist': playlist,
822 'playlist_id': ie_result.get('id'),
823 'playlist_title': ie_result.get('title'),
824 'playlist_index': i + playliststart,
825 'extractor': ie_result['extractor'],
826 'webpage_url': ie_result['webpage_url'],
827 'webpage_url_basename': url_basename(ie_result['webpage_url']),
828 'extractor_key': ie_result['extractor_key'],
831 reason = self._match_entry(entry, incomplete=True)
832 if reason is not None:
833 self.to_screen('[download] ' + reason)
836 entry_result = self.process_ie_result(entry,
839 playlist_results.append(entry_result)
840 ie_result['entries'] = playlist_results
841 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
843 elif result_type == 'compat_list':
845 'Extractor %s returned a compat_list result. '
846 'It needs to be updated.' % ie_result.get('extractor'))
852 'extractor': ie_result['extractor'],
853 'webpage_url': ie_result['webpage_url'],
854 'webpage_url_basename': url_basename(ie_result['webpage_url']),
855 'extractor_key': ie_result['extractor_key'],
859 ie_result['entries'] = [
860 self.process_ie_result(_fixup(r), download, extra_info)
861 for r in ie_result['entries']
865 raise Exception('Invalid result type: %s' % result_type)
867 def _build_format_filter(self, filter_spec):
868 " Returns a function to filter the formats according to the filter_spec "
878 operator_rex = re.compile(r'''(?x)\s*
879 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
880 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
881 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
883 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
884 m = operator_rex.search(filter_spec)
887 comparison_value = int(m.group('value'))
889 comparison_value = parse_filesize(m.group('value'))
890 if comparison_value is None:
891 comparison_value = parse_filesize(m.group('value') + 'B')
892 if comparison_value is None:
894 'Invalid value %r in format specification %r' % (
895 m.group('value'), filter_spec))
896 op = OPERATORS[m.group('op')]
902 '^=': lambda attr, value: attr.startswith(value),
903 '$=': lambda attr, value: attr.endswith(value),
904 '*=': lambda attr, value: value in attr,
906 str_operator_rex = re.compile(r'''(?x)
907 \s*(?P<key>ext|acodec|vcodec|container|protocol)
908 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
909 \s*(?P<value>[a-zA-Z0-9._-]+)
911 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
912 m = str_operator_rex.search(filter_spec)
914 comparison_value = m.group('value')
915 op = STR_OPERATORS[m.group('op')]
918 raise ValueError('Invalid filter specification %r' % filter_spec)
921 actual_value = f.get(m.group('key'))
922 if actual_value is None:
923 return m.group('none_inclusive')
924 return op(actual_value, comparison_value)
927 def build_format_selector(self, format_spec):
928 def syntax_error(note, start):
930 'Invalid format specification: '
931 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
932 return SyntaxError(message)
934 PICKFIRST = 'PICKFIRST'
938 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
940 def _parse_filter(tokens):
942 for type, string, start, _, _ in tokens:
943 if type == tokenize.OP and string == ']':
944 return ''.join(filter_parts)
946 filter_parts.append(string)
948 def _remove_unused_ops(tokens):
949 # Remove operators that we don't use and join them with the surrounding strings
950 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
951 ALLOWED_OPS = ('/', '+', ',', '(', ')')
952 last_string, last_start, last_end, last_line = None, None, None, None
953 for type, string, start, end, line in tokens:
954 if type == tokenize.OP and string == '[':
956 yield tokenize.NAME, last_string, last_start, last_end, last_line
958 yield type, string, start, end, line
959 # everything inside brackets will be handled by _parse_filter
960 for type, string, start, end, line in tokens:
961 yield type, string, start, end, line
962 if type == tokenize.OP and string == ']':
964 elif type == tokenize.OP and string in ALLOWED_OPS:
966 yield tokenize.NAME, last_string, last_start, last_end, last_line
968 yield type, string, start, end, line
969 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
975 last_string += string
977 yield tokenize.NAME, last_string, last_start, last_end, last_line
979 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
981 current_selector = None
982 for type, string, start, _, _ in tokens:
983 # ENCODING is only defined in python 3.x
