2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import, unicode_literals
34 compat_get_terminal_size,
38 compat_tokenize_tokenize,
40 compat_urllib_request,
41 compat_urllib_request_DataHandler,
62 PerRequestProxyHandler,
73 UnavailableVideoError,
78 YoutubeDLCookieProcessor,
85 from .cache import Cache
86 from .extractor import get_info_extractor, gen_extractors
87 from .downloader import get_suitable_downloader
88 from .downloader.rtmp import rtmpdump_version
89 from .postprocessor import (
91 FFmpegFixupStretchedPP,
96 from .version import __version__
99 class YoutubeDL(object):
102 YoutubeDL objects are the ones responsible of downloading the
103 actual video file and writing it to disk if the user has requested
104 it, among some other tasks. In most cases there should be one per
105 program. As, given a video URL, the downloader doesn't know how to
106 extract all the needed information, task that InfoExtractors do, it
107 has to pass the URL to one of them.
109 For this, YoutubeDL objects have a method that allows
110 InfoExtractors to be registered in a given order. When it is passed
111 a URL, the YoutubeDL object handles it to the first InfoExtractor it
112 finds that reports being able to handle it. The InfoExtractor extracts
113 all the information about the video or videos the URL refers to, and
114 YoutubeDL process the extracted information, possibly using a File
115 Downloader to download the video.
117 YoutubeDL objects accept a lot of parameters. In order not to saturate
118 the object constructor with arguments, it receives a dictionary of
119 options instead. These options are available through the params
120 attribute for the InfoExtractors to use. The YoutubeDL also
121 registers itself as the downloader in charge for the InfoExtractors
122 that are added to it, so this is a "mutual registration".
126 username: Username for authentication purposes.
127 password: Password for authentication purposes.
128 videopassword: Password for accessing a video.
129 usenetrc: Use netrc for authentication instead.
130 verbose: Print additional info to stdout.
131 quiet: Do not print messages to stdout.
132 no_warnings: Do not print out anything for warnings.
133 forceurl: Force printing final URL.
134 forcetitle: Force printing title.
135 forceid: Force printing ID.
136 forcethumbnail: Force printing thumbnail URL.
137 forcedescription: Force printing description.
138 forcefilename: Force printing final filename.
139 forceduration: Force printing duration.
140 forcejson: Force printing info_dict as JSON.
141 dump_single_json: Force printing the info_dict of the whole playlist
142 (or video) as a single JSON line.
143 simulate: Do not download the video files.
144 format: Video format code. See options.py for more information.
145 outtmpl: Template for output names.
146 restrictfilenames: Do not allow "&" and spaces in file names
147 ignoreerrors: Do not stop on download errors.
148 force_generic_extractor: Force downloader to use the generic extractor
149 nooverwrites: Prevent overwriting files.
150 playliststart: Playlist item to start at.
151 playlistend: Playlist item to end at.
152 playlist_items: Specific indices of playlist to download.
153 playlistreverse: Download playlist items in reverse order.
154 matchtitle: Download only matching titles.
155 rejecttitle: Reject downloads for matching titles.
156 logger: Log messages to a logging.Logger instance.
157 logtostderr: Log messages to stderr instead of stdout.
158 writedescription: Write the video description to a .description file
159 writeinfojson: Write the video description to a .info.json file
160 writeannotations: Write the video annotations to a .annotations.xml file
161 writethumbnail: Write the thumbnail image to a file
162 write_all_thumbnails: Write all thumbnail formats to files
163 writesubtitles: Write the video subtitles to a file
164 writeautomaticsub: Write the automatically generated subtitles to a file
165 allsubtitles: Downloads all the subtitles of the video
166 (requires writesubtitles or writeautomaticsub)
167 listsubtitles: Lists all available subtitles for the video
168 subtitlesformat: The format code for subtitles
169 subtitleslangs: List of languages of the subtitles to download
170 keepvideo: Keep the video file after post-processing
171 daterange: A DateRange object, download only if the upload_date is in the range.
172 skip_download: Skip the actual download of the video file
173 cachedir: Location of the cache files in the filesystem.
174 False to disable filesystem cache.
175 noplaylist: Download single video instead of a playlist if in doubt.
176 age_limit: An integer representing the user's age in years.
177 Unsuitable videos for the given age are skipped.
178 min_views: An integer representing the minimum view count the video
179 must have in order to not be skipped.
180 Videos without view count information are always
181 downloaded. None for no limit.
182 max_views: An integer representing the maximum view count.
183 Videos that are more popular than that are not
185 Videos without view count information are always
186 downloaded. None for no limit.
187 download_archive: File name of a file where all downloads are recorded.
188 Videos already present in the file are not downloaded
190 cookiefile: File name where cookies should be read from and dumped to.
191 nocheckcertificate:Do not verify SSL certificates
192 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
193 At the moment, this is only supported by YouTube.
194 proxy: URL of the proxy server to use
195 cn_verification_proxy: URL of the proxy to use for IP address verification
196 on Chinese sites. (Experimental)
197 socket_timeout: Time to wait for unresponsive hosts, in seconds
198 bidi_workaround: Work around buggy terminals without bidirectional text
199 support, using fridibi
200 debug_printtraffic:Print out sent and received HTTP traffic
201 include_ads: Download ads as well
202 default_search: Prepend this string if an input url is not valid.
203 'auto' for elaborate guessing
204 encoding: Use this encoding instead of the system-specified.
205 extract_flat: Do not resolve URLs, return the immediate result.
206 Pass in 'in_playlist' to only show this behavior for
208 postprocessors: A list of dictionaries, each with an entry
209 * key: The name of the postprocessor. See
210 youtube_dl/postprocessor/__init__.py for a list.
211 as well as any further keyword arguments for the
213 progress_hooks: A list of functions that get called on download
214 progress, with a dictionary with the entries
215 * status: One of "downloading", "error", or "finished".
216 Check this first and ignore unknown values.
218 If status is one of "downloading", or "finished", the
219 following properties may also be present:
220 * filename: The final filename (always present)
221 * tmpfilename: The filename we're currently writing to
222 * downloaded_bytes: Bytes on disk
223 * total_bytes: Size of the whole file, None if unknown
224 * total_bytes_estimate: Guess of the eventual file size,
226 * elapsed: The number of seconds since download started.
227 * eta: The estimated time in seconds, None if unknown
228 * speed: The download speed in bytes/second, None if
230 * fragment_index: The counter of the currently
231 downloaded video fragment.
232 * fragment_count: The number of fragments (= individual
233 files that will be merged)
235 Progress hooks are guaranteed to be called at least once
236 (with status "finished") if the download is successful.
237 merge_output_format: Extension to use when merging formats.
238 fixup: Automatically correct known faults of the file.
240 - "never": do nothing
241 - "warn": only emit a warning
242 - "detect_or_warn": check whether we can do anything
243 about it, warn otherwise (default)
244 source_address: (Experimental) Client-side IP address to bind to.
245 call_home: Boolean, true iff we are allowed to contact the
246 youtube-dl servers for debugging.
247 sleep_interval: Number of seconds to sleep before each download.
248 listformats: Print an overview of available video formats and exit.
249 list_thumbnails: Print a table of all thumbnails and exit.
250 match_filter: A function that gets called with the info_dict of
252 If it returns a message, the video is ignored.
253 If it returns None, the video is downloaded.
254 match_filter_func in utils.py is one example for this.
255 no_color: Do not emit color codes in output.
