2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import, unicode_literals
31 compat_urllib_request,
58 UnavailableVideoError,
66 from .cache import Cache
67 from .extractor import get_info_extractor, gen_extractors
68 from .downloader import get_suitable_downloader
69 from .downloader.rtmp import rtmpdump_version
70 from .postprocessor import FFmpegMergerPP, FFmpegPostProcessor
71 from .version import __version__
74 class YoutubeDL(object):
77 YoutubeDL objects are the ones responsible of downloading the
78 actual video file and writing it to disk if the user has requested
79 it, among some other tasks. In most cases there should be one per
80 program. As, given a video URL, the downloader doesn't know how to
81 extract all the needed information, task that InfoExtractors do, it
82 has to pass the URL to one of them.
84 For this, YoutubeDL objects have a method that allows
85 InfoExtractors to be registered in a given order. When it is passed
86 a URL, the YoutubeDL object handles it to the first InfoExtractor it
87 finds that reports being able to handle it. The InfoExtractor extracts
88 all the information about the video or videos the URL refers to, and
89 YoutubeDL process the extracted information, possibly using a File
90 Downloader to download the video.
92 YoutubeDL objects accept a lot of parameters. In order not to saturate
93 the object constructor with arguments, it receives a dictionary of
94 options instead. These options are available through the params
95 attribute for the InfoExtractors to use. The YoutubeDL also
96 registers itself as the downloader in charge for the InfoExtractors
97 that are added to it, so this is a "mutual registration".
101 username: Username for authentication purposes.
102 password: Password for authentication purposes.
103 videopassword: Password for acces a video.
104 usenetrc: Use netrc for authentication instead.
105 verbose: Print additional info to stdout.
106 quiet: Do not print messages to stdout.
107 no_warnings: Do not print out anything for warnings.
108 forceurl: Force printing final URL.
109 forcetitle: Force printing title.
110 forceid: Force printing ID.
111 forcethumbnail: Force printing thumbnail URL.
112 forcedescription: Force printing description.
113 forcefilename: Force printing final filename.
114 forceduration: Force printing duration.
115 forcejson: Force printing info_dict as JSON.
116 dump_single_json: Force printing the info_dict of the whole playlist
117 (or video) as a single JSON line.
118 simulate: Do not download the video files.
119 format: Video format code.
120 format_limit: Highest quality format to try.
121 outtmpl: Template for output names.
122 restrictfilenames: Do not allow "&" and spaces in file names
123 ignoreerrors: Do not stop on download errors.
124 nooverwrites: Prevent overwriting files.
125 playliststart: Playlist item to start at.
126 playlistend: Playlist item to end at.
127 matchtitle: Download only matching titles.
128 rejecttitle: Reject downloads for matching titles.
129 logger: Log messages to a logging.Logger instance.
130 logtostderr: Log messages to stderr instead of stdout.
131 writedescription: Write the video description to a .description file
132 writeinfojson: Write the video description to a .info.json file
133 writeannotations: Write the video annotations to a .annotations.xml file
134 writethumbnail: Write the thumbnail image to a file
135 writesubtitles: Write the video subtitles to a file
136 writeautomaticsub: Write the automatic subtitles to a file
137 allsubtitles: Downloads all the subtitles of the video
138 (requires writesubtitles or writeautomaticsub)
139 listsubtitles: Lists all available subtitles for the video
140 subtitlesformat: Subtitle format [srt/sbv/vtt] (default=srt)
141 subtitleslangs: List of languages of the subtitles to download
142 keepvideo: Keep the video file after post-processing
143 daterange: A DateRange object, download only if the upload_date is in the range.
144 skip_download: Skip the actual download of the video file
145 cachedir: Location of the cache files in the filesystem.
146 False to disable filesystem cache.
147 noplaylist: Download single video instead of a playlist if in doubt.
148 age_limit: An integer representing the user's age in years.
149 Unsuitable videos for the given age are skipped.
150 min_views: An integer representing the minimum view count the video
151 must have in order to not be skipped.
152 Videos without view count information are always
153 downloaded. None for no limit.
154 max_views: An integer representing the maximum view count.
155 Videos that are more popular than that are not
157 Videos without view count information are always
158 downloaded. None for no limit.
159 download_archive: File name of a file where all downloads are recorded.
160 Videos already present in the file are not downloaded
162 cookiefile: File name where cookies should be read from and dumped to.
163 nocheckcertificate:Do not verify SSL certificates
164 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
165 At the moment, this is only supported by YouTube.
166 proxy: URL of the proxy server to use
167 socket_timeout: Time to wait for unresponsive hosts, in seconds
168 bidi_workaround: Work around buggy terminals without bidirectional text
169 support, using fridibi
170 debug_printtraffic:Print out sent and received HTTP traffic
171 include_ads: Download ads as well
172 default_search: Prepend this string if an input url is not valid.
173 'auto' for elaborate guessing
174 encoding: Use this encoding instead of the system-specified.
175 extract_flat: Do not resolve URLs, return the immediate result.
176 Pass in 'in_playlist' to only show this behavior for
179 The following parameters are not used by YoutubeDL itself, they are used by
181 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
182 noresizebuffer, retries, continuedl, noprogress, consoletitle
184 The following options are used by the post processors:
185 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
186 otherwise prefer avconv.