984 if type == getattr(tokenize, 'ENCODING', None):
986 elif type in [tokenize.NAME, tokenize.NUMBER]:
987 current_selector = FormatSelector(SINGLE, string, [])
988 elif type == tokenize.OP:
991 # ')' will be handled by the parentheses group
992 tokens.restore_last_token()
994 elif inside_merge and string in ['/', ',']:
995 tokens.restore_last_token()
997 elif inside_choice and string == ',':
998 tokens.restore_last_token()
1001 if not current_selector:
1002 raise syntax_error('"," must follow a format selector', start)
1003 selectors.append(current_selector)
1004 current_selector = None
1006 if not current_selector:
1007 raise syntax_error('"/" must follow a format selector', start)
1008 first_choice = current_selector
1009 second_choice = _parse_format_selection(tokens, inside_choice=True)
1010 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1012 if not current_selector:
1013 current_selector = FormatSelector(SINGLE, 'best', [])
1014 format_filter = _parse_filter(tokens)
1015 current_selector.filters.append(format_filter)
1017 if current_selector:
1018 raise syntax_error('Unexpected "("', start)
1019 group = _parse_format_selection(tokens, inside_group=True)
1020 current_selector = FormatSelector(GROUP, group, [])
1022 video_selector = current_selector
1023 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1024 if not video_selector or not audio_selector:
1025 raise syntax_error('"+" must be between two format selectors', start)
1026 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1028 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1029 elif type == tokenize.ENDMARKER:
1031 if current_selector:
1032 selectors.append(current_selector)
1035 def _build_selector_function(selector):
1036 if isinstance(selector, list):
1037 fs = [_build_selector_function(s) for s in selector]
1039 def selector_function(formats):
1041 for format in f(formats):
1043 return selector_function
1044 elif selector.type == GROUP:
1045 selector_function = _build_selector_function(selector.selector)
1046 elif selector.type == PICKFIRST:
1047 fs = [_build_selector_function(s) for s in selector.selector]
1049 def selector_function(formats):
1051 picked_formats = list(f(formats))
1053 return picked_formats
1055 elif selector.type == SINGLE:
1056 format_spec = selector.selector
1058 def selector_function(formats):
1059 formats = list(formats)
1062 if format_spec == 'all':
1065 elif format_spec in ['best', 'worst', None]:
1066 format_idx = 0 if format_spec == 'worst' else -1
1067 audiovideo_formats = [
1069 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1070 if audiovideo_formats:
1071 yield audiovideo_formats[format_idx]
1072 # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format
1073 elif (all(f.get('acodec') != 'none' for f in formats) or
1074 all(f.get('vcodec') != 'none' for f in formats)):
1075 yield formats[format_idx]
1076 elif format_spec == 'bestaudio':
1079 if f.get('vcodec') == 'none']
1081 yield audio_formats[-1]
1082 elif format_spec == 'worstaudio':
1085 if f.get('vcodec') == 'none']
1087 yield audio_formats[0]
1088 elif format_spec == 'bestvideo':
1091 if f.get('acodec') == 'none']
1093 yield video_formats[-1]
1094 elif format_spec == 'worstvideo':
1097 if f.get('acodec') == 'none']
1099 yield video_formats[0]
1101 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1102 if format_spec in extensions:
1103 filter_f = lambda f: f['ext'] == format_spec
1105 filter_f = lambda f: f['format_id'] == format_spec
1106 matches = list(filter(filter_f, formats))
1109 elif selector.type == MERGE:
1110 def _merge(formats_info):
1111 format_1, format_2 = [f['format_id'] for f in formats_info]
1112 # The first format must contain the video and the
1114 if formats_info[0].get('vcodec') == 'none':
1115 self.report_error('The first format must '
1116 'contain the video, try using '
1117 '"-f %s+%s"' % (format_2, format_1))
1119 # Formats must be opposite (video+audio)
1120 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1122 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1123 % (format_1, format_2))
1126 formats_info[0]['ext']
1127 if self.params.get('merge_output_format') is None
1128 else self.params['merge_output_format'])
1130 'requested_formats': formats_info,
1131 'format': '%s+%s' % (formats_info[0].get('format'),
1132 formats_info[1].get('format')),
1133 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1134 formats_info[1].get('format_id')),
