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,
60 PerRequestProxyHandler,
70 UnavailableVideoError,
81 from .cache import Cache
82 from .extractor import get_info_extractor, gen_extractors
83 from .downloader import get_suitable_downloader
84 from .downloader.rtmp import rtmpdump_version
85 from .postprocessor import (
87 FFmpegFixupStretchedPP,
92 from .version import __version__
95 class YoutubeDL(object):
98 YoutubeDL objects are the ones responsible of downloading the
99 actual video file and writing it to disk if the user has requested
100 it, among some other tasks. In most cases there should be one per
101 program. As, given a video URL, the downloader doesn't know how to
102 extract all the needed information, task that InfoExtractors do, it
103 has to pass the URL to one of them.
105 For this, YoutubeDL objects have a method that allows
106 InfoExtractors to be registered in a given order. When it is passed
107 a URL, the YoutubeDL object handles it to the first InfoExtractor it
108 finds that reports being able to handle it. The InfoExtractor extracts
109 all the information about the video or videos the URL refers to, and
110 YoutubeDL process the extracted information, possibly using a File
111 Downloader to download the video.
113 YoutubeDL objects accept a lot of parameters. In order not to saturate
114 the object constructor with arguments, it receives a dictionary of
115 options instead. These options are available through the params
116 attribute for the InfoExtractors to use. The YoutubeDL also
117 registers itself as the downloader in charge for the InfoExtractors
118 that are added to it, so this is a "mutual registration".
122 username: Username for authentication purposes.
123 password: Password for authentication purposes.
124 videopassword: Password for accessing a video.
125 usenetrc: Use netrc for authentication instead.
126 verbose: Print additional info to stdout.
127 quiet: Do not print messages to stdout.
128 no_warnings: Do not print out anything for warnings.
129 forceurl: Force printing final URL.
130 forcetitle: Force printing title.
131 forceid: Force printing ID.
132 forcethumbnail: Force printing thumbnail URL.
133 forcedescription: Force printing description.
134 forcefilename: Force printing final filename.
135 forceduration: Force printing duration.
136 forcejson: Force printing info_dict as JSON.
137 dump_single_json: Force printing the info_dict of the whole playlist
138 (or video) as a single JSON line.
139 simulate: Do not download the video files.
140 format: Video format code. See options.py for more information.
141 outtmpl: Template for output names.
142 restrictfilenames: Do not allow "&" and spaces in file names
143 ignoreerrors: Do not stop on download errors.
144 force_generic_extractor: Force downloader to use the generic extractor
145 nooverwrites: Prevent overwriting files.
146 playliststart: Playlist item to start at.
147 playlistend: Playlist item to end at.
148 playlist_items: Specific indices of playlist to download.
149 playlistreverse: Download playlist items in reverse order.
150 matchtitle: Download only matching titles.
151 rejecttitle: Reject downloads for matching titles.
152 logger: Log messages to a logging.Logger instance.
153 logtostderr: Log messages to stderr instead of stdout.
154 writedescription: Write the video description to a .description file
155 writeinfojson: Write the video description to a .info.json file
156 writeannotations: Write the video annotations to a .annotations.xml file
157 writethumbnail: Write the thumbnail image to a file
158 write_all_thumbnails: Write all thumbnail formats to files
159 writesubtitles: Write the video subtitles to a file
160 writeautomaticsub: Write the automatic subtitles to a file
161 allsubtitles: Downloads all the subtitles of the video
162 (requires writesubtitles or writeautomaticsub)
163 listsubtitles: Lists all available subtitles for the video
164 subtitlesformat: The format code for subtitles
165 subtitleslangs: List of languages of the subtitles to download
166 keepvideo: Keep the video file after post-processing
167 daterange: A DateRange object, download only if the upload_date is in the range.
168 skip_download: Skip the actual download of the video file
169 cachedir: Location of the cache files in the filesystem.
170 False to disable filesystem cache.
171 noplaylist: Download single video instead of a playlist if in doubt.
172 age_limit: An integer representing the user's age in years.
173 Unsuitable videos for the given age are skipped.
174 min_views: An integer representing the minimum view count the video
175 must have in order to not be skipped.
176 Videos without view count information are always
177 downloaded. None for no limit.
178 max_views: An integer representing the maximum view count.
179 Videos that are more popular than that are not
181 Videos without view count information are always
182 downloaded. None for no limit.
183 download_archive: File name of a file where all downloads are recorded.
184 Videos already present in the file are not downloaded
186 cookiefile: File name where cookies should be read from and dumped to.
187 nocheckcertificate:Do not verify SSL certificates
188 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
189 At the moment, this is only supported by YouTube.
190 proxy: URL of the proxy server to use
191 cn_verification_proxy: URL of the proxy to use for IP address verification
192 on Chinese sites. (Experimental)
193 socket_timeout: Time to wait for unresponsive hosts, in seconds
194 bidi_workaround: Work around buggy terminals without bidirectional text
195 support, using fridibi
196 debug_printtraffic:Print out sent and received HTTP traffic
197 include_ads: Download ads as well
198 default_search: Prepend this string if an input url is not valid.
199 'auto' for elaborate guessing
200 encoding: Use this encoding instead of the system-specified.
201 extract_flat: Do not resolve URLs, return the immediate result.
202 Pass in 'in_playlist' to only show this behavior for
204 postprocessors: A list of dictionaries, each with an entry
205 * key: The name of the postprocessor. See
206 youtube_dl/postprocessor/__init__.py for a list.
207 as well as any further keyword arguments for the
209 progress_hooks: A list of functions that get called on download
210 progress, with a dictionary with the entries
211 * status: One of "downloading", "error", or "finished".
212 Check this first and ignore unknown values.
214 If status is one of "downloading", or "finished", the
215 following properties may also be present:
216 * filename: The final filename (always present)
217 * tmpfilename: The filename we're currently writing to
218 * downloaded_bytes: Bytes on disk
219 * total_bytes: Size of the whole file, None if unknown
220 * total_bytes_estimate: Guess of the eventual file size,
222 * elapsed: The number of seconds since download started.
223 * eta: The estimated time in seconds, None if unknown
224 * speed: The download speed in bytes/second, None if
226 * fragment_index: The counter of the currently
227 downloaded video fragment.
228 * fragment_count: The number of fragments (= individual
229 files that will be merged)
231 Progress hooks are guaranteed to be called at least once
232 (with status "finished") if the download is successful.
233 merge_output_format: Extension to use when merging formats.
234 fixup: Automatically correct known faults of the file.
236 - "never": do nothing
237 - "warn": only emit a warning
238 - "detect_or_warn": check whether we can do anything
239 about it, warn otherwise (default)
240 source_address: (Experimental) Client-side IP address to bind to.
241 call_home: Boolean, true iff we are allowed to contact the
242 youtube-dl servers for debugging.
243 sleep_interval: Number of seconds to sleep before each download.
244 listformats: Print an overview of available video formats and exit.
245 list_thumbnails: Print a table of all thumbnails and exit.
246 match_filter: A function that gets called with the info_dict of
248 If it returns a message, the video is ignored.
249 If it returns None, the video is downloaded.
250 match_filter_func in utils.py is one example for this.
251 no_color: Do not emit color codes in output.