187 exec_cmd: Arbitrary command to run after downloading
193 _download_retcode = None
194 _num_downloads = None
197 def __init__(self, params=None, auto_init=True):
198 """Create a FileDownloader object with the given options."""
202 self._ies_instances = {}
204 self._progress_hooks = []
205 self._download_retcode = 0
206 self._num_downloads = 0
207 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
208 self._err_file = sys.stderr
210 self.cache = Cache(self)
212 if params.get('bidi_workaround', False):
215 master, slave = pty.openpty()
216 width = get_term_width()
220 width_args = ['-w', str(width)]
222 stdin=subprocess.PIPE,
224 stderr=self._err_file)
226 self._output_process = subprocess.Popen(
227 ['bidiv'] + width_args, **sp_kwargs
230 self._output_process = subprocess.Popen(
231 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
232 self._output_channel = os.fdopen(master, 'rb')
233 except OSError as ose:
235 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 $PATH.')
239 if (sys.version_info >= (3,) and sys.platform != 'win32' and
240 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
241 and not params.get('restrictfilenames', False)):
242 # On Python 3, the Unicode filesystem API will throw errors (#1474)
244 'Assuming --restrict-filenames since file system encoding '
245 'cannot encode all characters. '
246 'Set the LC_ALL environment variable to fix this.')
247 self.params['restrictfilenames'] = True
249 if '%(stitle)s' in self.params.get('outtmpl', ''):
250 self.report_warning('%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.')
255 self.print_debug_header()
256 self.add_default_info_extractors()
258 def warn_if_short_id(self, argv):
259 # short YouTube ID starting with dash?
261 i for i, a in enumerate(argv)
262 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
266 [a for i, a in enumerate(argv) if i not in idxs] +
267 ['--'] + [argv[i] for i in idxs]
270 'Long argument string detected. '
271 'Use -- to separate parameters and URLs, like this:\n%s\n' %
272 args_to_str(correct_argv))
274 def add_info_extractor(self, ie):
275 """Add an InfoExtractor object to the end of the list."""
277 self._ies_instances[ie.ie_key()] = ie
278 ie.set_downloader(self)
280 def get_info_extractor(self, ie_key):
282 Get an instance of an IE with name ie_key, it will try to get one from
283 the _ies list, if there's no instance it will create a new one and add
284 it to the extractor list.
286 ie = self._ies_instances.get(ie_key)
288 ie = get_info_extractor(ie_key)()
289 self.add_info_extractor(ie)
292 def add_default_info_extractors(self):
294 Add the InfoExtractors returned by gen_extractors to the end of the list
296 for ie in gen_extractors():
297 self.add_info_extractor(ie)
299 def add_post_processor(self, pp):
300 """Add a PostProcessor object to the end of the chain."""
302 pp.set_downloader(self)
304 def add_progress_hook(self, ph):
305 """Add the progress hook (currently only for the file downloader)"""
306 self._progress_hooks.append(ph)
308 def _bidi_workaround(self, message):
309 if not hasattr(self, '_output_channel'):
312 assert hasattr(self, '_output_process')
313 assert isinstance(message, compat_str)
314 line_count = message.count('\n') + 1
315 self._output_process.stdin.write((message + '\n').encode('utf-8'))
316 self._output_process.stdin.flush()
317 res = ''.join(self._output_channel.readline().decode('utf-8')
318 for _ in range(line_count))
319 return res[:-len('\n')]
321 def to_screen(self, message, skip_eol=False):
322 """Print message to stdout if not in quiet mode."""
323 return self.to_stdout(message, skip_eol, check_quiet=True)
325 def _write_string(self, s, out=None):
326 write_string(s, out=out, encoding=self.params.get('encoding'))
328 def to_stdout(self, message, skip_eol=False, check_quiet=False):
329 """Print message to stdout if not in quiet mode."""
330 if self.params.get('logger'):
331 self.params['logger'].debug(message)
332 elif not check_quiet or not self.params.get('quiet', False):
333 message = self._bidi_workaround(message)
334 terminator = ['\n', ''][skip_eol]
335 output = message + terminator
337 self._write_string(output, self._screen_file)
339 def to_stderr(self, message):
340 """Print message to stderr."""
341 assert isinstance(message, compat_str)
342 if self.params.get('logger'):
343 self.params['logger'].error(message)
345 message = self._bidi_workaround(message)
346 output = message + '\n'
347 self._write_string(output, self._err_file)
349 def to_console_title(self, message):
350 if not self.params.get('consoletitle', False):
352 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
353 # c_wchar_p() might not be necessary if `message` is
354 # already of type unicode()
355 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
356 elif 'TERM' in os.environ:
357 self._write_string('\033]0;%s\007' % message, self._screen_file)
359 def save_console_title(self):
360 if not self.params.get('consoletitle', False):
362 if 'TERM' in os.environ:
363 # Save the title on stack
364 self._write_string('\033[22;0t', self._screen_file)
366 def restore_console_title(self):
367 if not self.params.get('consoletitle', False):
369 if 'TERM' in os.environ:
370 # Restore the title from stack
371 self._write_string('\033[23;0t', self._screen_file)
374 self.save_console_title()
377 def __exit__(self, *args):
378 self.restore_console_title()
380 if self.params.get('cookiefile') is not None:
381 self.cookiejar.save()
383 def trouble(self, message=None, tb=None):
384 """Determine action to take when a download problem appears.