257 The following options determine which downloader is picked:
258 external_downloader: Executable of the external downloader to call.
259 None or unset for standard (built-in) downloader.
260 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv.
262 The following parameters are not used by YoutubeDL itself, they are used by
263 the downloader (see youtube_dl/downloader/common.py):
264 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
265 noresizebuffer, retries, continuedl, noprogress, consoletitle,
266 xattr_set_filesize, external_downloader_args, hls_use_mpegts.
268 The following options are used by the post processors:
269 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
270 otherwise prefer avconv.
271 postprocessor_args: A list of additional command-line arguments for the
278 _download_retcode = None
279 _num_downloads = None
282 def __init__(self, params=None, auto_init=True):
283 """Create a FileDownloader object with the given options."""
287 self._ies_instances = {}
289 self._progress_hooks = []
290 self._download_retcode = 0
291 self._num_downloads = 0
292 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
293 self._err_file = sys.stderr
296 'nocheckcertificate': False,
298 self.params.update(params)
299 self.cache = Cache(self)
301 if params.get('bidi_workaround', False):
304 master, slave = pty.openpty()
305 width = compat_get_terminal_size().columns
309 width_args = ['-w', str(width)]
311 stdin=subprocess.PIPE,
313 stderr=self._err_file)
315 self._output_process = subprocess.Popen(
316 ['bidiv'] + width_args, **sp_kwargs
319 self._output_process = subprocess.Popen(
320 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
321 self._output_channel = os.fdopen(master, 'rb')
322 except OSError as ose:
324 self.report_warning('Could not find fribidi executable, ignoring --bidi-workaround . Make sure that fribidi is an executable file in one of the directories in your $PATH.')
328 if (sys.version_info >= (3,) and sys.platform != 'win32' and
329 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
330 not params.get('restrictfilenames', False)):
331 # On Python 3, the Unicode filesystem API will throw errors (#1474)
333 'Assuming --restrict-filenames since file system encoding '
334 'cannot encode all characters. '
335 'Set the LC_ALL environment variable to fix this.')
336 self.params['restrictfilenames'] = True
338 if isinstance(params.get('outtmpl'), bytes):
340 'Parameter outtmpl is bytes, but should be a unicode string. '
341 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
346 self.print_debug_header()
347 self.add_default_info_extractors()
349 for pp_def_raw in self.params.get('postprocessors', []):
350 pp_class = get_postprocessor(pp_def_raw['key'])
351 pp_def = dict(pp_def_raw)
353 pp = pp_class(self, **compat_kwargs(pp_def))
354 self.add_post_processor(pp)
356 for ph in self.params.get('progress_hooks', []):
357 self.add_progress_hook(ph)
359 def warn_if_short_id(self, argv):
360 # short YouTube ID starting with dash?
362 i for i, a in enumerate(argv)
363 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
367 [a for i, a in enumerate(argv) if i not in idxs] +
368 ['--'] + [argv[i] for i in idxs]
371 'Long argument string detected. '
372 'Use -- to separate parameters and URLs, like this:\n%s\n' %
373 args_to_str(correct_argv))
375 def add_info_extractor(self, ie):
376 """Add an InfoExtractor object to the end of the list."""
378 self._ies_instances[ie.ie_key()] = ie
379 ie.set_downloader(self)
381 def get_info_extractor(self, ie_key):
383 Get an instance of an IE with name ie_key, it will try to get one from
384 the _ies list, if there's no instance it will create a new one and add
385 it to the extractor list.
387 ie = self._ies_instances.get(ie_key)
389 ie = get_info_extractor(ie_key)()
390 self.add_info_extractor(ie)
393 def add_default_info_extractors(self):
395 Add the InfoExtractors returned by gen_extractors to the end of the list
397 for ie in gen_extractors():
398 self.add_info_extractor(ie)
400 def add_post_processor(self, pp):
401 """Add a PostProcessor object to the end of the chain."""
403 pp.set_downloader(self)
405 def add_progress_hook(self, ph):
406 """Add the progress hook (currently only for the file downloader)"""
407 self._progress_hooks.append(ph)
409 def _bidi_workaround(self, message):
410 if not hasattr(self, '_output_channel'):
413 assert hasattr(self, '_output_process')
414 assert isinstance(message, compat_str)
415 line_count = message.count('\n') + 1
416 self._output_process.stdin.write((message + '\n').encode('utf-8'))
417 self._output_process.stdin.flush()
418 res = ''.join(self._output_channel.readline().decode('utf-8')
419 for _ in range(line_count))
420 return res[:-len('\n')]
422 def to_screen(self, message, skip_eol=False):
423 """Print message to stdout if not in quiet mode."""
424 return self.to_stdout(message, skip_eol, check_quiet=True)
426 def _write_string(self, s, out=None):
427 write_string(s, out=out, encoding=self.params.get('encoding'))
429 def to_stdout(self, message, skip_eol=False, check_quiet=False):
430 """Print message to stdout if not in quiet mode."""
431 if self.params.get('logger'):
432 self.params['logger'].debug(message)
433 elif not check_quiet or not self.params.get('quiet', False):
434 message = self._bidi_workaround(message)
435 terminator = ['\n', ''][skip_eol]
436 output = message + terminator
438 self._write_string(output, self._screen_file)
440 def to_stderr(self, message):
441 """Print message to stderr."""
442 assert isinstance(message, compat_str)
443 if self.params.get('logger'):
444 self.params['logger'].error(message)
446 message = self._bidi_workaround(message)
447 output = message + '\n'
448 self._write_string(output, self._err_file)
450 def to_console_title(self, message):
451 if not self.params.get('consoletitle', False):
453 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
454 # c_wchar_p() might not be necessary if `message` is
455 # already of type unicode()
456 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
457 elif 'TERM' in os.environ:
458 self._write_string('\033]0;%s\007' % message, self._screen_file)
460 def save_console_title(self):
461 if not self.params.get('consoletitle', False):
463 if 'TERM' in os.environ:
464 # Save the title on stack
465 self._write_string('\033[22;0t', self._screen_file)
467 def restore_console_title(self):
468 if not self.params.get('consoletitle', False):
470 if 'TERM' in os.environ:
471 # Restore the title from stack
472 self._write_string('\033[23;0t', self._screen_file)
475 self.save_console_title()
478 def __exit__(self, *args):
479 self.restore_console_title()
481 if self.params.get('cookiefile') is not None:
482 self.cookiejar.save()
484 def trouble(self, message=None, tb=None):
485 """Determine action to take when a download problem appears.
487 Depending on if the downloader has been configured to ignore
488 download errors or not, this method may throw an exception or
489 not when errors are found, after printing the message.
491 tb, if given, is additional traceback information.
493 if message is not None:
494 self.to_stderr(message)
495 if self.params.get('verbose'):
497 if sys.exc_info()[0]: # if .trouble has been called from an except block
499 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
500 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
501 tb += encode_compat_str(traceback.format_exc())
503 tb_data = traceback.format_list(traceback.extract_stack())
504 tb = ''.join(tb_data)
506 if not self.params.get('ignoreerrors', False):
507 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
508 exc_info = sys.exc_info()[1].exc_info
510 exc_info = sys.exc_info()
511 raise DownloadError(message, exc_info)
512 self._download_retcode = 1
514 def report_warning(self, message):
516 Print the message to stderr, it will be prefixed with 'WARNING:'
517 If stderr is a tty file the 'WARNING:' will be colored
519 if self.params.get('logger') is not None:
520 self.params['logger'].warning(message)
522 if self.params.get('no_warnings'):
524 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
525 _msg_header = '\033[0;33mWARNING:\033[0m'
527 _msg_header = 'WARNING:'
528 warning_message = '%s %s' % (_msg_header, message)
529 self.to_stderr(warning_message)
531 def report_error(self, message, tb=None):
533 Do the same as trouble, but prefixes the message with 'ERROR:', colored
534 in red if stderr is a tty file.