1135 'width': formats_info[0].get('width'),
1136 'height': formats_info[0].get('height'),
1137 'resolution': formats_info[0].get('resolution'),
1138 'fps': formats_info[0].get('fps'),
1139 'vcodec': formats_info[0].get('vcodec'),
1140 'vbr': formats_info[0].get('vbr'),
1141 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1142 'acodec': formats_info[1].get('acodec'),
1143 'abr': formats_info[1].get('abr'),
1146 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1148 def selector_function(formats):
1149 formats = list(formats)
1150 for pair in itertools.product(video_selector(formats), audio_selector(formats)):
1153 filters = [self._build_format_filter(f) for f in selector.filters]
1155 def final_selector(formats):
1156 for _filter in filters:
1157 formats = list(filter(_filter, formats))
1158 return selector_function(formats)
1159 return final_selector
1161 stream = io.BytesIO(format_spec.encode('utf-8'))
1163 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1164 except tokenize.TokenError:
1165 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1167 class TokenIterator(object):
1168 def __init__(self, tokens):
1169 self.tokens = tokens
1176 if self.counter >= len(self.tokens):
1177 raise StopIteration()
1178 value = self.tokens[self.counter]
1184 def restore_last_token(self):
1187 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1188 return _build_selector_function(parsed_selector)
1190 def _calc_headers(self, info_dict):
1191 res = std_headers.copy()
1193 add_headers = info_dict.get('http_headers')
1195 res.update(add_headers)
1197 cookies = self._calc_cookies(info_dict)
1199 res['Cookie'] = cookies
1203 def _calc_cookies(self, info_dict):
1204 pr = sanitized_Request(info_dict['url'])
1205 self.cookiejar.add_cookie_header(pr)
1206 return pr.get_header('Cookie')
1208 def process_video_result(self, info_dict, download=True):
1209 assert info_dict.get('_type', 'video') == 'video'
1211 if 'id' not in info_dict:
1212 raise ExtractorError('Missing "id" field in extractor result')
1213 if 'title' not in info_dict:
1214 raise ExtractorError('Missing "title" field in extractor result')
1216 if 'playlist' not in info_dict:
1217 # It isn't part of a playlist
1218 info_dict['playlist'] = None
1219 info_dict['playlist_index'] = None
1221 thumbnails = info_dict.get('thumbnails')
1222 if thumbnails is None:
1223 thumbnail = info_dict.get('thumbnail')
1225 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1227 thumbnails.sort(key=lambda t: (
1228 t.get('preference'), t.get('width'), t.get('height'),
1229 t.get('id'), t.get('url')))
1230 for i, t in enumerate(thumbnails):
1231 if t.get('width') and t.get('height'):
1232 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1233 if t.get('id') is None:
1236 if thumbnails and 'thumbnail' not in info_dict:
1237 info_dict['thumbnail'] = thumbnails[-1]['url']
1239 if 'display_id' not in info_dict and 'id' in info_dict:
1240 info_dict['display_id'] = info_dict['id']
1242 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1243 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1244 # see http://bugs.python.org/issue1646728)
1246 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1247 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1248 except (ValueError, OverflowError, OSError):
1251 # Auto generate title fields corresponding to the *_number fields when missing
1252 # in order to always have clean titles. This is very common for TV series.
1253 for field in ('chapter', 'season', 'episode'):
1254 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1255 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1257 subtitles = info_dict.get('subtitles')
1259 for _, subtitle in subtitles.items():
1260 for subtitle_format in subtitle:
1261 if 'ext' not in subtitle_format:
1262 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1264 if self.params.get('listsubtitles', False):
1265 if 'automatic_captions' in info_dict:
1266 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1267 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1269 info_dict['requested_subtitles'] = self.process_subtitles(
1270 info_dict['id'], subtitles,
1271 info_dict.get('automatic_captions'))
1273 # We now pick which formats have to be downloaded
1274 if info_dict.get('formats') is None:
1275 # There's only one format available
1276 formats = [info_dict]
1278 formats = info_dict['formats']
1281 raise ExtractorError('No video formats found!')