253 The following options determine which downloader is picked:
254 external_downloader: Executable of the external downloader to call.
255 None or unset for standard (built-in) downloader.
256 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv.
258 The following parameters are not used by YoutubeDL itself, they are used by
259 the downloader (see youtube_dl/downloader/common.py):
260 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
261 noresizebuffer, retries, continuedl, noprogress, consoletitle,
262 xattr_set_filesize, external_downloader_args.
264 The following options are used by the post processors:
265 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
266 otherwise prefer avconv.
272 _download_retcode = None
273 _num_downloads = None
276 def __init__(self, params=None, auto_init=True):
277 """Create a FileDownloader object with the given options."""
281 self._ies_instances = {}
283 self._progress_hooks = []
284 self._download_retcode = 0
285 self._num_downloads = 0
286 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
287 self._err_file = sys.stderr
289 self.cache = Cache(self)
291 if params.get('bidi_workaround', False):
294 master, slave = pty.openpty()
295 width = compat_get_terminal_size().columns
299 width_args = ['-w', str(width)]
301 stdin=subprocess.PIPE,
303 stderr=self._err_file)
305 self._output_process = subprocess.Popen(
306 ['bidiv'] + width_args, **sp_kwargs
309 self._output_process = subprocess.Popen(
310 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
311 self._output_channel = os.fdopen(master, 'rb')
312 except OSError as ose:
314 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 $PATH.')
318 if (sys.version_info >= (3,) and sys.platform != 'win32' and
319 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
320 not params.get('restrictfilenames', False)):
321 # On Python 3, the Unicode filesystem API will throw errors (#1474)
323 'Assuming --restrict-filenames since file system encoding '
324 'cannot encode all characters. '
325 'Set the LC_ALL environment variable to fix this.')
326 self.params['restrictfilenames'] = True
328 if isinstance(params.get('outtmpl'), bytes):
330 'Parameter outtmpl is bytes, but should be a unicode string. '
331 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
336 self.print_debug_header()
337 self.add_default_info_extractors()
339 for pp_def_raw in self.params.get('postprocessors', []):
340 pp_class = get_postprocessor(pp_def_raw['key'])
341 pp_def = dict(pp_def_raw)
343 pp = pp_class(self, **compat_kwargs(pp_def))
344 self.add_post_processor(pp)
346 for ph in self.params.get('progress_hooks', []):
347 self.add_progress_hook(ph)
349 def warn_if_short_id(self, argv):
350 # short YouTube ID starting with dash?
352 i for i, a in enumerate(argv)
353 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
357 [a for i, a in enumerate(argv) if i not in idxs] +
358 ['--'] + [argv[i] for i in idxs]
361 'Long argument string detected. '
362 'Use -- to separate parameters and URLs, like this:\n%s\n' %
363 args_to_str(correct_argv))
365 def add_info_extractor(self, ie):
366 """Add an InfoExtractor object to the end of the list."""
368 self._ies_instances[ie.ie_key()] = ie
369 ie.set_downloader(self)
371 def get_info_extractor(self, ie_key):
373 Get an instance of an IE with name ie_key, it will try to get one from
374 the _ies list, if there's no instance it will create a new one and add
375 it to the extractor list.
377 ie = self._ies_instances.get(ie_key)
379 ie = get_info_extractor(ie_key)()
380 self.add_info_extractor(ie)
383 def add_default_info_extractors(self):
385 Add the InfoExtractors returned by gen_extractors to the end of the list
387 for ie in gen_extractors():
388 self.add_info_extractor(ie)
390 def add_post_processor(self, pp):
391 """Add a PostProcessor object to the end of the chain."""
393 pp.set_downloader(self)
395 def add_progress_hook(self, ph):
396 """Add the progress hook (currently only for the file downloader)"""
397 self._progress_hooks.append(ph)
399 def _bidi_workaround(self, message):
400 if not hasattr(self, '_output_channel'):
403 assert hasattr(self, '_output_process')
404 assert isinstance(message, compat_str)
405 line_count = message.count('\n') + 1
406 self._output_process.stdin.write((message + '\n').encode('utf-8'))
407 self._output_process.stdin.flush()
408 res = ''.join(self._output_channel.readline().decode('utf-8')
409 for _ in range(line_count))
410 return res[:-len('\n')]
412 def to_screen(self, message, skip_eol=False):
413 """Print message to stdout if not in quiet mode."""
414 return self.to_stdout(message, skip_eol, check_quiet=True)
416 def _write_string(self, s, out=None):
417 write_string(s, out=out, encoding=self.params.get('encoding'))
419 def to_stdout(self, message, skip_eol=False, check_quiet=False):
420 """Print message to stdout if not in quiet mode."""
421 if self.params.get('logger'):
422 self.params['logger'].debug(message)
423 elif not check_quiet or not self.params.get('quiet', False):
424 message = self._bidi_workaround(message)
425 terminator = ['\n', ''][skip_eol]
426 output = message + terminator
428 self._write_string(output, self._screen_file)
430 def to_stderr(self, message):
431 """Print message to stderr."""
432 assert isinstance(message, compat_str)
433 if self.params.get('logger'):
434 self.params['logger'].error(message)
436 message = self._bidi_workaround(message)
437 output = message + '\n'
438 self._write_string(output, self._err_file)
440 def to_console_title(self, message):
441 if not self.params.get('consoletitle', False):
443 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
444 # c_wchar_p() might not be necessary if `message` is
445 # already of type unicode()
446 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
447 elif 'TERM' in os.environ:
448 self._write_string('\033]0;%s\007' % message, self._screen_file)
450 def save_console_title(self):
451 if not self.params.get('consoletitle', False):
453 if 'TERM' in os.environ:
454 # Save the title on stack
455 self._write_string('\033[22;0t', self._screen_file)
457 def restore_console_title(self):
458 if not self.params.get('consoletitle', False):
460 if 'TERM' in os.environ:
461 # Restore the title from stack
462 self._write_string('\033[23;0t', self._screen_file)
465 self.save_console_title()
468 def __exit__(self, *args):
469 self.restore_console_title()
471 if self.params.get('cookiefile') is not None:
472 self.cookiejar.save()
474 def trouble(self, message=None, tb=None):
475 """Determine action to take when a download problem appears.
477 Depending on if the downloader has been configured to ignore
478 download errors or not, this method may throw an exception or
479 not when errors are found, after printing the message.
481 tb, if given, is additional traceback information.
483 if message is not None:
484 self.to_stderr(message)
485 if self.params.get('verbose'):
487 if sys.exc_info()[0]: # if .trouble has been called from an except block
489 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
490 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
491 tb += compat_str(traceback.format_exc())
493 tb_data = traceback.format_list(traceback.extract_stack())
494 tb = ''.join(tb_data)
496 if not self.params.get('ignoreerrors', False):
497 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
498 exc_info = sys.exc_info()[1].exc_info
500 exc_info = sys.exc_info()
501 raise DownloadError(message, exc_info)
502 self._download_retcode = 1
504 def report_warning(self, message):
506 Print the message to stderr, it will be prefixed with 'WARNING:'
507 If stderr is a tty file the 'WARNING:' will be colored
509 if self.params.get('logger') is not None:
510 self.params['logger'].warning(message)
512 if self.params.get('no_warnings'):
514 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
515 _msg_header = '\033[0;33mWARNING:\033[0m'
517 _msg_header = 'WARNING:'
518 warning_message = '%s %s' % (_msg_header, message)
519 self.to_stderr(warning_message)
521 def report_error(self, message, tb=None):
523 Do the same as trouble, but prefixes the message with 'ERROR:', colored
524 in red if stderr is a tty file.