386 Depending on if the downloader has been configured to ignore
387 download errors or not, this method may throw an exception or
388 not when errors are found, after printing the message.
390 tb, if given, is additional traceback information.
392 if message is not None:
393 self.to_stderr(message)
394 if self.params.get('verbose'):
396 if sys.exc_info()[0]: # if .trouble has been called from an except block
398 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
399 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
400 tb += compat_str(traceback.format_exc())
402 tb_data = traceback.format_list(traceback.extract_stack())
403 tb = ''.join(tb_data)
405 if not self.params.get('ignoreerrors', False):
406 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
407 exc_info = sys.exc_info()[1].exc_info
409 exc_info = sys.exc_info()
410 raise DownloadError(message, exc_info)
411 self._download_retcode = 1
413 def report_warning(self, message):
415 Print the message to stderr, it will be prefixed with 'WARNING:'
416 If stderr is a tty file the 'WARNING:' will be colored
418 if self.params.get('logger') is not None:
419 self.params['logger'].warning(message)
421 if self.params.get('no_warnings'):
423 if self._err_file.isatty() and os.name != 'nt':
424 _msg_header = '\033[0;33mWARNING:\033[0m'
426 _msg_header = 'WARNING:'
427 warning_message = '%s %s' % (_msg_header, message)
428 self.to_stderr(warning_message)
430 def report_error(self, message, tb=None):
432 Do the same as trouble, but prefixes the message with 'ERROR:', colored
433 in red if stderr is a tty file.
435 if self._err_file.isatty() and os.name != 'nt':
436 _msg_header = '\033[0;31mERROR:\033[0m'
438 _msg_header = 'ERROR:'
439 error_message = '%s %s' % (_msg_header, message)
440 self.trouble(error_message, tb)
442 def report_file_already_downloaded(self, file_name):
443 """Report file has already been fully downloaded."""
445 self.to_screen('[download] %s has already been downloaded' % file_name)
446 except UnicodeEncodeError:
447 self.to_screen('[download] The file has already been downloaded')
449 def prepare_filename(self, info_dict):
450 """Generate the output filename."""
452 template_dict = dict(info_dict)
454 template_dict['epoch'] = int(time.time())
455 autonumber_size = self.params.get('autonumber_size')
456 if autonumber_size is None:
458 autonumber_templ = '%0' + str(autonumber_size) + 'd'
459 template_dict['autonumber'] = autonumber_templ % self._num_downloads
460 if template_dict.get('playlist_index') is not None:
461 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
462 if template_dict.get('resolution') is None:
463 if template_dict.get('width') and template_dict.get('height'):
464 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
465 elif template_dict.get('height'):
466 template_dict['resolution'] = '%sp' % template_dict['height']
467 elif template_dict.get('width'):
468 template_dict['resolution'] = '?x%d' % template_dict['width']
470 sanitize = lambda k, v: sanitize_filename(
472 restricted=self.params.get('restrictfilenames'),
474 template_dict = dict((k, sanitize(k, v))
475 for k, v in template_dict.items()
477 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
479 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
480 tmpl = compat_expanduser(outtmpl)
481 filename = tmpl % template_dict
483 except ValueError as err:
484 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
487 def _match_entry(self, info_dict):
488 """ Returns None iff the file should be downloaded """
490 video_title = info_dict.get('title', info_dict.get('id', 'video'))
491 if 'title' in info_dict:
492 # This can happen when we're just evaluating the playlist
493 title = info_dict['title']
494 matchtitle = self.params.get('matchtitle', False)
496 if not re.search(matchtitle, title, re.IGNORECASE):
497 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
498 rejecttitle = self.params.get('rejecttitle', False)
500 if re.search(rejecttitle, title, re.IGNORECASE):
501 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
502 date = info_dict.get('upload_date', None)
504 dateRange = self.params.get('daterange', DateRange())
505 if date not in dateRange:
506 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
507 view_count = info_dict.get('view_count', None)
508 if view_count is not None:
509 min_views = self.params.get('min_views')
510 if min_views is not None and view_count < min_views:
511 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
512 max_views = self.params.get('max_views')
513 if max_views is not None and view_count > max_views:
514 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
515 age_limit = self.params.get('age_limit')
516 if age_limit is not None:
517 actual_age_limit = info_dict.get('age_limit')
518 if actual_age_limit is None:
520 if age_limit < actual_age_limit:
521 return 'Skipping "' + title + '" because it is age restricted'
522 if self.in_download_archive(info_dict):
523 return '%s has already been recorded in archive' % video_title
527 def add_extra_info(info_dict, extra_info):
528 '''Set the keys from extra_info in info dict if they are missing'''
529 for key, value in extra_info.items():
530 info_dict.setdefault(key, value)
532 def extract_info(self, url, download=True, ie_key=None, extra_info={},
535 Returns a list with a dictionary for each video we find.
536 If 'download', also downloads the videos.
537 extra_info is a dict containing the extra values to add to each result
541 ies = [self.get_info_extractor(ie_key)]
546 if not ie.suitable(url):
550 self.report_warning('The program functionality for this site has been marked as broken, '
551 'and will probably not work.')