1285 # We check that all the formats have the format and format_id fields
1286 for i, format in enumerate(formats):
1287 if 'url' not in format:
1288 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1290 if format.get('format_id') is None:
1291 format['format_id'] = compat_str(i)
1292 format_id = format['format_id']
1293 if format_id not in formats_dict:
1294 formats_dict[format_id] = []
1295 formats_dict[format_id].append(format)
1297 # Make sure all formats have unique format_id
1298 for format_id, ambiguous_formats in formats_dict.items():
1299 if len(ambiguous_formats) > 1:
1300 for i, format in enumerate(ambiguous_formats):
1301 format['format_id'] = '%s-%d' % (format_id, i)
1303 for i, format in enumerate(formats):
1304 if format.get('format') is None:
1305 format['format'] = '{id} - {res}{note}'.format(
1306 id=format['format_id'],
1307 res=self.format_resolution(format),
1308 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1310 # Automatically determine file extension if missing
1311 if 'ext' not in format:
1312 format['ext'] = determine_ext(format['url']).lower()
1313 # Automatically determine protocol if missing (useful for format
1314 # selection purposes)
1315 if 'protocol' not in format:
1316 format['protocol'] = determine_protocol(format)
1317 # Add HTTP headers, so that external programs can use them from the
1319 full_format_info = info_dict.copy()
1320 full_format_info.update(format)
1321 format['http_headers'] = self._calc_headers(full_format_info)
1323 # TODO Central sorting goes here
1325 if formats[0] is not info_dict:
1326 # only set the 'formats' fields if the original info_dict list them
1327 # otherwise we end up with a circular reference, the first (and unique)
1328 # element in the 'formats' field in info_dict is info_dict itself,
1329 # which can't be exported to json
1330 info_dict['formats'] = formats
1331 if self.params.get('listformats'):
1332 self.list_formats(info_dict)
1334 if self.params.get('list_thumbnails'):
1335 self.list_thumbnails(info_dict)
1338 req_format = self.params.get('format')
1339 if req_format is None:
1340 req_format_list = []
1341 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1342 info_dict['extractor'] in ['youtube', 'ted'] and
1343 not info_dict.get('is_live')):
1344 merger = FFmpegMergerPP(self)
1345 if merger.available and merger.can_merge():
1346 req_format_list.append('bestvideo+bestaudio')
1347 req_format_list.append('best')
1348 req_format = '/'.join(req_format_list)
1349 format_selector = self.build_format_selector(req_format)
1350 formats_to_download = list(format_selector(formats))
1351 if not formats_to_download:
1352 raise ExtractorError('requested format not available',
1356 if len(formats_to_download) > 1:
1357 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1358 for format in formats_to_download:
1359 new_info = dict(info_dict)
1360 new_info.update(format)
1361 self.process_info(new_info)
1362 # We update the info dict with the best quality format (backwards compatibility)
1363 info_dict.update(formats_to_download[-1])
1366 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1367 """Select the requested subtitles and their format"""
1369 if normal_subtitles and self.params.get('writesubtitles'):
1370 available_subs.update(normal_subtitles)
1371 if automatic_captions and self.params.get('writeautomaticsub'):
1372 for lang, cap_info in automatic_captions.items():
1373 if lang not in available_subs:
1374 available_subs[lang] = cap_info
1376 if (not self.params.get('writesubtitles') and not
1377 self.params.get('writeautomaticsub') or not
1381 if self.params.get('allsubtitles', False):
1382 requested_langs = available_subs.keys()
1384 if self.params.get('subtitleslangs', False):
1385 requested_langs = self.params.get('subtitleslangs')
1386 elif 'en' in available_subs:
1387 requested_langs = ['en']
1389 requested_langs = [list(available_subs.keys())[0]]
1391 formats_query = self.params.get('subtitlesformat', 'best')
1392 formats_preference = formats_query.split('/') if formats_query else []
1394 for lang in requested_langs:
1395 formats = available_subs.get(lang)
1397 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1399 for ext in formats_preference:
1403 matches = list(filter(lambda f: f['ext'] == ext, formats))
1409 self.report_warning(
1410 'No subtitle format found matching "%s" for language %s, '
1411 'using %s' % (formats_query, lang, f['ext']))
1415 def process_info(self, info_dict):
1416 """Process a single resolved IE result."""
1418 assert info_dict.get('_type', 'video') == 'video'
1420 max_downloads = self.params.get('max_downloads')
1421 if max_downloads is not None:
1422 if self._num_downloads >= int(max_downloads):
1423 raise MaxDownloadsReached()
1425 info_dict['fulltitle'] = info_dict['title']
1426 if len(info_dict['title']) > 200:
1427 info_dict['title'] = info_dict['title'][:197] + '...'
1429 if 'format' not in info_dict:
1430 info_dict['format'] = info_dict['ext']
1432 reason = self._match_entry(info_dict, incomplete=False)
1433 if reason is not None:
1434 self.to_screen('[download] ' + reason)
1437 self._num_downloads += 1
1439 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1442 if self.params.get('forcetitle', False):
1443 self.to_stdout(info_dict['fulltitle'])
1444 if self.params.get('forceid', False):
1445 self.to_stdout(info_dict['id'])
1446 if self.params.get('forceurl', False):
1447 if info_dict.get('requested_formats') is not None:
1448 for f in info_dict['requested_formats']:
1449 self.to_stdout(f['url'] + f.get('play_path', ''))
1451 # For RTMP URLs, also include the playpath
1452 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1453 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1454 self.to_stdout(info_dict['thumbnail'])
1455 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1456 self.to_stdout(info_dict['description'])
1457 if self.params.get('forcefilename', False) and filename is not None:
1458 self.to_stdout(filename)
1459 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1460 self.to_stdout(formatSeconds(info_dict['duration']))
1461 if self.params.get('forceformat', False):
1462 self.to_stdout(info_dict['format'])
1463 if self.params.get('forcejson', False):
1464 self.to_stdout(json.dumps(info_dict))
1466 # Do nothing else if in simulate mode
1467 if self.params.get('simulate', False):
1470 if filename is None:
1474 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1475 if dn and not os.path.exists(dn):
1477 except (OSError, IOError) as err:
1478 self.report_error('unable to create directory ' + error_to_compat_str(err))
1481 if self.params.get('writedescription', False):
1482 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1483 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1484 self.to_screen('[info] Video description is already present')
1485 elif info_dict.get('description') is None:
1486 self.report_warning('There\'s no description to write.')