526 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
527 _msg_header = '\033[0;31mERROR:\033[0m'
529 _msg_header = 'ERROR:'
530 error_message = '%s %s' % (_msg_header, message)
531 self.trouble(error_message, tb)
533 def report_file_already_downloaded(self, file_name):
534 """Report file has already been fully downloaded."""
536 self.to_screen('[download] %s has already been downloaded' % file_name)
537 except UnicodeEncodeError:
538 self.to_screen('[download] The file has already been downloaded')
540 def prepare_filename(self, info_dict):
541 """Generate the output filename."""
543 template_dict = dict(info_dict)
545 template_dict['epoch'] = int(time.time())
546 autonumber_size = self.params.get('autonumber_size')
547 if autonumber_size is None:
549 autonumber_templ = '%0' + str(autonumber_size) + 'd'
550 template_dict['autonumber'] = autonumber_templ % self._num_downloads
551 if template_dict.get('playlist_index') is not None:
552 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
553 if template_dict.get('resolution') is None:
554 if template_dict.get('width') and template_dict.get('height'):
555 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
556 elif template_dict.get('height'):
557 template_dict['resolution'] = '%sp' % template_dict['height']
558 elif template_dict.get('width'):
559 template_dict['resolution'] = '?x%d' % template_dict['width']
561 sanitize = lambda k, v: sanitize_filename(
563 restricted=self.params.get('restrictfilenames'),
565 template_dict = dict((k, sanitize(k, v))
566 for k, v in template_dict.items()
568 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
570 outtmpl = sanitize_path(self.params.get('outtmpl', DEFAULT_OUTTMPL))
571 tmpl = compat_expanduser(outtmpl)
572 filename = tmpl % template_dict
573 # Temporary fix for #4787
574 # 'Treat' all problem characters by passing filename through preferredencoding
575 # to workaround encoding issues with subprocess on python2 @ Windows
576 if sys.version_info < (3, 0) and sys.platform == 'win32':
577 filename = encodeFilename(filename, True).decode(preferredencoding())
579 except ValueError as err:
580 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
583 def _match_entry(self, info_dict, incomplete):
584 """ Returns None iff the file should be downloaded """
586 video_title = info_dict.get('title', info_dict.get('id', 'video'))
587 if 'title' in info_dict:
588 # This can happen when we're just evaluating the playlist
589 title = info_dict['title']
590 matchtitle = self.params.get('matchtitle', False)
592 if not re.search(matchtitle, title, re.IGNORECASE):
593 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
594 rejecttitle = self.params.get('rejecttitle', False)
596 if re.search(rejecttitle, title, re.IGNORECASE):
597 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
598 date = info_dict.get('upload_date', None)
600 dateRange = self.params.get('daterange', DateRange())
601 if date not in dateRange:
602 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
603 view_count = info_dict.get('view_count', None)
604 if view_count is not None:
605 min_views = self.params.get('min_views')
606 if min_views is not None and view_count < min_views:
607 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
608 max_views = self.params.get('max_views')
609 if max_views is not None and view_count > max_views:
610 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
611 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
612 return 'Skipping "%s" because it is age restricted' % video_title
613 if self.in_download_archive(info_dict):
614 return '%s has already been recorded in archive' % video_title
617 match_filter = self.params.get('match_filter')
618 if match_filter is not None:
619 ret = match_filter(info_dict)
626 def add_extra_info(info_dict, extra_info):
627 '''Set the keys from extra_info in info dict if they are missing'''
628 for key, value in extra_info.items():
629 info_dict.setdefault(key, value)
631 def extract_info(self, url, download=True, ie_key=None, extra_info={},
632 process=True, force_generic_extractor=False):
634 Returns a list with a dictionary for each video we find.
635 If 'download', also downloads the videos.
636 extra_info is a dict containing the extra values to add to each result
639 if not ie_key and force_generic_extractor:
643 ies = [self.get_info_extractor(ie_key)]
648 if not ie.suitable(url):
652 self.report_warning('The program functionality for this site has been marked as broken, '
653 'and will probably not work.')
656 ie_result = ie.extract(url)
657 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
659 if isinstance(ie_result, list):
660 # Backwards compatibility: old IE result format
662 '_type': 'compat_list',
663 'entries': ie_result,
665 self.add_default_extra_info(ie_result, ie, url)
667 return self.process_ie_result(ie_result, download, extra_info)
670 except ExtractorError as de: # An error we somewhat expected
671 self.report_error(compat_str(de), de.format_traceback())
673 except MaxDownloadsReached:
675 except Exception as e:
676 if self.params.get('ignoreerrors', False):
677 self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
682 self.report_error('no suitable InfoExtractor for URL %s' % url)
684 def add_default_extra_info(self, ie_result, ie, url):
685 self.add_extra_info(ie_result, {
686 'extractor': ie.IE_NAME,
688 'webpage_url_basename': url_basename(url),
689 'extractor_key': ie.ie_key(),
692 def process_ie_result(self, ie_result, download=True, extra_info={}):
694 Take the result of the ie(may be modified) and resolve all unresolved
695 references (URLs, playlist items).