554 ie_result = ie.extract(url)
555 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
557 if isinstance(ie_result, list):
558 # Backwards compatibility: old IE result format
560 '_type': 'compat_list',
561 'entries': ie_result,
563 self.add_default_extra_info(ie_result, ie, url)
565 return self.process_ie_result(ie_result, download, extra_info)
568 except ExtractorError as de: # An error we somewhat expected
569 self.report_error(compat_str(de), de.format_traceback())
571 except MaxDownloadsReached:
573 except Exception as e:
574 if self.params.get('ignoreerrors', False):
575 self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
580 self.report_error('no suitable InfoExtractor for URL %s' % url)
582 def add_default_extra_info(self, ie_result, ie, url):
583 self.add_extra_info(ie_result, {
584 'extractor': ie.IE_NAME,
586 'webpage_url_basename': url_basename(url),
587 'extractor_key': ie.ie_key(),
590 def process_ie_result(self, ie_result, download=True, extra_info={}):
592 Take the result of the ie(may be modified) and resolve all unresolved
593 references (URLs, playlist items).
595 It will also download the videos if 'download'.
596 Returns the resolved ie_result.
599 result_type = ie_result.get('_type', 'video')
601 if result_type in ('url', 'url_transparent'):
602 extract_flat = self.params.get('extract_flat', False)
603 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
604 extract_flat is True):
605 if self.params.get('forcejson', False):
606 self.to_stdout(json.dumps(ie_result))
609 if result_type == 'video':
610 self.add_extra_info(ie_result, extra_info)
611 return self.process_video_result(ie_result, download=download)
612 elif result_type == 'url':
613 # We have to add extra_info to the results because it may be
614 # contained in a playlist
615 return self.extract_info(ie_result['url'],
617 ie_key=ie_result.get('ie_key'),
618 extra_info=extra_info)
619 elif result_type == 'url_transparent':
620 # Use the information from the embedding page
621 info = self.extract_info(
622 ie_result['url'], ie_key=ie_result.get('ie_key'),
623 extra_info=extra_info, download=False, process=False)
625 def make_result(embedded_info):
626 new_result = ie_result.copy()
627 for f in ('_type', 'url', 'ext', 'player_url', 'formats',
628 'entries', 'ie_key', 'duration',
629 'subtitles', 'annotations', 'format',
630 'thumbnail', 'thumbnails'):
633 if f in embedded_info:
634 new_result[f] = embedded_info[f]
636 new_result = make_result(info)
638 assert new_result.get('_type') != 'url_transparent'
639 if new_result.get('_type') == 'compat_list':
640 new_result['entries'] = [
641 make_result(e) for e in new_result['entries']]
643 return self.process_ie_result(
644 new_result, download=download, extra_info=extra_info)
645 elif result_type == 'playlist' or result_type == 'multi_video':
646 # We process each entry in the playlist
647 playlist = ie_result.get('title', None) or ie_result.get('id', None)
648 self.to_screen('[download] Downloading playlist: %s' % playlist)
650 playlist_results = []
652 playliststart = self.params.get('playliststart', 1) - 1
653 playlistend = self.params.get('playlistend', None)
654 # For backwards compatibility, interpret -1 as whole list
655 if playlistend == -1:
658 if isinstance(ie_result['entries'], list):
659 n_all_entries = len(ie_result['entries'])
660 entries = ie_result['entries'][playliststart:playlistend]
661 n_entries = len(entries)
663 "[%s] playlist %s: Collected %d video ids (downloading %d of them)" %
664 (ie_result['extractor'], playlist, n_all_entries, n_entries))
666 assert isinstance(ie_result['entries'], PagedList)
667 entries = ie_result['entries'].getslice(
668 playliststart, playlistend)
669 n_entries = len(entries)
671 "[%s] playlist %s: Downloading %d videos" %
672 (ie_result['extractor'], playlist, n_entries))
674 for i, entry in enumerate(entries, 1):
675 self.to_screen('[download] Downloading video #%s of %s' % (i, n_entries))
677 'n_entries': n_entries,
678 'playlist': playlist,
679 'playlist_id': ie_result.get('id'),
680 'playlist_title': ie_result.get('title'),
681 'playlist_index': i + playliststart,
682 'extractor': ie_result['extractor'],
683 'webpage_url': ie_result['webpage_url'],
684 'webpage_url_basename': url_basename(ie_result['webpage_url']),
685 'extractor_key': ie_result['extractor_key'],
688 reason = self._match_entry(entry)
689 if reason is not None:
690 self.to_screen('[download] ' + reason)
693 entry_result = self.process_ie_result(entry,
696 playlist_results.append(entry_result)
697 ie_result['entries'] = playlist_results
699 elif result_type == 'compat_list':
701 'Extractor %s returned a compat_list result. '
702 'It needs to be updated.' % ie_result.get('extractor'))
705 self.add_extra_info(r,
707 'extractor': ie_result['extractor'],
708 'webpage_url': ie_result['webpage_url'],
709 'webpage_url_basename': url_basename(ie_result['webpage_url']),
710 'extractor_key': ie_result['extractor_key'],
713 ie_result['entries'] = [
714 self.process_ie_result(_fixup(r), download, extra_info)
715 for r in ie_result['entries']
719 raise Exception('Invalid result type: %s' % result_type)
721 def select_format(self, format_spec, available_formats):
722 if format_spec == 'best' or format_spec is None:
723 return available_formats[-1]
724 elif format_spec == 'worst':
725 return available_formats[0]
726 elif format_spec == 'bestaudio':
728 f for f in available_formats
729 if f.get('vcodec') == 'none']