1489 self.to_screen('[info] Writing video description to: ' + descfn)
1490 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1491 descfile.write(info_dict['description'])
1492 except (OSError, IOError):
1493 self.report_error('Cannot write description file ' + descfn)
1496 if self.params.get('writeannotations', False):
1497 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1498 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1499 self.to_screen('[info] Video annotations are already present')
1502 self.to_screen('[info] Writing video annotations to: ' + annofn)
1503 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1504 annofile.write(info_dict['annotations'])
1505 except (KeyError, TypeError):
1506 self.report_warning('There are no annotations to write.')
1507 except (OSError, IOError):
1508 self.report_error('Cannot write annotations file: ' + annofn)
1511 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1512 self.params.get('writeautomaticsub')])
1514 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1515 # subtitles download errors are already managed as troubles in relevant IE
1516 # that way it will silently go on when used with unsupporting IE
1517 subtitles = info_dict['requested_subtitles']
1518 ie = self.get_info_extractor(info_dict['extractor_key'])
1519 for sub_lang, sub_info in subtitles.items():
1520 sub_format = sub_info['ext']
1521 if sub_info.get('data') is not None:
1522 sub_data = sub_info['data']
1525 sub_data = ie._download_webpage(
1526 sub_info['url'], info_dict['id'], note=False)
1527 except ExtractorError as err:
1528 self.report_warning('Unable to download subtitle for "%s": %s' %
1529 (sub_lang, error_to_compat_str(err.cause)))
1532 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1533 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1534 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1536 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1537 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1538 subfile.write(sub_data)
1539 except (OSError, IOError):
1540 self.report_error('Cannot write subtitles file ' + sub_filename)
1543 if self.params.get('writeinfojson', False):
1544 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1545 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1546 self.to_screen('[info] Video description metadata is already present')
1548 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1550 write_json_file(self.filter_requested_info(info_dict), infofn)
1551 except (OSError, IOError):
1552 self.report_error('Cannot write metadata to JSON file ' + infofn)
1555 self._write_thumbnails(info_dict, filename)
1557 if not self.params.get('skip_download', False):
1560 fd = get_suitable_downloader(info, self.params)(self, self.params)
1561 for ph in self._progress_hooks:
1562 fd.add_progress_hook(ph)
1563 if self.params.get('verbose'):
1564 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1565 return fd.download(name, info)
1567 if info_dict.get('requested_formats') is not None:
1570 merger = FFmpegMergerPP(self)
1571 if not merger.available:
1573 self.report_warning('You have requested multiple '
1574 'formats but ffmpeg or avconv are not installed.'
1575 ' The formats won\'t be merged.')
1577 postprocessors = [merger]
1579 def compatible_formats(formats):
1580 video, audio = formats
1582 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1583 if video_ext and audio_ext:
1585 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1588 for exts in COMPATIBLE_EXTS:
1589 if video_ext in exts and audio_ext in exts:
1591 # TODO: Check acodec/vcodec
1594 filename_real_ext = os.path.splitext(filename)[1][1:]
1596 os.path.splitext(filename)[0]
1597 if filename_real_ext == info_dict['ext']
1599 requested_formats = info_dict['requested_formats']
1600 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1601 info_dict['ext'] = 'mkv'
1602 self.report_warning(
1603 'Requested formats are incompatible for merge and will be merged into mkv.')