697 It will also download the videos if 'download'.
698 Returns the resolved ie_result.
701 result_type = ie_result.get('_type', 'video')
703 if result_type in ('url', 'url_transparent'):
704 extract_flat = self.params.get('extract_flat', False)
705 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
706 extract_flat is True):
707 if self.params.get('forcejson', False):
708 self.to_stdout(json.dumps(ie_result))
711 if result_type == 'video':
712 self.add_extra_info(ie_result, extra_info)
713 return self.process_video_result(ie_result, download=download)
714 elif result_type == 'url':
715 # We have to add extra_info to the results because it may be
716 # contained in a playlist
717 return self.extract_info(ie_result['url'],
719 ie_key=ie_result.get('ie_key'),
720 extra_info=extra_info)
721 elif result_type == 'url_transparent':
722 # Use the information from the embedding page
723 info = self.extract_info(
724 ie_result['url'], ie_key=ie_result.get('ie_key'),
725 extra_info=extra_info, download=False, process=False)
727 force_properties = dict(
728 (k, v) for k, v in ie_result.items() if v is not None)
729 for f in ('_type', 'url'):
730 if f in force_properties:
731 del force_properties[f]
732 new_result = info.copy()
733 new_result.update(force_properties)
735 assert new_result.get('_type') != 'url_transparent'
737 return self.process_ie_result(
738 new_result, download=download, extra_info=extra_info)
739 elif result_type == 'playlist' or result_type == 'multi_video':
740 # We process each entry in the playlist
741 playlist = ie_result.get('title', None) or ie_result.get('id', None)
742 self.to_screen('[download] Downloading playlist: %s' % playlist)
744 playlist_results = []
746 playliststart = self.params.get('playliststart', 1) - 1
747 playlistend = self.params.get('playlistend', None)
748 # For backwards compatibility, interpret -1 as whole list
749 if playlistend == -1:
752 playlistitems_str = self.params.get('playlist_items', None)
754 if playlistitems_str is not None:
755 def iter_playlistitems(format):
756 for string_segment in format.split(','):
757 if '-' in string_segment:
758 start, end = string_segment.split('-')
759 for item in range(int(start), int(end) + 1):
762 yield int(string_segment)
763 playlistitems = iter_playlistitems(playlistitems_str)
765 ie_entries = ie_result['entries']
766 if isinstance(ie_entries, list):
767 n_all_entries = len(ie_entries)
770 ie_entries[i - 1] for i in playlistitems
771 if -n_all_entries <= i - 1 < n_all_entries]
773 entries = ie_entries[playliststart:playlistend]
774 n_entries = len(entries)
776 "[%s] playlist %s: Collected %d video ids (downloading %d of them)" %
777 (ie_result['extractor'], playlist, n_all_entries, n_entries))
778 elif isinstance(ie_entries, PagedList):
781 for item in playlistitems:
782 entries.extend(ie_entries.getslice(
786 entries = ie_entries.getslice(
787 playliststart, playlistend)
788 n_entries = len(entries)
790 "[%s] playlist %s: Downloading %d videos" %
791 (ie_result['extractor'], playlist, n_entries))
794 entry_list = list(ie_entries)
795 entries = [entry_list[i - 1] for i in playlistitems]
797 entries = list(itertools.islice(
798 ie_entries, playliststart, playlistend))
799 n_entries = len(entries)
801 "[%s] playlist %s: Downloading %d videos" %
802 (ie_result['extractor'], playlist, n_entries))
804 if self.params.get('playlistreverse', False):
805 entries = entries[::-1]
807 for i, entry in enumerate(entries, 1):
808 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
810 'n_entries': n_entries,
811 'playlist': playlist,
812 'playlist_id': ie_result.get('id'),
813 'playlist_title': ie_result.get('title'),
814 'playlist_index': i + playliststart,
815 'extractor': ie_result['extractor'],
816 'webpage_url': ie_result['webpage_url'],
817 'webpage_url_basename': url_basename(ie_result['webpage_url']),
818 'extractor_key': ie_result['extractor_key'],
821 reason = self._match_entry(entry, incomplete=True)
822 if reason is not None:
823 self.to_screen('[download] ' + reason)
826 entry_result = self.process_ie_result(entry,
829 playlist_results.append(entry_result)
830 ie_result['entries'] = playlist_results
832 elif result_type == 'compat_list':
834 'Extractor %s returned a compat_list result. '
835 'It needs to be updated.' % ie_result.get('extractor'))
841 'extractor': ie_result['extractor'],
842 'webpage_url': ie_result['webpage_url'],
843 'webpage_url_basename': url_basename(ie_result['webpage_url']),
844 'extractor_key': ie_result['extractor_key'],
848 ie_result['entries'] = [
849 self.process_ie_result(_fixup(r), download, extra_info)
850 for r in ie_result['entries']
854 raise Exception('Invalid result type: %s' % result_type)
856 def _build_format_filter(self, filter_spec):
857 " Returns a function to filter the formats according to the filter_spec "
867 operator_rex = re.compile(r'''(?x)\s*
868 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
869 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
870 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
872 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
873 m = operator_rex.search(filter_spec)
876 comparison_value = int(m.group('value'))
878 comparison_value = parse_filesize(m.group('value'))
879 if comparison_value is None:
880 comparison_value = parse_filesize(m.group('value') + 'B')
881 if comparison_value is None:
883 'Invalid value %r in format specification %r' % (
884 m.group('value'), filter_spec))
885 op = OPERATORS[m.group('op')]
892 str_operator_rex = re.compile(r'''(?x)
893 \s*(?P<key>ext|acodec|vcodec|container|protocol)
894 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
895 \s*(?P<value>[a-zA-Z0-9_-]+)
897 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
898 m = str_operator_rex.search(filter_spec)
900 comparison_value = m.group('value')
901 op = STR_OPERATORS[m.group('op')]
904 raise ValueError('Invalid filter specification %r' % filter_spec)
907 actual_value = f.get(m.group('key'))
908 if actual_value is None:
909 return m.group('none_inclusive')
910 return op(actual_value, comparison_value)
913 def build_format_selector(self, format_spec):
914 def syntax_error(note, start):
916 'Invalid format specification: '
917 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
918 return SyntaxError(message)
920 PICKFIRST = 'PICKFIRST'
924 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
926 def _parse_filter(tokens):
928 for type, string, start, _, _ in tokens:
929 if type == tokenize.OP and string == ']':
930 return ''.join(filter_parts)
932 filter_parts.append(string)
934 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
936 current_selector = None
937 for type, string, start, _, _ in tokens:
938 # ENCODING is only defined in python 3.x
939 if type == getattr(tokenize, 'ENCODING', None):
941 elif type in [tokenize.NAME, tokenize.NUMBER]:
942 current_selector = FormatSelector(SINGLE, string, [])
943 elif type == tokenize.OP:
946 # ')' will be handled by the parentheses group
947 tokens.restore_last_token()
949 elif inside_merge and string in ['/', ',']:
950 tokens.restore_last_token()
952 elif inside_choice and string == ',':
953 tokens.restore_last_token()
956 if not current_selector:
957 raise syntax_error('"," must follow a format selector', start)
958 selectors.append(current_selector)
959 current_selector = None
961 first_choice = current_selector