731 return audio_formats[-1]
732 elif format_spec == 'worstaudio':
734 f for f in available_formats
735 if f.get('vcodec') == 'none']
737 return audio_formats[0]
738 elif format_spec == 'bestvideo':
740 f for f in available_formats
741 if f.get('acodec') == 'none']
743 return video_formats[-1]
744 elif format_spec == 'worstvideo':
746 f for f in available_formats
747 if f.get('acodec') == 'none']
749 return video_formats[0]
751 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a']
752 if format_spec in extensions:
753 filter_f = lambda f: f['ext'] == format_spec
755 filter_f = lambda f: f['format_id'] == format_spec
756 matches = list(filter(filter_f, available_formats))
761 def process_video_result(self, info_dict, download=True):
762 assert info_dict.get('_type', 'video') == 'video'
764 if 'id' not in info_dict:
765 raise ExtractorError('Missing "id" field in extractor result')
766 if 'title' not in info_dict:
767 raise ExtractorError('Missing "title" field in extractor result')
769 if 'playlist' not in info_dict:
770 # It isn't part of a playlist
771 info_dict['playlist'] = None
772 info_dict['playlist_index'] = None
774 thumbnails = info_dict.get('thumbnails')
776 thumbnails.sort(key=lambda t: (
777 t.get('width'), t.get('height'), t.get('url')))
779 if 'width' in t and 'height' in t:
780 t['resolution'] = '%dx%d' % (t['width'], t['height'])
782 if thumbnails and 'thumbnail' not in info_dict:
783 info_dict['thumbnail'] = thumbnails[-1]['url']
785 if 'display_id' not in info_dict and 'id' in info_dict:
786 info_dict['display_id'] = info_dict['id']
788 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
789 upload_date = datetime.datetime.utcfromtimestamp(
790 info_dict['timestamp'])
791 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
793 # This extractors handle format selection themselves
794 if info_dict['extractor'] in ['Youku']:
796 self.process_info(info_dict)
799 # We now pick which formats have to be downloaded
800 if info_dict.get('formats') is None:
801 # There's only one format available
802 formats = [info_dict]
804 formats = info_dict['formats']
807 raise ExtractorError('No video formats found!')
809 # We check that all the formats have the format and format_id fields
810 for i, format in enumerate(formats):
811 if 'url' not in format:
812 raise ExtractorError('Missing "url" key in result (index %d)' % i)
814 if format.get('format_id') is None:
815 format['format_id'] = compat_str(i)
816 if format.get('format') is None:
817 format['format'] = '{id} - {res}{note}'.format(
818 id=format['format_id'],
819 res=self.format_resolution(format),
820 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
822 # Automatically determine file extension if missing
823 if 'ext' not in format:
824 format['ext'] = determine_ext(format['url']).lower()
826 format_limit = self.params.get('format_limit', None)
828 formats = list(takewhile_inclusive(
829 lambda f: f['format_id'] != format_limit, formats
832 # TODO Central sorting goes here
834 if formats[0] is not info_dict:
835 # only set the 'formats' fields if the original info_dict list them
836 # otherwise we end up with a circular reference, the first (and unique)
837 # element in the 'formats' field in info_dict is info_dict itself,
838 # wich can't be exported to json
839 info_dict['formats'] = formats
840 if self.params.get('listformats', None):
841 self.list_formats(info_dict)
844 req_format = self.params.get('format')
845 if req_format is None:
847 formats_to_download = []
848 # The -1 is for supporting YoutubeIE
849 if req_format in ('-1', 'all'):
850 formats_to_download = formats
852 for rfstr in req_format.split(','):
853 # We can accept formats requested in the format: 34/5/best, we pick
854 # the first that is available, starting from left
855 req_formats = rfstr.split('/')
856 for rf in req_formats:
857 if re.match(r'.+?\+.+?', rf) is not None:
858 # Two formats have been requested like '137+139'
859 format_1, format_2 = rf.split('+')
860 formats_info = (self.select_format(format_1, formats),
861 self.select_format(format_2, formats))
862 if all(formats_info):
863 # The first format must contain the video and the
865 if formats_info[0].get('vcodec') == 'none':
866 self.report_error('The first format must '
867 'contain the video, try using '
868 '"-f %s+%s"' % (format_2, format_1))
871 'requested_formats': formats_info,
873 'ext': formats_info[0]['ext'],
876 selected_format = None
878 selected_format = self.select_format(rf, formats)
879 if selected_format is not None:
880 formats_to_download.append(selected_format)
882 if not formats_to_download:
883 raise ExtractorError('requested format not available',
887 if len(formats_to_download) > 1:
888 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
889 for format in formats_to_download:
890 new_info = dict(info_dict)
891 new_info.update(format)
892 self.process_info(new_info)
893 # We update the info dict with the best quality format (backwards compatibility)
894 info_dict.update(formats_to_download[-1])
897 def process_info(self, info_dict):
898 """Process a single resolved IE result."""
900 assert info_dict.get('_type', 'video') == 'video'
902 max_downloads = self.params.get('max_downloads')
903 if max_downloads is not None:
904 if self._num_downloads >= int(max_downloads):
905 raise MaxDownloadsReached()
907 info_dict['fulltitle'] = info_dict['title']
908 if len(info_dict['title']) > 200:
909 info_dict['title'] = info_dict['title'][:197] + '...'