536 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
537 _msg_header = '\033[0;31mERROR:\033[0m'
539 _msg_header = 'ERROR:'
540 error_message = '%s %s' % (_msg_header, message)
541 self.trouble(error_message, tb)
543 def report_file_already_downloaded(self, file_name):
544 """Report file has already been fully downloaded."""
546 self.to_screen('[download] %s has already been downloaded' % file_name)
547 except UnicodeEncodeError:
548 self.to_screen('[download] The file has already been downloaded')
550 def prepare_filename(self, info_dict):
551 """Generate the output filename."""
553 template_dict = dict(info_dict)
555 template_dict['epoch'] = int(time.time())
556 autonumber_size = self.params.get('autonumber_size')
557 if autonumber_size is None:
559 autonumber_templ = '%0' + str(autonumber_size) + 'd'
560 template_dict['autonumber'] = autonumber_templ % self._num_downloads
561 if template_dict.get('playlist_index') is not None:
562 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
563 if template_dict.get('resolution') is None:
564 if template_dict.get('width') and template_dict.get('height'):
565 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
566 elif template_dict.get('height'):
567 template_dict['resolution'] = '%sp' % template_dict['height']
568 elif template_dict.get('width'):
569 template_dict['resolution'] = '?x%d' % template_dict['width']
571 sanitize = lambda k, v: sanitize_filename(
573 restricted=self.params.get('restrictfilenames'),
575 template_dict = dict((k, sanitize(k, v))
576 for k, v in template_dict.items()
578 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
580 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
581 tmpl = compat_expanduser(outtmpl)
582 filename = tmpl % template_dict
583 # Temporary fix for #4787
584 # 'Treat' all problem characters by passing filename through preferredencoding
585 # to workaround encoding issues with subprocess on python2 @ Windows
586 if sys.version_info < (3, 0) and sys.platform == 'win32':
587 filename = encodeFilename(filename, True).decode(preferredencoding())
588 return sanitize_path(filename)
589 except ValueError as err:
590 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
593 def _match_entry(self, info_dict, incomplete):
594 """ Returns None iff the file should be downloaded """
596 video_title = info_dict.get('title', info_dict.get('id', 'video'))
597 if 'title' in info_dict:
598 # This can happen when we're just evaluating the playlist
599 title = info_dict['title']
600 matchtitle = self.params.get('matchtitle', False)
602 if not re.search(matchtitle, title, re.IGNORECASE):
603 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
604 rejecttitle = self.params.get('rejecttitle', False)
606 if re.search(rejecttitle, title, re.IGNORECASE):
607 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
608 date = info_dict.get('upload_date', None)
610 dateRange = self.params.get('daterange', DateRange())
611 if date not in dateRange:
612 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
613 view_count = info_dict.get('view_count', None)
614 if view_count is not None:
615 min_views = self.params.get('min_views')
616 if min_views is not None and view_count < min_views:
617 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
618 max_views = self.params.get('max_views')
619 if max_views is not None and view_count > max_views:
620 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
621 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
622 return 'Skipping "%s" because it is age restricted' % video_title
623 if self.in_download_archive(info_dict):
624 return '%s has already been recorded in archive' % video_title
627 match_filter = self.params.get('match_filter')
628 if match_filter is not None:
629 ret = match_filter(info_dict)
636 def add_extra_info(info_dict, extra_info):
637 '''Set the keys from extra_info in info dict if they are missing'''
638 for key, value in extra_info.items():
639 info_dict.setdefault(key, value)
641 def extract_info(self, url, download=True, ie_key=None, extra_info={},
642 process=True, force_generic_extractor=False):
644 Returns a list with a dictionary for each video we find.
645 If 'download', also downloads the videos.
646 extra_info is a dict containing the extra values to add to each result
649 if not ie_key and force_generic_extractor:
653 ies = [self.get_info_extractor(ie_key)]
658 if not ie.suitable(url):
662 self.report_warning('The program functionality for this site has been marked as broken, '
663 'and will probably not work.')
666 ie_result = ie.extract(url)
667 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
669 if isinstance(ie_result, list):
670 # Backwards compatibility: old IE result format
672 '_type': 'compat_list',
673 'entries': ie_result,
675 self.add_default_extra_info(ie_result, ie, url)
677 return self.process_ie_result(ie_result, download, extra_info)
680 except ExtractorError as e: # An error we somewhat expected
681 self.report_error(compat_str(e), e.format_traceback())
683 except MaxDownloadsReached:
685 except Exception as e:
686 if self.params.get('ignoreerrors', False):
687 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
692 self.report_error('no suitable InfoExtractor for URL %s' % url)
694 def add_default_extra_info(self, ie_result, ie, url):
695 self.add_extra_info(ie_result, {
696 'extractor': ie.IE_NAME,
698 'webpage_url_basename': url_basename(url),
699 'extractor_key': ie.ie_key(),
702 def process_ie_result(self, ie_result, download=True, extra_info={}):
704 Take the result of the ie(may be modified) and resolve all unresolved
705 references (URLs, playlist items).