1604 # Ensure filename always has a correct extension for successful merge
1605 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1606 if os.path.exists(encodeFilename(filename)):
1608 '[download] %s has already been downloaded and '
1609 'merged' % filename)
1611 for f in requested_formats:
1612 new_info = dict(info_dict)
1614 fname = self.prepare_filename(new_info)
1615 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1616 downloaded.append(fname)
1617 partial_success = dl(fname, new_info)
1618 success = success and partial_success
1619 info_dict['__postprocessors'] = postprocessors
1620 info_dict['__files_to_merge'] = downloaded
1622 # Just a single file
1623 success = dl(filename, info_dict)
1624 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1625 self.report_error('unable to download video data: %s' % str(err))
1627 except (OSError, IOError) as err:
1628 raise UnavailableVideoError(err)
1629 except (ContentTooShortError, ) as err:
1630 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1635 fixup_policy = self.params.get('fixup')
1636 if fixup_policy is None:
1637 fixup_policy = 'detect_or_warn'
1639 stretched_ratio = info_dict.get('stretched_ratio')
1640 if stretched_ratio is not None and stretched_ratio != 1:
1641 if fixup_policy == 'warn':
1642 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1643 info_dict['id'], stretched_ratio))
1644 elif fixup_policy == 'detect_or_warn':
1645 stretched_pp = FFmpegFixupStretchedPP(self)
1646 if stretched_pp.available:
1647 info_dict.setdefault('__postprocessors', [])
1648 info_dict['__postprocessors'].append(stretched_pp)
1650 self.report_warning(
1651 '%s: Non-uniform pixel ratio (%s). Install ffmpeg or avconv to fix this automatically.' % (
1652 info_dict['id'], stretched_ratio))
1654 assert fixup_policy in ('ignore', 'never')
1656 if info_dict.get('requested_formats') is None and info_dict.get('container') == 'm4a_dash':
1657 if fixup_policy == 'warn':
1658 self.report_warning('%s: writing DASH m4a. Only some players support this container.' % (
1660 elif fixup_policy == 'detect_or_warn':
1661 fixup_pp = FFmpegFixupM4aPP(self)
1662 if fixup_pp.available:
1663 info_dict.setdefault('__postprocessors', [])
1664 info_dict['__postprocessors'].append(fixup_pp)
1666 self.report_warning(
1667 '%s: writing DASH m4a. Only some players support this container. Install ffmpeg or avconv to fix this automatically.' % (
1670 assert fixup_policy in ('ignore', 'never')
1673 self.post_process(filename, info_dict)
1674 except (PostProcessingError) as err:
1675 self.report_error('postprocessing: %s' % str(err))
1677 self.record_download_archive(info_dict)
1679 def download(self, url_list):
1680 """Download a given list of URLs."""
1681 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1682 if (len(url_list) > 1 and
1683 '%' not in outtmpl and
1684 self.params.get('max_downloads') != 1):
1685 raise SameFileError(outtmpl)
1687 for url in url_list:
1689 # It also downloads the videos
1690 res = self.extract_info(
1691 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1692 except UnavailableVideoError:
1693 self.report_error('unable to download video')
1694 except MaxDownloadsReached:
1695 self.to_screen('[info] Maximum number of downloaded files reached.')
1698 if self.params.get('dump_single_json', False):
1699 self.to_stdout(json.dumps(res))
1701 return self._download_retcode
1703 def download_with_info_file(self, info_filename):
1704 with contextlib.closing(fileinput.FileInput(
1705 [info_filename], mode='r',
1706 openhook=fileinput.hook_encoded('utf-8'))) as f:
1707 # FileInput doesn't have a read method, we can't call json.load
1708 info = self.filter_requested_info(json.loads('\n'.join(f)))
1710 self.process_ie_result(info, download=True)
1711 except DownloadError:
1712 webpage_url = info.get('webpage_url')
1713 if webpage_url is not None:
1714 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1715 return self.download([webpage_url])
1718 return self._download_retcode
1721 def filter_requested_info(info_dict):
1723 (k, v) for k, v in info_dict.items()
1724 if k not in ['requested_formats', 'requested_subtitles'])
1726 def post_process(self, filename, ie_info):
1727 """Run all the postprocessors on the given file."""