911 # Keep for backwards compatibility
912 info_dict['stitle'] = info_dict['title']
914 if 'format' not in info_dict:
915 info_dict['format'] = info_dict['ext']
917 reason = self._match_entry(info_dict)
918 if reason is not None:
919 self.to_screen('[download] ' + reason)
922 self._num_downloads += 1
924 filename = self.prepare_filename(info_dict)
927 if self.params.get('forcetitle', False):
928 self.to_stdout(info_dict['fulltitle'])
929 if self.params.get('forceid', False):
930 self.to_stdout(info_dict['id'])
931 if self.params.get('forceurl', False):
932 # For RTMP URLs, also include the playpath
933 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
934 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
935 self.to_stdout(info_dict['thumbnail'])
936 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
937 self.to_stdout(info_dict['description'])
938 if self.params.get('forcefilename', False) and filename is not None:
939 self.to_stdout(filename)
940 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
941 self.to_stdout(formatSeconds(info_dict['duration']))
942 if self.params.get('forceformat', False):
943 self.to_stdout(info_dict['format'])
944 if self.params.get('forcejson', False):
945 info_dict['_filename'] = filename
946 self.to_stdout(json.dumps(info_dict))
947 if self.params.get('dump_single_json', False):
948 info_dict['_filename'] = filename
950 # Do nothing else if in simulate mode
951 if self.params.get('simulate', False):
958 dn = os.path.dirname(encodeFilename(filename))
959 if dn and not os.path.exists(dn):
961 except (OSError, IOError) as err:
962 self.report_error('unable to create directory ' + compat_str(err))
965 if self.params.get('writedescription', False):
966 descfn = filename + '.description'
967 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
968 self.to_screen('[info] Video description is already present')
971 self.to_screen('[info] Writing video description to: ' + descfn)
972 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
973 descfile.write(info_dict['description'])
974 except (KeyError, TypeError):
975 self.report_warning('There\'s no description to write.')
976 except (OSError, IOError):
977 self.report_error('Cannot write description file ' + descfn)
980 if self.params.get('writeannotations', False):
981 annofn = filename + '.annotations.xml'
982 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
983 self.to_screen('[info] Video annotations are already present')
986 self.to_screen('[info] Writing video annotations to: ' + annofn)
987 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
988 annofile.write(info_dict['annotations'])
989 except (KeyError, TypeError):
990 self.report_warning('There are no annotations to write.')
991 except (OSError, IOError):
992 self.report_error('Cannot write annotations file: ' + annofn)
995 subtitles_are_requested = any([self.params.get('writesubtitles', False),
996 self.params.get('writeautomaticsub')])
998 if subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']:
999 # subtitles download errors are already managed as troubles in relevant IE
1000 # that way it will silently go on when used with unsupporting IE
1001 subtitles = info_dict['subtitles']
1002 sub_format = self.params.get('subtitlesformat', 'srt')
1003 for sub_lang in subtitles.keys():
1004 sub = subtitles[sub_lang]
1008 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1009 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1010 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1012 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1013 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1015 except (OSError, IOError):
1016 self.report_error('Cannot write subtitles file ' + sub_filename)
1019 if self.params.get('writeinfojson', False):
1020 infofn = os.path.splitext(filename)[0] + '.info.json'
1021 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1022 self.to_screen('[info] Video description metadata is already present')
1024 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1026 write_json_file(info_dict, infofn)
1027 except (OSError, IOError):
1028 self.report_error('Cannot write metadata to JSON file ' + infofn)
1031 if self.params.get('writethumbnail', False):
1032 if info_dict.get('thumbnail') is not None:
1033 thumb_format = determine_ext(info_dict['thumbnail'], 'jpg')
1034 thumb_filename = os.path.splitext(filename)[0] + '.' + thumb_format
1035 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
1036 self.to_screen('[%s] %s: Thumbnail is already present' %
1037 (info_dict['extractor'], info_dict['id']))
1039 self.to_screen('[%s] %s: Downloading thumbnail ...' %
1040 (info_dict['extractor'], info_dict['id']))
1042 uf = self.urlopen(info_dict['thumbnail'])
1043 with open(thumb_filename, 'wb') as thumbf:
1044 shutil.copyfileobj(uf, thumbf)
1045 self.to_screen('[%s] %s: Writing thumbnail to: %s' %
1046 (info_dict['extractor'], info_dict['id'], thumb_filename))
1047 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1048 self.report_warning('Unable to download thumbnail "%s": %s' %
1049 (info_dict['thumbnail'], compat_str(err)))
1051 if not self.params.get('skip_download', False):
1052 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
1057 fd = get_suitable_downloader(info)(self, self.params)
1058 for ph in self._progress_hooks:
1059 fd.add_progress_hook(ph)
1060 if self.params.get('verbose'):
1061 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1062 return fd.download(name, info)
1063 if info_dict.get('requested_formats') is not None:
1066 merger = FFmpegMergerPP(self, not self.params.get('keepvideo'))
1067 if not merger._executable:
1069 self.report_warning('You have requested multiple '
1070 'formats but ffmpeg or avconv are not installed.'