707 It will also download the videos if 'download'.
708 Returns the resolved ie_result.
710 result_type = ie_result.get('_type', 'video')
712 if result_type in ('url', 'url_transparent'):
713 extract_flat = self.params.get('extract_flat', False)
714 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
715 extract_flat is True):
716 if self.params.get('forcejson', False):
717 self.to_stdout(json.dumps(ie_result))
720 if result_type == 'video':
721 self.add_extra_info(ie_result, extra_info)
722 return self.process_video_result(ie_result, download=download)
723 elif result_type == 'url':
724 # We have to add extra_info to the results because it may be
725 # contained in a playlist
726 return self.extract_info(ie_result['url'],
728 ie_key=ie_result.get('ie_key'),
729 extra_info=extra_info)
730 elif result_type == 'url_transparent':
731 # Use the information from the embedding page
732 info = self.extract_info(
733 ie_result['url'], ie_key=ie_result.get('ie_key'),
734 extra_info=extra_info, download=False, process=False)
736 force_properties = dict(
737 (k, v) for k, v in ie_result.items() if v is not None)
738 for f in ('_type', 'url', 'ie_key'):
739 if f in force_properties:
740 del force_properties[f]
741 new_result = info.copy()
742 new_result.update(force_properties)
744 assert new_result.get('_type') != 'url_transparent'
746 return self.process_ie_result(
747 new_result, download=download, extra_info=extra_info)
748 elif result_type == 'playlist' or result_type == 'multi_video':
749 # We process each entry in the playlist
750 playlist = ie_result.get('title', None) or ie_result.get('id', None)
751 self.to_screen('[download] Downloading playlist: %s' % playlist)
753 playlist_results = []
755 playliststart = self.params.get('playliststart', 1) - 1
756 playlistend = self.params.get('playlistend', None)
757 # For backwards compatibility, interpret -1 as whole list
758 if playlistend == -1:
761 playlistitems_str = self.params.get('playlist_items', None)
763 if playlistitems_str is not None:
764 def iter_playlistitems(format):
765 for string_segment in format.split(','):
766 if '-' in string_segment:
767 start, end = string_segment.split('-')
768 for item in range(int(start), int(end) + 1):
771 yield int(string_segment)
772 playlistitems = iter_playlistitems(playlistitems_str)
774 ie_entries = ie_result['entries']
775 if isinstance(ie_entries, list):
776 n_all_entries = len(ie_entries)
779 ie_entries[i - 1] for i in playlistitems
780 if -n_all_entries <= i - 1 < n_all_entries]
782 entries = ie_entries[playliststart:playlistend]
783 n_entries = len(entries)
785 "[%s] playlist %s: Collected %d video ids (downloading %d of them)" %
786 (ie_result['extractor'], playlist, n_all_entries, n_entries))
787 elif isinstance(ie_entries, PagedList):
790 for item in playlistitems:
791 entries.extend(ie_entries.getslice(
795 entries = ie_entries.getslice(
796 playliststart, playlistend)
797 n_entries = len(entries)
799 "[%s] playlist %s: Downloading %d videos" %
800 (ie_result['extractor'], playlist, n_entries))
803 entry_list = list(ie_entries)
804 entries = [entry_list[i - 1] for i in playlistitems]
806 entries = list(itertools.islice(
807 ie_entries, playliststart, playlistend))
808 n_entries = len(entries)
810 "[%s] playlist %s: Downloading %d videos" %
811 (ie_result['extractor'], playlist, n_entries))
813 if self.params.get('playlistreverse', False):
814 entries = entries[::-1]
816 for i, entry in enumerate(entries, 1):
817 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
819 'n_entries': n_entries,
820 'playlist': playlist,
821 'playlist_id': ie_result.get('id'),
822 'playlist_title': ie_result.get('title'),
823 'playlist_index': i + playliststart,
824 'extractor': ie_result['extractor'],
825 'webpage_url': ie_result['webpage_url'],
826 'webpage_url_basename': url_basename(ie_result['webpage_url']),
827 'extractor_key': ie_result['extractor_key'],
830 reason = self._match_entry(entry, incomplete=True)
831 if reason is not None:
832 self.to_screen('[download] ' + reason)
835 entry_result = self.process_ie_result(entry,
838 playlist_results.append(entry_result)
839 ie_result['entries'] = playlist_results
840 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
842 elif result_type == 'compat_list':
844 'Extractor %s returned a compat_list result. '
845 'It needs to be updated.' % ie_result.get('extractor'))
851 'extractor': ie_result['extractor'],
852 'webpage_url': ie_result['webpage_url'],
853 'webpage_url_basename': url_basename(ie_result['webpage_url']),
854 'extractor_key': ie_result['extractor_key'],
858 ie_result['entries'] = [
859 self.process_ie_result(_fixup(r), download, extra_info)
860 for r in ie_result['entries']
864 raise Exception('Invalid result type: %s' % result_type)
866 def _build_format_filter(self, filter_spec):
867 " Returns a function to filter the formats according to the filter_spec "
877 operator_rex = re.compile(r'''(?x)\s*
878 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
879 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
880 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
882 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
883 m = operator_rex.search(filter_spec)
886 comparison_value = int(m.group('value'))
888 comparison_value = parse_filesize(m.group('value'))
889 if comparison_value is None:
890 comparison_value = parse_filesize(m.group('value') + 'B')
891 if comparison_value is None:
893 'Invalid value %r in format specification %r' % (
894 m.group('value'), filter_spec))
895 op = OPERATORS[m.group('op')]
901 '^=': lambda attr, value: attr.startswith(value),
902 '$=': lambda attr, value: attr.endswith(value),
903 '*=': lambda attr, value: value in attr,
905 str_operator_rex = re.compile(r'''(?x)
906 \s*(?P<key>ext|acodec|vcodec|container|protocol)
907 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
908 \s*(?P<value>[a-zA-Z0-9._-]+)
910 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
911 m = str_operator_rex.search(filter_spec)
913 comparison_value = m.group('value')
914 op = STR_OPERATORS[m.group('op')]
917 raise ValueError('Invalid filter specification %r' % filter_spec)
920 actual_value = f.get(m.group('key'))
921 if actual_value is None:
922 return m.group('none_inclusive')
923 return op(actual_value, comparison_value)
926 def build_format_selector(self, format_spec):
927 def syntax_error(note, start):
929 'Invalid format specification: '
930 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
931 return SyntaxError(message)
933 PICKFIRST = 'PICKFIRST'
937 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
939 def _parse_filter(tokens):
941 for type, string, start, _, _ in tokens:
942 if type == tokenize.OP and string == ']':
943 return ''.join(filter_parts)
945 filter_parts.append(string)
947 def _remove_unused_ops(tokens):
948 # Remove operators that we don't use and join them with the surrounding strings
949 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
950 ALLOWED_OPS = ('/', '+', ',', '(', ')')
951 last_string, last_start, last_end, last_line = None, None, None, None
952 for type, string, start, end, line in tokens:
953 if type == tokenize.OP and string == '[':
955 yield tokenize.NAME, last_string, last_start, last_end, last_line
957 yield type, string, start, end, line
958 # everything inside brackets will be handled by _parse_filter
959 for type, string, start, end, line in tokens:
960 yield type, string, start, end, line
961 if type == tokenize.OP and string == ']':
963 elif type == tokenize.OP and string in ALLOWED_OPS:
965 yield tokenize.NAME, last_string, last_start, last_end, last_line
967 yield type, string, start, end, line
968 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
974 last_string += string
976 yield tokenize.NAME, last_string, last_start, last_end, last_line
978 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
980 current_selector = None
981 for type, string, start, _, _ in tokens:
982 # ENCODING is only defined in python 3.x
983 if type == getattr(tokenize, 'ENCODING', None):
985 elif type in [tokenize.NAME, tokenize.NUMBER]:
986 current_selector = FormatSelector(SINGLE, string, [])
987 elif type == tokenize.OP:
990 # ')' will be handled by the parentheses group
991 tokens.restore_last_token()
993 elif inside_merge and string in ['/', ',']:
994 tokens.restore_last_token()
996 elif inside_choice and string == ',':
997 tokens.restore_last_token()
1000 if not current_selector:
1001 raise syntax_error('"," must follow a format selector', start)
1002 selectors.append(current_selector)
1003 current_selector = None
1005 if not current_selector:
1006 raise syntax_error('"/" must follow a format selector', start)
1007 first_choice = current_selector
1008 second_choice = _parse_format_selection(tokens, inside_choice=True)
1009 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1011 if not current_selector:
1012 current_selector = FormatSelector(SINGLE, 'best', [])
1013 format_filter = _parse_filter(tokens)
1014 current_selector.filters.append(format_filter)
1016 if current_selector:
1017 raise syntax_error('Unexpected "("', start)
1018 group = _parse_format_selection(tokens, inside_group=True)
1019 current_selector = FormatSelector(GROUP, group, [])
1021 video_selector = current_selector
1022 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1023 if not video_selector or not audio_selector:
1024 raise syntax_error('"+" must be between two format selectors', start)
1025 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1027 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1028 elif type == tokenize.ENDMARKER:
1030 if current_selector:
1031 selectors.append(current_selector)
1034 def _build_selector_function(selector):
1035 if isinstance(selector, list):
1036 fs = [_build_selector_function(s) for s in selector]
1038 def selector_function(formats):
1040 for format in f(formats):
1042 return selector_function
1043 elif selector.type == GROUP:
1044 selector_function = _build_selector_function(selector.selector)
1045 elif selector.type == PICKFIRST:
1046 fs = [_build_selector_function(s) for s in selector.selector]
1048 def selector_function(formats):
1050 picked_formats = list(f(formats))
1052 return picked_formats
1054 elif selector.type == SINGLE:
1055 format_spec = selector.selector
1057 def selector_function(formats):
1058 formats = list(formats)
1061 if format_spec == 'all':
1064 elif format_spec in ['best', 'worst', None]:
1065 format_idx = 0 if format_spec == 'worst' else -1
1066 audiovideo_formats = [
1068 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1069 if audiovideo_formats:
1070 yield audiovideo_formats[format_idx]
1071 # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format
1072 elif (all(f.get('acodec') != 'none' for f in formats) or
1073 all(f.get('vcodec') != 'none' for f in formats)):
1074 yield formats[format_idx]
1075 elif format_spec == 'bestaudio':
1078 if f.get('vcodec') == 'none']
1080 yield audio_formats[-1]
1081 elif format_spec == 'worstaudio':
1084 if f.get('vcodec') == 'none']
1086 yield audio_formats[0]
1087 elif format_spec == 'bestvideo':
1090 if f.get('acodec') == 'none']
1092 yield video_formats[-1]
1093 elif format_spec == 'worstvideo':
1096 if f.get('acodec') == 'none']
1098 yield video_formats[0]
1100 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1101 if format_spec in extensions:
1102 filter_f = lambda f: f['ext'] == format_spec
1104 filter_f = lambda f: f['format_id'] == format_spec
1105 matches = list(filter(filter_f, formats))
1108 elif selector.type == MERGE:
1109 def _merge(formats_info):
1110 format_1, format_2 = [f['format_id'] for f in formats_info]
1111 # The first format must contain the video and the
1113 if formats_info[0].get('vcodec') == 'none':
1114 self.report_error('The first format must '
1115 'contain the video, try using '
1116 '"-f %s+%s"' % (format_2, format_1))
1118 # Formats must be opposite (video+audio)
1119 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1121 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1122 % (format_1, format_2))
1125 formats_info[0]['ext']
1126 if self.params.get('merge_output_format') is None
1127 else self.params['merge_output_format'])
1129 'requested_formats': formats_info,
1130 'format': '%s+%s' % (formats_info[0].get('format'),
1131 formats_info[1].get('format')),
1132 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1133 formats_info[1].get('format_id')),
1134 'width': formats_info[0].get('width'),
1135 'height': formats_info[0].get('height'),
1136 'resolution': formats_info[0].get('resolution'),
1137 'fps': formats_info[0].get('fps'),
1138 'vcodec': formats_info[0].get('vcodec'),
1139 'vbr': formats_info[0].get('vbr'),
1140 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1141 'acodec': formats_info[1].get('acodec'),
1142 'abr': formats_info[1].get('abr'),
1145 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1147 def selector_function(formats):
1148 formats = list(formats)
1149 for pair in itertools.product(video_selector(formats), audio_selector(formats)):
1152 filters = [self._build_format_filter(f) for f in selector.filters]
1154 def final_selector(formats):
1155 for _filter in filters:
1156 formats = list(filter(_filter, formats))
1157 return selector_function(formats)
1158 return final_selector
1160 stream = io.BytesIO(format_spec.encode('utf-8'))
1162 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1163 except tokenize.TokenError:
1164 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1166 class TokenIterator(object):
1167 def __init__(self, tokens):
1168 self.tokens = tokens
1175 if self.counter >= len(self.tokens):
1176 raise StopIteration()
1177 value = self.tokens[self.counter]
1183 def restore_last_token(self):
1186 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1187 return _build_selector_function(parsed_selector)
1189 def _calc_headers(self, info_dict):
1190 res = std_headers.copy()
1192 add_headers = info_dict.get('http_headers')
1194 res.update(add_headers)
1196 cookies = self._calc_cookies(info_dict)
1198 res['Cookie'] = cookies
1202 def _calc_cookies(self, info_dict):
1203 pr = sanitized_Request(info_dict['url'])
1204 self.cookiejar.add_cookie_header(pr)
1205 return pr.get_header('Cookie')
1207 def process_video_result(self, info_dict, download=True):
1208 assert info_dict.get('_type', 'video') == 'video'
1210 if 'id' not in info_dict:
1211 raise ExtractorError('Missing "id" field in extractor result')
1212 if 'title' not in info_dict:
1213 raise ExtractorError('Missing "title" field in extractor result')
1215 if 'playlist' not in info_dict:
1216 # It isn't part of a playlist
1217 info_dict['playlist'] = None
1218 info_dict['playlist_index'] = None
1220 thumbnails = info_dict.get('thumbnails')
1221 if thumbnails is None:
1222 thumbnail = info_dict.get('thumbnail')
1224 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1226 thumbnails.sort(key=lambda t: (
1227 t.get('preference'), t.get('width'), t.get('height'),
1228 t.get('id'), t.get('url')))
1229 for i, t in enumerate(thumbnails):
1230 if t.get('width') and t.get('height'):
1231 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1232 if t.get('id') is None:
1235 if thumbnails and 'thumbnail' not in info_dict:
1236 info_dict['thumbnail'] = thumbnails[-1]['url']
1238 if 'display_id' not in info_dict and 'id' in info_dict:
1239 info_dict['display_id'] = info_dict['id']
1241 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1242 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1243 # see http://bugs.python.org/issue1646728)
1245 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1246 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1247 except (ValueError, OverflowError, OSError):
1250 # Auto generate title fields corresponding to the *_number fields when missing
1251 # in order to always have clean titles. This is very common for TV series.
1252 for field in ('chapter', 'season', 'episode'):
1253 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1254 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1256 subtitles = info_dict.get('subtitles')
1258 for _, subtitle in subtitles.items():
1259 for subtitle_format in subtitle:
1260 if 'ext' not in subtitle_format:
1261 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1263 if self.params.get('listsubtitles', False):
1264 if 'automatic_captions' in info_dict:
1265 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1266 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1268 info_dict['requested_subtitles'] = self.process_subtitles(
1269 info_dict['id'], subtitles,
1270 info_dict.get('automatic_captions'))
1272 # We now pick which formats have to be downloaded
1273 if info_dict.get('formats') is None:
1274 # There's only one format available
1275 formats = [info_dict]
1277 formats = info_dict['formats']
1280 raise ExtractorError('No video formats found!')