1728 info = dict(ie_info)
1729 info['filepath'] = filename
1731 if ie_info.get('__postprocessors') is not None:
1732 pps_chain.extend(ie_info['__postprocessors'])
1733 pps_chain.extend(self._pps)
1734 for pp in pps_chain:
1735 files_to_delete = []
1737 files_to_delete, info = pp.run(info)
1738 except PostProcessingError as e:
1739 self.report_error(e.msg)
1740 if files_to_delete and not self.params.get('keepvideo', False):
1741 for old_filename in files_to_delete:
1742 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1744 os.remove(encodeFilename(old_filename))
1745 except (IOError, OSError):
1746 self.report_warning('Unable to remove downloaded original file')
1748 def _make_archive_id(self, info_dict):
1749 # Future-proof against any change in case
1750 # and backwards compatibility with prior versions
1751 extractor = info_dict.get('extractor_key')
1752 if extractor is None:
1753 if 'id' in info_dict:
1754 extractor = info_dict.get('ie_key') # key in a playlist
1755 if extractor is None:
1756 return None # Incomplete video information
1757 return extractor.lower() + ' ' + info_dict['id']
1759 def in_download_archive(self, info_dict):
1760 fn = self.params.get('download_archive')
1764 vid_id = self._make_archive_id(info_dict)
1766 return False # Incomplete video information
1769 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1770 for line in archive_file:
1771 if line.strip() == vid_id:
1773 except IOError as ioe:
1774 if ioe.errno != errno.ENOENT:
1778 def record_download_archive(self, info_dict):
1779 fn = self.params.get('download_archive')
1782 vid_id = self._make_archive_id(info_dict)
1784 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1785 archive_file.write(vid_id + '\n')
1788 def format_resolution(format, default='unknown'):
1789 if format.get('vcodec') == 'none':
1791 if format.get('resolution') is not None:
1792 return format['resolution']
1793 if format.get('height') is not None:
1794 if format.get('width') is not None:
1795 res = '%sx%s' % (format['width'], format['height'])
1797 res = '%sp' % format['height']
1798 elif format.get('width') is not None:
1799 res = '?x%d' % format['width']
1804 def _format_note(self, fdict):
1806 if fdict.get('ext') in ['f4f', 'f4m']:
1807 res += '(unsupported) '
1808 if fdict.get('language'):
1811 res += '[%s]' % fdict['language']
1812 if fdict.get('format_note') is not None:
1813 res += fdict['format_note'] + ' '
1814 if fdict.get('tbr') is not None:
1815 res += '%4dk ' % fdict['tbr']
1816 if fdict.get('container') is not None:
1819 res += '%s container' % fdict['container']
1820 if (fdict.get('vcodec') is not None and
1821 fdict.get('vcodec') != 'none'):
1824 res += fdict['vcodec']
1825 if fdict.get('vbr') is not None:
1827 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1829 if fdict.get('vbr') is not None:
1830 res += '%4dk' % fdict['vbr']
1831 if fdict.get('fps') is not None:
1832 res += ', %sfps' % fdict['fps']
1833 if fdict.get('acodec') is not None:
1836 if fdict['acodec'] == 'none':
1839 res += '%-5s' % fdict['acodec']
1840 elif fdict.get('abr') is not None:
1844 if fdict.get('abr') is not None:
1845 res += '@%3dk' % fdict['abr']
1846 if fdict.get('asr') is not None:
1847 res += ' (%5dHz)' % fdict['asr']
1848 if fdict.get('filesize') is not None:
1851 res += format_bytes(fdict['filesize'])
1852 elif fdict.get('filesize_approx') is not None:
1855 res += '~' + format_bytes(fdict['filesize_approx'])
1858 def list_formats(self, info_dict):
1859 formats = info_dict.get('formats', [info_dict])
1861 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1863 if f.get('preference') is None or f['preference'] >= -1000]
1864 if len(formats) > 1:
1865 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1867 header_line = ['format code', 'extension', 'resolution', 'note']
1869 '[info] Available formats for %s:\n%s' %
1870 (info_dict['id'], render_table(header_line, table)))
1872 def list_thumbnails(self, info_dict):
1873 thumbnails = info_dict.get('thumbnails')
1875 tn_url = info_dict.get('thumbnail')
1877 thumbnails = [{'id': '0', 'url': tn_url}]
1880 '[info] No thumbnails present for %s' % info_dict['id'])
1884 '[info] Thumbnails for %s:' % info_dict['id'])
1885 self.to_screen(render_table(
1886 ['ID', 'width', 'height', 'URL'],
1887 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1889 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1891 self.to_screen('%s has no %s' % (video_id, name))
1894 'Available %s for %s:' % (name, video_id))
1895 self.to_screen(render_table(
1896 ['Language', 'formats'],
1897 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1898 for lang, formats in subtitles.items()]))
1900 def urlopen(self, req):
1901 """ Start an HTTP download """
1902 if isinstance(req, compat_basestring):
1903 req = sanitized_Request(req)
1904 return self._opener.open(req, timeout=self._socket_timeout)
1906 def print_debug_header(self):
1907 if not self.params.get('verbose'):
1910 if type('') is not compat_str:
1911 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1912 self.report_warning(
1913 'Your Python is broken! Update to a newer and supported version')
1915 stdout_encoding = getattr(