1071 ' The formats won\'t be merged')
1073 postprocessors = [merger]
1074 for f in info_dict['requested_formats']:
1075 new_info = dict(info_dict)
1077 fname = self.prepare_filename(new_info)
1078 fname = prepend_extension(fname, 'f%s' % f['format_id'])
1079 downloaded.append(fname)
1080 partial_success = dl(fname, new_info)
1081 success = success and partial_success
1082 info_dict['__postprocessors'] = postprocessors
1083 info_dict['__files_to_merge'] = downloaded
1085 # Just a single file
1086 success = dl(filename, info_dict)
1087 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1088 self.report_error('unable to download video data: %s' % str(err))
1090 except (OSError, IOError) as err:
1091 raise UnavailableVideoError(err)
1092 except (ContentTooShortError, ) as err:
1093 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1098 self.post_process(filename, info_dict)
1099 except (PostProcessingError) as err:
1100 self.report_error('postprocessing: %s' % str(err))
1103 self.record_download_archive(info_dict)
1105 def download(self, url_list):
1106 """Download a given list of URLs."""
1107 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1108 if (len(url_list) > 1 and
1110 and self.params.get('max_downloads') != 1):
1111 raise SameFileError(outtmpl)
1113 for url in url_list:
1115 # It also downloads the videos
1116 res = self.extract_info(url)
1117 except UnavailableVideoError:
1118 self.report_error('unable to download video')
1119 except MaxDownloadsReached:
1120 self.to_screen('[info] Maximum number of downloaded files reached.')
1123 if self.params.get('dump_single_json', False):
1124 self.to_stdout(json.dumps(res))
1126 return self._download_retcode
1128 def download_with_info_file(self, info_filename):
1129 with io.open(info_filename, 'r', encoding='utf-8') as f:
1132 self.process_ie_result(info, download=True)
1133 except DownloadError:
1134 webpage_url = info.get('webpage_url')
1135 if webpage_url is not None:
1136 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1137 return self.download([webpage_url])
1140 return self._download_retcode
1142 def post_process(self, filename, ie_info):
1143 """Run all the postprocessors on the given file."""
1144 info = dict(ie_info)
1145 info['filepath'] = filename
1148 if ie_info.get('__postprocessors') is not None:
1149 pps_chain.extend(ie_info['__postprocessors'])
1150 pps_chain.extend(self._pps)
1151 for pp in pps_chain:
1153 keep_video_wish, new_info = pp.run(info)
1154 if keep_video_wish is not None:
1156 keep_video = keep_video_wish
1157 elif keep_video is None:
1158 # No clear decision yet, let IE decide
1159 keep_video = keep_video_wish
1160 except PostProcessingError as e:
1161 self.report_error(e.msg)
1162 if keep_video is False and not self.params.get('keepvideo', False):
1164 self.to_screen('Deleting original file %s (pass -k to keep)' % filename)
1165 os.remove(encodeFilename(filename))
1166 except (IOError, OSError):
1167 self.report_warning('Unable to remove downloaded video file')
1169 def _make_archive_id(self, info_dict):
1170 # Future-proof against any change in case
1171 # and backwards compatibility with prior versions
1172 extractor = info_dict.get('extractor_key')
1173 if extractor is None:
1174 if 'id' in info_dict:
1175 extractor = info_dict.get('ie_key') # key in a playlist
1176 if extractor is None:
1177 return None # Incomplete video information
1178 return extractor.lower() + ' ' + info_dict['id']
1180 def in_download_archive(self, info_dict):
1181 fn = self.params.get('download_archive')
1185 vid_id = self._make_archive_id(info_dict)
1187 return False # Incomplete video information
1190 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1191 for line in archive_file:
1192 if line.strip() == vid_id:
1194 except IOError as ioe:
1195 if ioe.errno != errno.ENOENT:
1199 def record_download_archive(self, info_dict):
1200 fn = self.params.get('download_archive')
1203 vid_id = self._make_archive_id(info_dict)
1205 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1206 archive_file.write(vid_id + '\n')
1209 def format_resolution(format, default='unknown'):
1210 if format.get('vcodec') == 'none':
1212 if format.get('resolution') is not None:
1213 return format['resolution']
1214 if format.get('height') is not None:
1215 if format.get('width') is not None:
1216 res = '%sx%s' % (format['width'], format['height'])
1218 res = '%sp' % format['height']
1219 elif format.get('width') is not None:
1220 res = '?x%d' % format['width']
1225 def _format_note(self, fdict):
1227 if fdict.get('ext') in ['f4f', 'f4m']:
1228 res += '(unsupported) '
1229 if fdict.get('format_note') is not None:
1230 res += fdict['format_note'] + ' '
1231 if fdict.get('tbr') is not None:
1232 res += '%4dk ' % fdict['tbr']
1233 if fdict.get('container') is not None:
1236 res += '%s container' % fdict['container']
1237 if (fdict.get('vcodec') is not None and
1238 fdict.get('vcodec') != 'none'):
1241 res += fdict['vcodec']
1242 if fdict.get('vbr') is not None:
1244 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1246 if fdict.get('vbr') is not None:
1247 res += '%4dk' % fdict['vbr']
1248 if fdict.get('fps') is not None:
1249 res += ', %sfps' % fdict['fps']
1250 if fdict.get('acodec') is not None:
1253 if fdict['acodec'] == 'none':
1256 res += '%-5s' % fdict['acodec']
1257 elif fdict.get('abr') is not None:
1261 if fdict.get('abr') is not None:
1262 res += '@%3dk' % fdict['abr']