1284 # We check that all the formats have the format and format_id fields
1285 for i, format in enumerate(formats):
1286 if 'url' not in format:
1287 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1289 if format.get('format_id') is None:
1290 format['format_id'] = compat_str(i)
1291 format_id = format['format_id']
1292 if format_id not in formats_dict:
1293 formats_dict[format_id] = []
1294 formats_dict[format_id].append(format)
1296 # Make sure all formats have unique format_id
1297 for format_id, ambiguous_formats in formats_dict.items():
1298 if len(ambiguous_formats) > 1:
1299 for i, format in enumerate(ambiguous_formats):
1300 format['format_id'] = '%s-%d' % (format_id, i)
1302 for i, format in enumerate(formats):
1303 if format.get('format') is None:
1304 format['format'] = '{id} - {res}{note}'.format(
1305 id=format['format_id'],
1306 res=self.format_resolution(format),
1307 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1309 # Automatically determine file extension if missing
1310 if 'ext' not in format:
1311 format['ext'] = determine_ext(format['url']).lower()
1312 # Automatically determine protocol if missing (useful for format
1313 # selection purposes)
1314 if 'protocol' not in format:
1315 format['protocol'] = determine_protocol(format)
1316 # Add HTTP headers, so that external programs can use them from the
1318 full_format_info = info_dict.copy()
1319 full_format_info.update(format)
1320 format['http_headers'] = self._calc_headers(full_format_info)
1322 # TODO Central sorting goes here
1324 if formats[0] is not info_dict:
1325 # only set the 'formats' fields if the original info_dict list them
1326 # otherwise we end up with a circular reference, the first (and unique)
1327 # element in the 'formats' field in info_dict is info_dict itself,
1328 # which can't be exported to json
1329 info_dict['formats'] = formats
1330 if self.params.get('listformats'):
1331 self.list_formats(info_dict)
1333 if self.params.get('list_thumbnails'):
1334 self.list_thumbnails(info_dict)
1337 req_format = self.params.get('format')
1338 if req_format is None:
1339 req_format_list = []
1340 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1341 info_dict['extractor'] in ['youtube', 'ted'] and
1342 not info_dict.get('is_live')):
1343 merger = FFmpegMergerPP(self)
1344 if merger.available and merger.can_merge():
1345 req_format_list.append('bestvideo+bestaudio')
1346 req_format_list.append('best')
1347 req_format = '/'.join(req_format_list)
1348 format_selector = self.build_format_selector(req_format)
1349 formats_to_download = list(format_selector(formats))
1350 if not formats_to_download:
1351 raise ExtractorError('requested format not available',
1355 if len(formats_to_download) > 1:
1356 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1357 for format in formats_to_download:
1358 new_info = dict(info_dict)
1359 new_info.update(format)
1360 self.process_info(new_info)
1361 # We update the info dict with the best quality format (backwards compatibility)
1362 info_dict.update(formats_to_download[-1])
1365 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1366 """Select the requested subtitles and their format"""
1368 if normal_subtitles and self.params.get('writesubtitles'):
1369 available_subs.update(normal_subtitles)
1370 if automatic_captions and self.params.get('writeautomaticsub'):
1371 for lang, cap_info in automatic_captions.items():
1372 if lang not in available_subs:
1373 available_subs[lang] = cap_info
1375 if (not self.params.get('writesubtitles') and not
1376 self.params.get('writeautomaticsub') or not
1380 if self.params.get('allsubtitles', False):
1381 requested_langs = available_subs.keys()
1383 if self.params.get('subtitleslangs', False):
1384 requested_langs = self.params.get('subtitleslangs')
1385 elif 'en' in available_subs:
1386 requested_langs = ['en']
1388 requested_langs = [list(available_subs.keys())[0]]
1390 formats_query = self.params.get('subtitlesformat', 'best')
1391 formats_preference = formats_query.split('/') if formats_query else []
1393 for lang in requested_langs:
1394 formats = available_subs.get(lang)
1396 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1398 for ext in formats_preference:
1402 matches = list(filter(lambda f: f['ext'] == ext, formats))
1408 self.report_warning(
1409 'No subtitle format found matching "%s" for language %s, '
1410 'using %s' % (formats_query, lang, f['ext']))
1414 def process_info(self, info_dict):
1415 """Process a single resolved IE result."""
1417 assert info_dict.get('_type', 'video') == 'video'
1419 max_downloads = self.params.get('max_downloads')
1420 if max_downloads is not None:
1421 if self._num_downloads >= int(max_downloads):
1422 raise MaxDownloadsReached()
1424 info_dict['fulltitle'] = info_dict['title']
1425 if len(info_dict['title']) > 200:
1426 info_dict['title'] = info_dict['title'][:197] + '...'
1428 if 'format' not in info_dict:
1429 info_dict['format'] = info_dict['ext']
1431 reason = self._match_entry(info_dict, incomplete=False)
1432 if reason is not None:
1433 self.to_screen('[download] ' + reason)
1436 self._num_downloads += 1
1438 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1441 if self.params.get('forcetitle', False):
1442 self.to_stdout(info_dict['fulltitle'])
1443 if self.params.get('forceid', False):
1444 self.to_stdout(info_dict['id'])
1445 if self.params.get('forceurl', False):
1446 if info_dict.get('requested_formats') is not None:
1447 for f in info_dict['requested_formats']:
1448 self.to_stdout(f['url'] + f.get('play_path', ''))
1450 # For RTMP URLs, also include the playpath
1451 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1452 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1453 self.to_stdout(info_dict['thumbnail'])
1454 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1455 self.to_stdout(info_dict['description'])
1456 if self.params.get('forcefilename', False) and filename is not None:
1457 self.to_stdout(filename)
1458 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1459 self.to_stdout(formatSeconds(info_dict['duration']))
1460 if self.params.get('forceformat', False):
1461 self.to_stdout(info_dict['format'])
1462 if self.params.get('forcejson', False):
1463 self.to_stdout(json.dumps(info_dict))
1465 # Do nothing else if in simulate mode
1466 if self.params.get('simulate', False):
1469 if filename is None:
1473 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1474 if dn and not os.path.exists(dn):
1476 except (OSError, IOError) as err:
1477 self.report_error('unable to create directory ' + error_to_compat_str(err))
1480 if self.params.get('writedescription', False):
1481 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1482 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1483 self.to_screen('[info] Video description is already present')
1484 elif info_dict.get('description') is None:
1485 self.report_warning('There\'s no description to write.')
1488 self.to_screen('[info] Writing video description to: ' + descfn)
1489 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1490 descfile.write(info_dict['description'])
1491 except (OSError, IOError):
1492 self.report_error('Cannot write description file ' + descfn)
1495 if self.params.get('writeannotations', False):
1496 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1497 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1498 self.to_screen('[info] Video annotations are already present')
1501 self.to_screen('[info] Writing video annotations to: ' + annofn)
1502 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1503 annofile.write(info_dict['annotations'])
1504 except (KeyError, TypeError):
1505 self.report_warning('There are no annotations to write.')
1506 except (OSError, IOError):
1507 self.report_error('Cannot write annotations file: ' + annofn)
1510 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1511 self.params.get('writeautomaticsub')])
1513 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1514 # subtitles download errors are already managed as troubles in relevant IE
1515 # that way it will silently go on when used with unsupporting IE
1516 subtitles = info_dict['requested_subtitles']
1517 ie = self.get_info_extractor(info_dict['extractor_key'])
1518 for sub_lang, sub_info in subtitles.items():
1519 sub_format = sub_info['ext']
1520 if sub_info.get('data') is not None:
1521 sub_data = sub_info['data']
1524 sub_data = ie._download_webpage(
1525 sub_info['url'], info_dict['id'], note=False)
1526 except ExtractorError as err:
1527 self.report_warning('Unable to download subtitle for "%s": %s' %
1528 (sub_lang, error_to_compat_str(err.cause)))
1531 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1532 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1533 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1535 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1536 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1537 subfile.write(sub_data)
1538 except (OSError, IOError):
1539 self.report_error('Cannot write subtitles file ' + sub_filename)
1542 if self.params.get('writeinfojson', False):
1543 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1544 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1545 self.to_screen('[info] Video description metadata is already present')
1547 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1549 write_json_file(self.filter_requested_info(info_dict), infofn)
1550 except (OSError, IOError):
1551 self.report_error('Cannot write metadata to JSON file ' + infofn)
1554 self._write_thumbnails(info_dict, filename)
1556 if not self.params.get('skip_download', False):
1559 fd = get_suitable_downloader(info, self.params)(self, self.params)
1560 for ph in self._progress_hooks:
1561 fd.add_progress_hook(ph)
1562 if self.params.get('verbose'):
1563 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1564 return fd.download(name, info)
1566 if info_dict.get('requested_formats') is not None:
1569 merger = FFmpegMergerPP(self)
1570 if not merger.available:
1572 self.report_warning('You have requested multiple '
1573 'formats but ffmpeg or avconv are not installed.'