962 second_choice = _parse_format_selection(tokens, inside_choice=True)
963 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
965 if not current_selector:
966 current_selector = FormatSelector(SINGLE, 'best', [])
967 format_filter = _parse_filter(tokens)
968 current_selector.filters.append(format_filter)
971 raise syntax_error('Unexpected "("', start)
972 group = _parse_format_selection(tokens, inside_group=True)
973 current_selector = FormatSelector(GROUP, group, [])
975 video_selector = current_selector
976 audio_selector = _parse_format_selection(tokens, inside_merge=True)
977 if not video_selector or not audio_selector:
978 raise syntax_error('"+" must be between two format selectors', start)
979 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
981 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
982 elif type == tokenize.ENDMARKER:
985 selectors.append(current_selector)
988 def _build_selector_function(selector):
989 if isinstance(selector, list):
990 fs = [_build_selector_function(s) for s in selector]
992 def selector_function(formats):
994 for format in f(formats):
996 return selector_function
997 elif selector.type == GROUP:
998 selector_function = _build_selector_function(selector.selector)
999 elif selector.type == PICKFIRST:
1000 fs = [_build_selector_function(s) for s in selector.selector]
1002 def selector_function(formats):
1004 picked_formats = list(f(formats))
1006 return picked_formats
1008 elif selector.type == SINGLE:
1009 format_spec = selector.selector
1011 def selector_function(formats):
1012 formats = list(formats)
1015 if format_spec == 'all':
1018 elif format_spec in ['best', 'worst', None]:
1019 format_idx = 0 if format_spec == 'worst' else -1
1020 audiovideo_formats = [
1022 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1023 if audiovideo_formats:
1024 yield audiovideo_formats[format_idx]
1025 # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format
1026 elif (all(f.get('acodec') != 'none' for f in formats) or
1027 all(f.get('vcodec') != 'none' for f in formats)):
1028 yield formats[format_idx]
1029 elif format_spec == 'bestaudio':
1032 if f.get('vcodec') == 'none']
1034 yield audio_formats[-1]
1035 elif format_spec == 'worstaudio':
1038 if f.get('vcodec') == 'none']
1040 yield audio_formats[0]
1041 elif format_spec == 'bestvideo':
1044 if f.get('acodec') == 'none']
1046 yield video_formats[-1]
1047 elif format_spec == 'worstvideo':
1050 if f.get('acodec') == 'none']
1052 yield video_formats[0]
1054 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1055 if format_spec in extensions:
1056 filter_f = lambda f: f['ext'] == format_spec
1058 filter_f = lambda f: f['format_id'] == format_spec
1059 matches = list(filter(filter_f, formats))
1062 elif selector.type == MERGE:
1063 def _merge(formats_info):
1064 format_1, format_2 = [f['format_id'] for f in formats_info]
1065 # The first format must contain the video and the
1067 if formats_info[0].get('vcodec') == 'none':
1068 self.report_error('The first format must '
1069 'contain the video, try using '
1070 '"-f %s+%s"' % (format_2, format_1))
1073 formats_info[0]['ext']
1074 if self.params.get('merge_output_format') is None
1075 else self.params['merge_output_format'])
1077 'requested_formats': formats_info,
1078 'format': '%s+%s' % (formats_info[0].get('format'),
1079 formats_info[1].get('format')),
1080 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1081 formats_info[1].get('format_id')),
1082 'width': formats_info[0].get('width'),
1083 'height': formats_info[0].get('height'),
1084 'resolution': formats_info[0].get('resolution'),
1085 'fps': formats_info[0].get('fps'),
1086 'vcodec': formats_info[0].get('vcodec'),
1087 'vbr': formats_info[0].get('vbr'),
1088 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1089 'acodec': formats_info[1].get('acodec'),
1090 'abr': formats_info[1].get('abr'),
1093 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1095 def selector_function(formats):
1096 formats = list(formats)
1097 for pair in itertools.product(video_selector(formats), audio_selector(formats)):
1100 filters = [self._build_format_filter(f) for f in selector.filters]
1102 def final_selector(formats):
1103 for _filter in filters:
1104 formats = list(filter(_filter, formats))
1105 return selector_function(formats)
1106 return final_selector
1108 stream = io.BytesIO(format_spec.encode('utf-8'))
1110 tokens = list(compat_tokenize_tokenize(stream.readline))
1111 except tokenize.TokenError:
1112 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1114 class TokenIterator(object):
1115 def __init__(self, tokens):
1116 self.tokens = tokens
1123 if self.counter >= len(self.tokens):
1124 raise StopIteration()
1125 value = self.tokens[self.counter]
1131 def restore_last_token(self):
1134 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1135 return _build_selector_function(parsed_selector)
1137 def _calc_headers(self, info_dict):
1138 res = std_headers.copy()
1140 add_headers = info_dict.get('http_headers')
1142 res.update(add_headers)
1144 cookies = self._calc_cookies(info_dict)
1146 res['Cookie'] = cookies
1150 def _calc_cookies(self, info_dict):
1151 pr = compat_urllib_request.Request(info_dict['url'])
1152 self.cookiejar.add_cookie_header(pr)
1153 return pr.get_header('Cookie')
1155 def process_video_result(self, info_dict, download=True):
1156 assert info_dict.get('_type', 'video') == 'video'
1158 if 'id' not in info_dict:
1159 raise ExtractorError('Missing "id" field in extractor result')
1160 if 'title' not in info_dict:
1161 raise ExtractorError('Missing "title" field in extractor result')
1163 if 'playlist' not in info_dict:
1164 # It isn't part of a playlist
1165 info_dict['playlist'] = None
1166 info_dict['playlist_index'] = None
1168 thumbnails = info_dict.get('thumbnails')
1169 if thumbnails is None:
1170 thumbnail = info_dict.get('thumbnail')
1172 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1174 thumbnails.sort(key=lambda t: (
1175 t.get('preference'), t.get('width'), t.get('height'),
1176 t.get('id'), t.get('url')))
1177 for i, t in enumerate(thumbnails):
1178 if 'width' in t and 'height' in t:
1179 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1180 if t.get('id') is None:
1183 if thumbnails and 'thumbnail' not in info_dict:
1184 info_dict['thumbnail'] = thumbnails[-1]['url']
1186 if 'display_id' not in info_dict and 'id' in info_dict:
1187 info_dict['display_id'] = info_dict['id']
1189 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1190 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1191 # see http://bugs.python.org/issue1646728)
1193 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1194 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1195 except (ValueError, OverflowError, OSError):
1198 if self.params.get('listsubtitles', False):
1199 if 'automatic_captions' in info_dict:
1200 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1201 self.list_subtitles(info_dict['id'], info_dict.get('subtitles'), 'subtitles')
1203 info_dict['requested_subtitles'] = self.process_subtitles(
1204 info_dict['id'], info_dict.get('subtitles'),
1205 info_dict.get('automatic_captions'))
1207 # We now pick which formats have to be downloaded
1208 if info_dict.get('formats') is None:
1209 # There's only one format available
1210 formats = [info_dict]
1212 formats = info_dict['formats']
1215 raise ExtractorError('No video formats found!')