1263 if fdict.get('asr') is not None:
1264 res += ' (%5dHz)' % fdict['asr']
1265 if fdict.get('filesize') is not None:
1268 res += format_bytes(fdict['filesize'])
1269 elif fdict.get('filesize_approx') is not None:
1272 res += '~' + format_bytes(fdict['filesize_approx'])
1275 def list_formats(self, info_dict):
1276 def line(format, idlen=20):
1277 return (('%-' + compat_str(idlen + 1) + 's%-10s%-12s%s') % (
1278 format['format_id'],
1280 self.format_resolution(format),
1281 self._format_note(format),
1284 formats = info_dict.get('formats', [info_dict])
1285 idlen = max(len('format code'),
1286 max(len(f['format_id']) for f in formats))
1287 formats_s = [line(f, idlen) for f in formats]
1288 if len(formats) > 1:
1289 formats_s[0] += (' ' if self._format_note(formats[0]) else '') + '(worst)'
1290 formats_s[-1] += (' ' if self._format_note(formats[-1]) else '') + '(best)'
1292 header_line = line({
1293 'format_id': 'format code', 'ext': 'extension',
1294 'resolution': 'resolution', 'format_note': 'note'}, idlen=idlen)
1295 self.to_screen('[info] Available formats for %s:\n%s\n%s' %
1296 (info_dict['id'], header_line, '\n'.join(formats_s)))
1298 def urlopen(self, req):
1299 """ Start an HTTP download """
1301 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1302 # always respected by websites, some tend to give out URLs with non percent-encoded
1303 # non-ASCII characters (see telemb.py, ard.py [#3412])
1304 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1305 # To work around aforementioned issue we will replace request's original URL with
1306 # percent-encoded one
1307 req_is_string = isinstance(req, basestring if sys.version_info < (3, 0) else compat_str)
1308 url = req if req_is_string else req.get_full_url()
1309 url_escaped = escape_url(url)
1311 # Substitute URL if any change after escaping
1312 if url != url_escaped:
1316 req = compat_urllib_request.Request(
1317 url_escaped, data=req.data, headers=req.headers,
1318 origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
1320 return self._opener.open(req, timeout=self._socket_timeout)
1322 def print_debug_header(self):
1323 if not self.params.get('verbose'):
1326 if type('') is not compat_str:
1327 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1328 self.report_warning(
1329 'Your Python is broken! Update to a newer and supported version')
1331 stdout_encoding = getattr(
1332 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1334 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1335 locale.getpreferredencoding(),
1336 sys.getfilesystemencoding(),
1338 self.get_encoding()))
1339 write_string(encoding_str, encoding=None)
1341 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1343 sp = subprocess.Popen(
1344 ['git', 'rev-parse', '--short', 'HEAD'],
1345 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1346 cwd=os.path.dirname(os.path.abspath(__file__)))
1347 out, err = sp.communicate()
1348 out = out.decode().strip()
1349 if re.match('[0-9a-f]+', out):
1350 self._write_string('[debug] Git HEAD: ' + out + '\n')
1356 self._write_string('[debug] Python version %s - %s\n' % (
1357 platform.python_version(), platform_name()))
1359 exe_versions = FFmpegPostProcessor.get_versions()
1360 exe_versions['rtmpdump'] = rtmpdump_version()
1361 exe_str = ', '.join(
1363 for exe, v in sorted(exe_versions.items())
1368 self._write_string('[debug] exe versions: %s\n' % exe_str)
1371 for handler in self._opener.handlers:
1372 if hasattr(handler, 'proxies'):
1373 proxy_map.update(handler.proxies)
1374 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1376 def _setup_opener(self):
1377 timeout_val = self.params.get('socket_timeout')
1378 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
1380 opts_cookiefile = self.params.get('cookiefile')
1381 opts_proxy = self.params.get('proxy')
1383 if opts_cookiefile is None:
1384 self.cookiejar = compat_cookiejar.CookieJar()
1386 self.cookiejar = compat_cookiejar.MozillaCookieJar(
1388 if os.access(opts_cookiefile, os.R_OK):
1389 self.cookiejar.load()
1391 cookie_processor = compat_urllib_request.HTTPCookieProcessor(
1393 if opts_proxy is not None:
1394 if opts_proxy == '':
1397 proxies = {'http': opts_proxy, 'https': opts_proxy}
1399 proxies = compat_urllib_request.getproxies()
1400 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
1401 if 'http' in proxies and 'https' not in proxies:
1402 proxies['https'] = proxies['http']
1403 proxy_handler = compat_urllib_request.ProxyHandler(proxies)
1405 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
1406 https_handler = make_HTTPS_handler(
1407 self.params.get('nocheckcertificate', False), debuglevel=debuglevel)
1408 ydlh = YoutubeDLHandler(debuglevel=debuglevel)
1409 opener = compat_urllib_request.build_opener(
1410 https_handler, proxy_handler, cookie_processor, ydlh)
1411 # Delete the default user-agent header, which would otherwise apply in
1412 # cases where our custom HTTP handler doesn't come into play
1413 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
1414 opener.addheaders = []
1415 self._opener = opener
1417 def encode(self, s):
1418 if isinstance(s, bytes):
1419 return s # Already encoded
1422 return s.encode(self.get_encoding())
1423 except UnicodeEncodeError as err:
1424 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
1427 def get_encoding(self):
1428 encoding = self.params.get('encoding')
1429 if encoding is None:
1430 encoding = preferredencoding()