1916 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1918 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1919 locale.getpreferredencoding(),
1920 sys.getfilesystemencoding(),
1922 self.get_encoding()))
1923 write_string(encoding_str, encoding=None)
1925 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1927 sp = subprocess.Popen(
1928 ['git', 'rev-parse', '--short', 'HEAD'],
1929 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1930 cwd=os.path.dirname(os.path.abspath(__file__)))
1931 out, err = sp.communicate()
1932 out = out.decode().strip()
1933 if re.match('[0-9a-f]+', out):
1934 self._write_string('[debug] Git HEAD: ' + out + '\n')
1940 self._write_string('[debug] Python version %s - %s\n' % (
1941 platform.python_version(), platform_name()))
1943 exe_versions = FFmpegPostProcessor.get_versions(self)
1944 exe_versions['rtmpdump'] = rtmpdump_version()
1945 exe_str = ', '.join(
1947 for exe, v in sorted(exe_versions.items())
1952 self._write_string('[debug] exe versions: %s\n' % exe_str)
1955 for handler in self._opener.handlers:
1956 if hasattr(handler, 'proxies'):
1957 proxy_map.update(handler.proxies)
1958 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1960 if self.params.get('call_home', False):
1961 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
1962 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
1963 latest_version = self.urlopen(
1964 'https://yt-dl.org/latest/version').read().decode('utf-8')
1965 if version_tuple(latest_version) > version_tuple(__version__):
1966 self.report_warning(
1967 'You are using an outdated version (newest version: %s)! '
1968 'See https://yt-dl.org/update if you need help updating.' %
1971 def _setup_opener(self):
1972 timeout_val = self.params.get('socket_timeout')
1973 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
1975 opts_cookiefile = self.params.get('cookiefile')
1976 opts_proxy = self.params.get('proxy')
1978 if opts_cookiefile is None:
1979 self.cookiejar = compat_cookiejar.CookieJar()
1981 self.cookiejar = compat_cookiejar.MozillaCookieJar(
1983 if os.access(opts_cookiefile, os.R_OK):
1984 self.cookiejar.load()
1986 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
1987 if opts_proxy is not None:
1988 if opts_proxy == '':
1991 proxies = {'http': opts_proxy, 'https': opts_proxy}
1993 proxies = compat_urllib_request.getproxies()
1994 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
1995 if 'http' in proxies and 'https' not in proxies:
1996 proxies['https'] = proxies['http']
1997 proxy_handler = PerRequestProxyHandler(proxies)
1999 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2000 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2001 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2002 data_handler = compat_urllib_request_DataHandler()
2004 # When passing our own FileHandler instance, build_opener won't add the
2005 # default FileHandler and allows us to disable the file protocol, which
2006 # can be used for malicious purposes (see
2007 # https://github.com/rg3/youtube-dl/issues/8227)
2008 file_handler = compat_urllib_request.FileHandler()
2010 def file_open(*args, **kwargs):
2011 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2012 file_handler.file_open = file_open
2014 opener = compat_urllib_request.build_opener(
2015 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2017 # Delete the default user-agent header, which would otherwise apply in
2018 # cases where our custom HTTP handler doesn't come into play
2019 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2020 opener.addheaders = []
2021 self._opener = opener
2023 def encode(self, s):
2024 if isinstance(s, bytes):
2025 return s # Already encoded
2028 return s.encode(self.get_encoding())
2029 except UnicodeEncodeError as err:
2030 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2033 def get_encoding(self):
2034 encoding = self.params.get('encoding')
2035 if encoding is None:
2036 encoding = preferredencoding()
2039 def _write_thumbnails(self, info_dict, filename):
2040 if self.params.get('writethumbnail', False):
2041 thumbnails = info_dict.get('thumbnails')
2043 thumbnails = [thumbnails[-1]]
2044 elif self.params.get('write_all_thumbnails', False):
2045 thumbnails = info_dict.get('thumbnails')
2050 # No thumbnails present, so return immediately
2053 for t in thumbnails:
2054 thumb_ext = determine_ext(t['url'], 'jpg')
2055 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2056 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2057 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2059 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2060 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2061 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2063 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2064 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2066 uf = self.urlopen(t['url'])
2067 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2068 shutil.copyfileobj(uf, thumbf)
2069 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2070 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2071 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2072 self.report_warning('Unable to download thumbnail "%s": %s' %
2073 (t['url'], error_to_compat_str(err)))