1219 # We check that all the formats have the format and format_id fields
1220 for i, format in enumerate(formats):
1221 if 'url' not in format:
1222 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1224 if format.get('format_id') is None:
1225 format['format_id'] = compat_str(i)
1226 format_id = format['format_id']
1227 if format_id not in formats_dict:
1228 formats_dict[format_id] = []
1229 formats_dict[format_id].append(format)
1231 # Make sure all formats have unique format_id
1232 for format_id, ambiguous_formats in formats_dict.items():
1233 if len(ambiguous_formats) > 1:
1234 for i, format in enumerate(ambiguous_formats):
1235 format['format_id'] = '%s-%d' % (format_id, i)
1237 for i, format in enumerate(formats):
1238 if format.get('format') is None:
1239 format['format'] = '{id} - {res}{note}'.format(
1240 id=format['format_id'],
1241 res=self.format_resolution(format),
1242 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1244 # Automatically determine file extension if missing
1245 if 'ext' not in format:
1246 format['ext'] = determine_ext(format['url']).lower()
1247 # Add HTTP headers, so that external programs can use them from the
1249 full_format_info = info_dict.copy()
1250 full_format_info.update(format)
1251 format['http_headers'] = self._calc_headers(full_format_info)
1253 # TODO Central sorting goes here
1255 if formats[0] is not info_dict:
1256 # only set the 'formats' fields if the original info_dict list them
1257 # otherwise we end up with a circular reference, the first (and unique)
1258 # element in the 'formats' field in info_dict is info_dict itself,
1259 # wich can't be exported to json
1260 info_dict['formats'] = formats
1261 if self.params.get('listformats'):
1262 self.list_formats(info_dict)
1264 if self.params.get('list_thumbnails'):
1265 self.list_thumbnails(info_dict)
1268 req_format = self.params.get('format')
1269 if req_format is None:
1270 req_format_list = []
1271 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1272 info_dict['extractor'] in ['youtube', 'ted']):
1273 merger = FFmpegMergerPP(self)
1274 if merger.available and merger.can_merge():
1275 req_format_list.append('bestvideo+bestaudio')
1276 req_format_list.append('best')
1277 req_format = '/'.join(req_format_list)
1278 format_selector = self.build_format_selector(req_format)
1279 formats_to_download = list(format_selector(formats))
1280 if not formats_to_download:
1281 raise ExtractorError('requested format not available',
1285 if len(formats_to_download) > 1:
1286 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1287 for format in formats_to_download:
1288 new_info = dict(info_dict)
1289 new_info.update(format)
1290 self.process_info(new_info)
1291 # We update the info dict with the best quality format (backwards compatibility)
1292 info_dict.update(formats_to_download[-1])
1295 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1296 """Select the requested subtitles and their format"""
1298 if normal_subtitles and self.params.get('writesubtitles'):
1299 available_subs.update(normal_subtitles)
1300 if automatic_captions and self.params.get('writeautomaticsub'):
1301 for lang, cap_info in automatic_captions.items():
1302 if lang not in available_subs:
1303 available_subs[lang] = cap_info
1305 if (not self.params.get('writesubtitles') and not
1306 self.params.get('writeautomaticsub') or not
1310 if self.params.get('allsubtitles', False):
1311 requested_langs = available_subs.keys()
1313 if self.params.get('subtitleslangs', False):
1314 requested_langs = self.params.get('subtitleslangs')
1315 elif 'en' in available_subs:
1316 requested_langs = ['en']
1318 requested_langs = [list(available_subs.keys())[0]]
1320 formats_query = self.params.get('subtitlesformat', 'best')
1321 formats_preference = formats_query.split('/') if formats_query else []
1323 for lang in requested_langs:
1324 formats = available_subs.get(lang)
1326 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1328 for ext in formats_preference:
1332 matches = list(filter(lambda f: f['ext'] == ext, formats))
1338 self.report_warning(
1339 'No subtitle format found matching "%s" for language %s, '
1340 'using %s' % (formats_query, lang, f['ext']))
1344 def process_info(self, info_dict):
1345 """Process a single resolved IE result."""
1347 assert info_dict.get('_type', 'video') == 'video'
1349 max_downloads = self.params.get('max_downloads')
1350 if max_downloads is not None:
1351 if self._num_downloads >= int(max_downloads):
1352 raise MaxDownloadsReached()
1354 info_dict['fulltitle'] = info_dict['title']
1355 if len(info_dict['title']) > 200:
1356 info_dict['title'] = info_dict['title'][:197] + '...'
1358 if 'format' not in info_dict:
1359 info_dict['format'] = info_dict['ext']
1361 reason = self._match_entry(info_dict, incomplete=False)
1362 if reason is not None:
1363 self.to_screen('[download] ' + reason)
1366 self._num_downloads += 1
1368 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1371 if self.params.get('forcetitle', False):
1372 self.to_stdout(info_dict['fulltitle'])
1373 if self.params.get('forceid', False):
1374 self.to_stdout(info_dict['id'])
1375 if self.params.get('forceurl', False):
1376 if info_dict.get('requested_formats') is not None:
1377 for f in info_dict['requested_formats']:
1378 self.to_stdout(f['url'] + f.get('play_path', ''))
1380 # For RTMP URLs, also include the playpath
1381 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1382 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1383 self.to_stdout(info_dict['thumbnail'])
1384 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1385 self.to_stdout(info_dict['description'])
1386 if self.params.get('forcefilename', False) and filename is not None:
1387 self.to_stdout(filename)
1388 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1389 self.to_stdout(formatSeconds(info_dict['duration']))
1390 if self.params.get('forceformat', False):
1391 self.to_stdout(info_dict['format'])
1392 if self.params.get('forcejson', False):
1393 self.to_stdout(json.dumps(info_dict))
1395 # Do nothing else if in simulate mode
1396 if self.params.get('simulate', False):
1399 if filename is None:
1403 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1404 if dn and not os.path.exists(dn):
1406 except (OSError, IOError) as err:
1407 self.report_error('unable to create directory ' + compat_str(err))
1410 if self.params.get('writedescription', False):
1411 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1412 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1413 self.to_screen('[info] Video description is already present')
1414 elif info_dict.get('description') is None:
1415 self.report_warning('There\'s no description to write.')
1418 self.to_screen('[info] Writing video description to: ' + descfn)
1419 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1420 descfile.write(info_dict['description'])
1421 except (OSError, IOError):
1422 self.report_error('Cannot write description file ' + descfn)
1425 if self.params.get('writeannotations', False):
1426 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1427 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1428 self.to_screen('[info] Video annotations are already present')
1431 self.to_screen('[info] Writing video annotations to: ' + annofn)
1432 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1433 annofile.write(info_dict['annotations'])
1434 except (KeyError, TypeError):
1435 self.report_warning('There are no annotations to write.')
1436 except (OSError, IOError):
1437 self.report_error('Cannot write annotations file: ' + annofn)
1440 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1441 self.params.get('writeautomaticsub')])
1443 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1444 # subtitles download errors are already managed as troubles in relevant IE
1445 # that way it will silently go on when used with unsupporting IE
1446 subtitles = info_dict['requested_subtitles']
1447 ie = self.get_info_extractor(info_dict['extractor_key'])
1448 for sub_lang, sub_info in subtitles.items():
1449 sub_format = sub_info['ext']
1450 if sub_info.get('data') is not None:
1451 sub_data = sub_info['data']
1454 sub_data = ie._download_webpage(
1455 sub_info['url'], info_dict['id'], note=False)
1456 except ExtractorError as err:
1457 self.report_warning('Unable to download subtitle for "%s": %s' %
1458 (sub_lang, compat_str(err.cause)))
1461 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1462 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1463 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1465 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1466 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1467 subfile.write(sub_data)
1468 except (OSError, IOError):
1469 self.report_error('Cannot write subtitles file ' + sub_filename)
1472 if self.params.get('writeinfojson', False):
1473 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1474 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1475 self.to_screen('[info] Video description metadata is already present')
1477 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1479 write_json_file(self.filter_requested_info(info_dict), infofn)
1480 except (OSError, IOError):
1481 self.report_error('Cannot write metadata to JSON file ' + infofn)
1484 self._write_thumbnails(info_dict, filename)
1486 if not self.params.get('skip_download', False):
1489 fd = get_suitable_downloader(info, self.params)(self, self.params)
1490 for ph in self._progress_hooks:
1491 fd.add_progress_hook(ph)
1492 if self.params.get('verbose'):
1493 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1494 return fd.download(name, info)
1496 if info_dict.get('requested_formats') is not None:
1499 merger = FFmpegMergerPP(self)
1500 if not merger.available:
1502 self.report_warning('You have requested multiple '
1503 'formats but ffmpeg or avconv are not installed.'
1504 ' The formats won\'t be merged.')
1506 postprocessors = [merger]
1508 def compatible_formats(formats):
1509 video, audio = formats
1511 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1512 if video_ext and audio_ext:
1514 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1517 for exts in COMPATIBLE_EXTS:
1518 if video_ext in exts and audio_ext in exts:
1520 # TODO: Check acodec/vcodec
1523 filename_real_ext = os.path.splitext(filename)[1][1:]
1525 os.path.splitext(filename)[0]
1526 if filename_real_ext == info_dict['ext']
1528 requested_formats = info_dict['requested_formats']
1529 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1530 info_dict['ext'] = 'mkv'
1531 self.report_warning(
1532 'Requested formats are incompatible for merge and will be merged into mkv.')