1574 ' The formats won\'t be merged.')
1576 postprocessors = [merger]
1578 def compatible_formats(formats):
1579 video, audio = formats
1581 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1582 if video_ext and audio_ext:
1584 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1587 for exts in COMPATIBLE_EXTS:
1588 if video_ext in exts and audio_ext in exts:
1590 # TODO: Check acodec/vcodec
1593 filename_real_ext = os.path.splitext(filename)[1][1:]
1595 os.path.splitext(filename)[0]
1596 if filename_real_ext == info_dict['ext']
1598 requested_formats = info_dict['requested_formats']
1599 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1600 info_dict['ext'] = 'mkv'
1601 self.report_warning(
1602 'Requested formats are incompatible for merge and will be merged into mkv.')
1603 # Ensure filename always has a correct extension for successful merge
1604 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1605 if os.path.exists(encodeFilename(filename)):
1607 '[download] %s has already been downloaded and '
1608 'merged' % filename)
1610 for f in requested_formats:
1611 new_info = dict(info_dict)
1613 fname = self.prepare_filename(new_info)
1614 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1615 downloaded.append(fname)
1616 partial_success = dl(fname, new_info)
1617 success = success and partial_success
1618 info_dict['__postprocessors'] = postprocessors
1619 info_dict['__files_to_merge'] = downloaded
1621 # Just a single file
1622 success = dl(filename, info_dict)
1623 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1624 self.report_error('unable to download video data: %s' % str(err))
1626 except (OSError, IOError) as err:
1627 raise UnavailableVideoError(err)
1628 except (ContentTooShortError, ) as err:
1629 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1634 fixup_policy = self.params.get('fixup')
1635 if fixup_policy is None:
1636 fixup_policy = 'detect_or_warn'
1638 stretched_ratio = info_dict.get('stretched_ratio')
1639 if stretched_ratio is not None and stretched_ratio != 1:
1640 if fixup_policy == 'warn':
1641 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1642 info_dict['id'], stretched_ratio))
1643 elif fixup_policy == 'detect_or_warn':
1644 stretched_pp = FFmpegFixupStretchedPP(self)
1645 if stretched_pp.available:
1646 info_dict.setdefault('__postprocessors', [])
1647 info_dict['__postprocessors'].append(stretched_pp)
1649 self.report_warning(
1650 '%s: Non-uniform pixel ratio (%s). Install ffmpeg or avconv to fix this automatically.' % (
1651 info_dict['id'], stretched_ratio))
1653 assert fixup_policy in ('ignore', 'never')
1655 if info_dict.get('requested_formats') is None and info_dict.get('container') == 'm4a_dash':
1656 if fixup_policy == 'warn':
1657 self.report_warning('%s: writing DASH m4a. Only some players support this container.' % (
1659 elif fixup_policy == 'detect_or_warn':
1660 fixup_pp = FFmpegFixupM4aPP(self)
1661 if fixup_pp.available:
1662 info_dict.setdefault('__postprocessors', [])
1663 info_dict['__postprocessors'].append(fixup_pp)
1665 self.report_warning(
1666 '%s: writing DASH m4a. Only some players support this container. Install ffmpeg or avconv to fix this automatically.' % (
1669 assert fixup_policy in ('ignore', 'never')
1672 self.post_process(filename, info_dict)
1673 except (PostProcessingError) as err:
1674 self.report_error('postprocessing: %s' % str(err))
1676 self.record_download_archive(info_dict)
1678 def download(self, url_list):
1679 """Download a given list of URLs."""
1680 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1681 if (len(url_list) > 1 and
1682 '%' not in outtmpl and
1683 self.params.get('max_downloads') != 1):
1684 raise SameFileError(outtmpl)
1686 for url in url_list:
1688 # It also downloads the videos
1689 res = self.extract_info(
1690 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1691 except UnavailableVideoError:
1692 self.report_error('unable to download video')
1693 except MaxDownloadsReached:
1694 self.to_screen('[info] Maximum number of downloaded files reached.')
1697 if self.params.get('dump_single_json', False):
1698 self.to_stdout(json.dumps(res))
1700 return self._download_retcode
1702 def download_with_info_file(self, info_filename):
1703 with contextlib.closing(fileinput.FileInput(
1704 [info_filename], mode='r',
1705 openhook=fileinput.hook_encoded('utf-8'))) as f:
1706 # FileInput doesn't have a read method, we can't call json.load
1707 info = self.filter_requested_info(json.loads('\n'.join(f)))
1709 self.process_ie_result(info, download=True)
1710 except DownloadError:
1711 webpage_url = info.get('webpage_url')
1712 if webpage_url is not None:
1713 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1714 return self.download([webpage_url])
1717 return self._download_retcode
1720 def filter_requested_info(info_dict):
1722 (k, v) for k, v in info_dict.items()
1723 if k not in ['requested_formats', 'requested_subtitles'])
1725 def post_process(self, filename, ie_info):
1726 """Run all the postprocessors on the given file."""