1533 # Ensure filename always has a correct extension for successful merge
1534 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1535 if os.path.exists(encodeFilename(filename)):
1537 '[download] %s has already been downloaded and '
1538 'merged' % filename)
1540 for f in requested_formats:
1541 new_info = dict(info_dict)
1543 fname = self.prepare_filename(new_info)
1544 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1545 downloaded.append(fname)
1546 partial_success = dl(fname, new_info)
1547 success = success and partial_success
1548 info_dict['__postprocessors'] = postprocessors
1549 info_dict['__files_to_merge'] = downloaded
1551 # Just a single file
1552 success = dl(filename, info_dict)
1553 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1554 self.report_error('unable to download video data: %s' % str(err))
1556 except (OSError, IOError) as err:
1557 raise UnavailableVideoError(err)
1558 except (ContentTooShortError, ) as err:
1559 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1564 fixup_policy = self.params.get('fixup')
1565 if fixup_policy is None:
1566 fixup_policy = 'detect_or_warn'
1568 stretched_ratio = info_dict.get('stretched_ratio')
1569 if stretched_ratio is not None and stretched_ratio != 1:
1570 if fixup_policy == 'warn':
1571 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1572 info_dict['id'], stretched_ratio))
1573 elif fixup_policy == 'detect_or_warn':
1574 stretched_pp = FFmpegFixupStretchedPP(self)
1575 if stretched_pp.available:
1576 info_dict.setdefault('__postprocessors', [])
1577 info_dict['__postprocessors'].append(stretched_pp)
1579 self.report_warning(
1580 '%s: Non-uniform pixel ratio (%s). Install ffmpeg or avconv to fix this automatically.' % (
1581 info_dict['id'], stretched_ratio))
1583 assert fixup_policy in ('ignore', 'never')
1585 if info_dict.get('requested_formats') is None and info_dict.get('container') == 'm4a_dash':
1586 if fixup_policy == 'warn':
1587 self.report_warning('%s: writing DASH m4a. Only some players support this container.' % (
1589 elif fixup_policy == 'detect_or_warn':
1590 fixup_pp = FFmpegFixupM4aPP(self)
1591 if fixup_pp.available:
1592 info_dict.setdefault('__postprocessors', [])
1593 info_dict['__postprocessors'].append(fixup_pp)
1595 self.report_warning(
1596 '%s: writing DASH m4a. Only some players support this container. Install ffmpeg or avconv to fix this automatically.' % (
1599 assert fixup_policy in ('ignore', 'never')
1602 self.post_process(filename, info_dict)
1603 except (PostProcessingError) as err:
1604 self.report_error('postprocessing: %s' % str(err))
1606 self.record_download_archive(info_dict)
1608 def download(self, url_list):
1609 """Download a given list of URLs."""
1610 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1611 if (len(url_list) > 1 and
1612 '%' not in outtmpl and
1613 self.params.get('max_downloads') != 1):
1614 raise SameFileError(outtmpl)
1616 for url in url_list:
1618 # It also downloads the videos
1619 res = self.extract_info(
1620 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1621 except UnavailableVideoError:
1622 self.report_error('unable to download video')
1623 except MaxDownloadsReached:
1624 self.to_screen('[info] Maximum number of downloaded files reached.')
1627 if self.params.get('dump_single_json', False):
1628 self.to_stdout(json.dumps(res))
1630 return self._download_retcode
1632 def download_with_info_file(self, info_filename):
1633 with contextlib.closing(fileinput.FileInput(
1634 [info_filename], mode='r',
1635 openhook=fileinput.hook_encoded('utf-8'))) as f:
1636 # FileInput doesn't have a read method, we can't call json.load
1637 info = self.filter_requested_info(json.loads('\n'.join(f)))
1639 self.process_ie_result(info, download=True)
1640 except DownloadError:
1641 webpage_url = info.get('webpage_url')
1642 if webpage_url is not None:
1643 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1644 return self.download([webpage_url])
1647 return self._download_retcode
1650 def filter_requested_info(info_dict):
1652 (k, v) for k, v in info_dict.items()
1653 if k not in ['requested_formats', 'requested_subtitles'])
1655 def post_process(self, filename, ie_info):
1656 """Run all the postprocessors on the given file."""
1657 info = dict(ie_info)
1658 info['filepath'] = filename
1660 if ie_info.get('__postprocessors') is not None:
1661 pps_chain.extend(ie_info['__postprocessors'])
1662 pps_chain.extend(self._pps)
1663 for pp in pps_chain:
1664 files_to_delete = []
1666 files_to_delete, info = pp.run(info)
1667 except PostProcessingError as e:
1668 self.report_error(e.msg)
1669 if files_to_delete and not self.params.get('keepvideo', False):
1670 for old_filename in files_to_delete:
1671 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1673 os.remove(encodeFilename(old_filename))
1674 except (IOError, OSError):
1675 self.report_warning('Unable to remove downloaded original file')
1677 def _make_archive_id(self, info_dict):
1678 # Future-proof against any change in case
1679 # and backwards compatibility with prior versions
1680 extractor = info_dict.get('extractor_key')
1681 if extractor is None:
1682 if 'id' in info_dict:
1683 extractor = info_dict.get('ie_key') # key in a playlist
1684 if extractor is None:
1685 return None # Incomplete video information
1686 return extractor.lower() + ' ' + info_dict['id']
1688 def in_download_archive(self, info_dict):
1689 fn = self.params.get('download_archive')
1693 vid_id = self._make_archive_id(info_dict)
1695 return False # Incomplete video information
1698 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1699 for line in archive_file:
1700 if line.strip() == vid_id:
1702 except IOError as ioe:
1703 if ioe.errno != errno.ENOENT:
1707 def record_download_archive(self, info_dict):
1708 fn = self.params.get('download_archive')
1711 vid_id = self._make_archive_id(info_dict)
1713 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1714 archive_file.write(vid_id + '\n')
1717 def format_resolution(format, default='unknown'):
1718 if format.get('vcodec') == 'none':
1720 if format.get('resolution') is not None:
1721 return format['resolution']
1722 if format.get('height') is not None:
1723 if format.get('width') is not None:
1724 res = '%sx%s' % (format['width'], format['height'])
1726 res = '%sp' % format['height']
1727 elif format.get('width') is not None:
1728 res = '?x%d' % format['width']
1733 def _format_note(self, fdict):
1735 if fdict.get('ext') in ['f4f', 'f4m']:
1736 res += '(unsupported) '
1737 if fdict.get('format_note') is not None:
1738 res += fdict['format_note'] + ' '
1739 if fdict.get('tbr') is not None:
1740 res += '%4dk ' % fdict['tbr']
1741 if fdict.get('container') is not None:
1744 res += '%s container' % fdict['container']
1745 if (fdict.get('vcodec') is not None and
1746 fdict.get('vcodec') != 'none'):
1749 res += fdict['vcodec']
1750 if fdict.get('vbr') is not None:
1752 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1754 if fdict.get('vbr') is not None:
1755 res += '%4dk' % fdict['vbr']
1756 if fdict.get('fps') is not None:
1757 res += ', %sfps' % fdict['fps']
1758 if fdict.get('acodec') is not None:
1761 if fdict['acodec'] == 'none':
1764 res += '%-5s' % fdict['acodec']
1765 elif fdict.get('abr') is not None:
1769 if fdict.get('abr') is not None:
1770 res += '@%3dk' % fdict['abr']
1771 if fdict.get('asr') is not None:
1772 res += ' (%5dHz)' % fdict['asr']
1773 if fdict.get('filesize') is not None:
1776 res += format_bytes(fdict['filesize'])
1777 elif fdict.get('filesize_approx') is not None:
1780 res += '~' + format_bytes(fdict['filesize_approx'])
1783 def list_formats(self, info_dict):
1784 formats = info_dict.get('formats', [info_dict])
1786 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1788 if f.get('preference') is None or f['preference'] >= -1000]
1789 if len(formats) > 1:
1790 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1792 header_line = ['format code', 'extension', 'resolution', 'note']
1794 '[info] Available formats for %s:\n%s' %
1795 (info_dict['id'], render_table(header_line, table)))
1797 def list_thumbnails(self, info_dict):
1798 thumbnails = info_dict.get('thumbnails')
1800 tn_url = info_dict.get('thumbnail')
1802 thumbnails = [{'id': '0', 'url': tn_url}]
1805 '[info] No thumbnails present for %s' % info_dict['id'])
1809 '[info] Thumbnails for %s:' % info_dict['id'])
1810 self.to_screen(render_table(
1811 ['ID', 'width', 'height', 'URL'],
1812 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1814 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1816 self.to_screen('%s has no %s' % (video_id, name))
1819 'Available %s for %s:' % (name, video_id))
1820 self.to_screen(render_table(
1821 ['Language', 'formats'],
1822 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1823 for lang, formats in subtitles.items()]))
1825 def urlopen(self, req):
1826 """ Start an HTTP download """
1828 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1829 # always respected by websites, some tend to give out URLs with non percent-encoded
1830 # non-ASCII characters (see telemb.py, ard.py [#3412])
1831 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1832 # To work around aforementioned issue we will replace request's original URL with
1833 # percent-encoded one
1834 req_is_string = isinstance(req, compat_basestring)
1835 url = req if req_is_string else req.get_full_url()
1836 url_escaped = escape_url(url)
1838 # Substitute URL if any change after escaping
1839 if url != url_escaped:
1843 req_type = HEADRequest if req.get_method() == 'HEAD' else compat_urllib_request.Request
1845 url_escaped, data=req.data, headers=req.headers,
1846 origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
1848 return self._opener.open(req, timeout=self._socket_timeout)
1850 def print_debug_header(self):
1851 if not self.params.get('verbose'):
1854 if type('') is not compat_str:
1855 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1856 self.report_warning(
1857 'Your Python is broken! Update to a newer and supported version')
1859 stdout_encoding = getattr(
1860 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1862 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1863 locale.getpreferredencoding(),
1864 sys.getfilesystemencoding(),
1866 self.get_encoding()))
1867 write_string(encoding_str, encoding=None)
1869 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1871 sp = subprocess.Popen(
1872 ['git', 'rev-parse', '--short', 'HEAD'],
1873 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1874 cwd=os.path.dirname(os.path.abspath(__file__)))
1875 out, err = sp.communicate()
1876 out = out.decode().strip()
1877 if re.match('[0-9a-f]+', out):
1878 self._write_string('[debug] Git HEAD: ' + out + '\n')
1884 self._write_string('[debug] Python version %s - %s\n' % (
1885 platform.python_version(), platform_name()))
1887 exe_versions = FFmpegPostProcessor.get_versions(self)
1888 exe_versions['rtmpdump'] = rtmpdump_version()
1889 exe_str = ', '.join(
1891 for exe, v in sorted(exe_versions.items())
1896 self._write_string('[debug] exe versions: %s\n' % exe_str)
1899 for handler in self._opener.handlers:
1900 if hasattr(handler, 'proxies'):
1901 proxy_map.update(handler.proxies)
1902 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1904 if self.params.get('call_home', False):
1905 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
1906 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
1907 latest_version = self.urlopen(
1908 'https://yt-dl.org/latest/version').read().decode('utf-8')
1909 if version_tuple(latest_version) > version_tuple(__version__):
1910 self.report_warning(
1911 'You are using an outdated version (newest version: %s)! '
1912 'See https://yt-dl.org/update if you need help updating.' %
1915 def _setup_opener(self):
1916 timeout_val = self.params.get('socket_timeout')
1917 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
1919 opts_cookiefile = self.params.get('cookiefile')
1920 opts_proxy = self.params.get('proxy')
1922 if opts_cookiefile is None:
1923 self.cookiejar = compat_cookiejar.CookieJar()
1925 self.cookiejar = compat_cookiejar.MozillaCookieJar(
1927 if os.access(opts_cookiefile, os.R_OK):
1928 self.cookiejar.load()
1930 cookie_processor = compat_urllib_request.HTTPCookieProcessor(
1932 if opts_proxy is not None:
1933 if opts_proxy == '':
1936 proxies = {'http': opts_proxy, 'https': opts_proxy}
1938 proxies = compat_urllib_request.getproxies()
1939 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
1940 if 'http' in proxies and 'https' not in proxies:
1941 proxies['https'] = proxies['http']
1942 proxy_handler = PerRequestProxyHandler(proxies)
1944 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
1945 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
1946 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
1947 opener = compat_urllib_request.build_opener(
1948 proxy_handler, https_handler, cookie_processor, ydlh)
1950 # Delete the default user-agent header, which would otherwise apply in
1951 # cases where our custom HTTP handler doesn't come into play
1952 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
1953 opener.addheaders = []
1954 self._opener = opener
1956 def encode(self, s):
1957 if isinstance(s, bytes):
1958 return s # Already encoded
1961 return s.encode(self.get_encoding())
1962 except UnicodeEncodeError as err:
1963 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
1966 def get_encoding(self):
1967 encoding = self.params.get('encoding')
1968 if encoding is None:
1969 encoding = preferredencoding()
1972 def _write_thumbnails(self, info_dict, filename):
1973 if self.params.get('writethumbnail', False):
1974 thumbnails = info_dict.get('thumbnails')
1976 thumbnails = [thumbnails[-1]]
1977 elif self.params.get('write_all_thumbnails', False):
1978 thumbnails = info_dict.get('thumbnails')
1983 # No thumbnails present, so return immediately
1986 for t in thumbnails:
1987 thumb_ext = determine_ext(t['url'], 'jpg')
1988 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
1989 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
1990 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
1992 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
1993 self.to_screen('[%s] %s: Thumbnail %sis already present' %
1994 (info_dict['extractor'], info_dict['id'], thumb_display_id))
1996 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
1997 (info_dict['extractor'], info_dict['id'], thumb_display_id))
1999 uf = self.urlopen(t['url'])
2000 with open(thumb_filename, 'wb') as thumbf:
2001 shutil.copyfileobj(uf, thumbf)
2002 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2003 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2004 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2005 self.report_warning('Unable to download thumbnail "%s": %s' %
2006 (t['url'], compat_str(err)))