1727 info = dict(ie_info)
1728 info['filepath'] = filename
1730 if ie_info.get('__postprocessors') is not None:
1731 pps_chain.extend(ie_info['__postprocessors'])
1732 pps_chain.extend(self._pps)
1733 for pp in pps_chain:
1734 files_to_delete = []
1736 files_to_delete, info = pp.run(info)
1737 except PostProcessingError as e:
1738 self.report_error(e.msg)
1739 if files_to_delete and not self.params.get('keepvideo', False):
1740 for old_filename in files_to_delete:
1741 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1743 os.remove(encodeFilename(old_filename))
1744 except (IOError, OSError):
1745 self.report_warning('Unable to remove downloaded original file')
1747 def _make_archive_id(self, info_dict):
1748 # Future-proof against any change in case
1749 # and backwards compatibility with prior versions
1750 extractor = info_dict.get('extractor_key')
1751 if extractor is None:
1752 if 'id' in info_dict:
1753 extractor = info_dict.get('ie_key') # key in a playlist
1754 if extractor is None:
1755 return None # Incomplete video information
1756 return extractor.lower() + ' ' + info_dict['id']
1758 def in_download_archive(self, info_dict):
1759 fn = self.params.get('download_archive')
1763 vid_id = self._make_archive_id(info_dict)
1765 return False # Incomplete video information
1768 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1769 for line in archive_file:
1770 if line.strip() == vid_id:
1772 except IOError as ioe:
1773 if ioe.errno != errno.ENOENT:
1777 def record_download_archive(self, info_dict):
1778 fn = self.params.get('download_archive')
1781 vid_id = self._make_archive_id(info_dict)
1783 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1784 archive_file.write(vid_id + '\n')
1787 def format_resolution(format, default='unknown'):
1788 if format.get('vcodec') == 'none':
1790 if format.get('resolution') is not None:
1791 return format['resolution']
1792 if format.get('height') is not None:
1793 if format.get('width') is not None:
1794 res = '%sx%s' % (format['width'], format['height'])
1796 res = '%sp' % format['height']
1797 elif format.get('width') is not None:
1798 res = '?x%d' % format['width']
1803 def _format_note(self, fdict):
1805 if fdict.get('ext') in ['f4f', 'f4m']:
1806 res += '(unsupported) '
1807 if fdict.get('language'):
1810 res += '[%s]' % fdict['language']
1811 if fdict.get('format_note') is not None:
1812 res += fdict['format_note'] + ' '
1813 if fdict.get('tbr') is not None:
1814 res += '%4dk ' % fdict['tbr']
1815 if fdict.get('container') is not None:
1818 res += '%s container' % fdict['container']
1819 if (fdict.get('vcodec') is not None and
1820 fdict.get('vcodec') != 'none'):
1823 res += fdict['vcodec']
1824 if fdict.get('vbr') is not None:
1826 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1828 if fdict.get('vbr') is not None:
1829 res += '%4dk' % fdict['vbr']
1830 if fdict.get('fps') is not None:
1831 res += ', %sfps' % fdict['fps']
1832 if fdict.get('acodec') is not None:
1835 if fdict['acodec'] == 'none':
1838 res += '%-5s' % fdict['acodec']
1839 elif fdict.get('abr') is not None:
1843 if fdict.get('abr') is not None:
1844 res += '@%3dk' % fdict['abr']
1845 if fdict.get('asr') is not None:
1846 res += ' (%5dHz)' % fdict['asr']
1847 if fdict.get('filesize') is not None:
1850 res += format_bytes(fdict['filesize'])
1851 elif fdict.get('filesize_approx') is not None:
1854 res += '~' + format_bytes(fdict['filesize_approx'])
1857 def list_formats(self, info_dict):
1858 formats = info_dict.get('formats', [info_dict])
1860 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1862 if f.get('preference') is None or f['preference'] >= -1000]
1863 if len(formats) > 1:
1864 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1866 header_line = ['format code', 'extension', 'resolution', 'note']
1868 '[info] Available formats for %s:\n%s' %
1869 (info_dict['id'], render_table(header_line, table)))
1871 def list_thumbnails(self, info_dict):
1872 thumbnails = info_dict.get('thumbnails')
1874 tn_url = info_dict.get('thumbnail')
1876 thumbnails = [{'id': '0', 'url': tn_url}]
1879 '[info] No thumbnails present for %s' % info_dict['id'])
1883 '[info] Thumbnails for %s:' % info_dict['id'])
1884 self.to_screen(render_table(
1885 ['ID', 'width', 'height', 'URL'],
1886 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1888 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1890 self.to_screen('%s has no %s' % (video_id, name))
1893 'Available %s for %s:' % (name, video_id))
1894 self.to_screen(render_table(
1895 ['Language', 'formats'],
1896 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1897 for lang, formats in subtitles.items()]))
1899 def urlopen(self, req):
1900 """ Start an HTTP download """
1901 if isinstance(req, compat_basestring):
1902 req = sanitized_Request(req)
1903 return self._opener.open(req, timeout=self._socket_timeout)
1905 def print_debug_header(self):
1906 if not self.params.get('verbose'):
1909 if type('') is not compat_str:
1910 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1911 self.report_warning(
1912 'Your Python is broken! Update to a newer and supported version')
1914 stdout_encoding = getattr(
1915 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1917 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1918 locale.getpreferredencoding(),
1919 sys.getfilesystemencoding(),
1921 self.get_encoding()))
1922 write_string(encoding_str, encoding=None)
1924 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1926 sp = subprocess.Popen(
1927 ['git', 'rev-parse', '--short', 'HEAD'],
1928 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1929 cwd=os.path.dirname(os.path.abspath(__file__)))
1930 out, err = sp.communicate()
1931 out = out.decode().strip()
1932 if re.match('[0-9a-f]+', out):
1933 self._write_string('[debug] Git HEAD: ' + out + '\n')
1939 self._write_string('[debug] Python version %s - %s\n' % (
1940 platform.python_version(), platform_name()))
1942 exe_versions = FFmpegPostProcessor.get_versions(self)
1943 exe_versions['rtmpdump'] = rtmpdump_version()
1944 exe_str = ', '.join(
1946 for exe, v in sorted(exe_versions.items())
1951 self._write_string('[debug] exe versions: %s\n' % exe_str)
1954 for handler in self._opener.handlers:
1955 if hasattr(handler, 'proxies'):
1956 proxy_map.update(handler.proxies)
1957 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1959 if self.params.get('call_home', False):
1960 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
1961 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
1962 latest_version = self.urlopen(
1963 'https://yt-dl.org/latest/version').read().decode('utf-8')
1964 if version_tuple(latest_version) > version_tuple(__version__):
1965 self.report_warning(
1966 'You are using an outdated version (newest version: %s)! '
1967 'See https://yt-dl.org/update if you need help updating.' %
1970 def _setup_opener(self):
1971 timeout_val = self.params.get('socket_timeout')
1972 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
1974 opts_cookiefile = self.params.get('cookiefile')
1975 opts_proxy = self.params.get('proxy')
1977 if opts_cookiefile is None:
1978 self.cookiejar = compat_cookiejar.CookieJar()
1980 self.cookiejar = compat_cookiejar.MozillaCookieJar(
1982 if os.access(opts_cookiefile, os.R_OK):
1983 self.cookiejar.load()
1985 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
1986 if opts_proxy is not None:
1987 if opts_proxy == '':
1990 proxies = {'http': opts_proxy, 'https': opts_proxy}
1992 proxies = compat_urllib_request.getproxies()
1993 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
1994 if 'http' in proxies and 'https' not in proxies:
1995 proxies['https'] = proxies['http']
1996 proxy_handler = PerRequestProxyHandler(proxies)
1998 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
1999 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2000 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2001 data_handler = compat_urllib_request_DataHandler()
2003 # When passing our own FileHandler instance, build_opener won't add the
2004 # default FileHandler and allows us to disable the file protocol, which
2005 # can be used for malicious purposes (see
2006 # https://github.com/rg3/youtube-dl/issues/8227)
2007 file_handler = compat_urllib_request.FileHandler()
2009 def file_open(*args, **kwargs):
2010 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2011 file_handler.file_open = file_open
2013 opener = compat_urllib_request.build_opener(
2014 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2016 # Delete the default user-agent header, which would otherwise apply in
2017 # cases where our custom HTTP handler doesn't come into play
2018 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2019 opener.addheaders = []
2020 self._opener = opener
2022 def encode(self, s):
2023 if isinstance(s, bytes):
2024 return s # Already encoded
2027 return s.encode(self.get_encoding())
2028 except UnicodeEncodeError as err:
2029 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2032 def get_encoding(self):
2033 encoding = self.params.get('encoding')
2034 if encoding is None:
2035 encoding = preferredencoding()
2038 def _write_thumbnails(self, info_dict, filename):
2039 if self.params.get('writethumbnail', False):
2040 thumbnails = info_dict.get('thumbnails')
2042 thumbnails = [thumbnails[-1]]
2043 elif self.params.get('write_all_thumbnails', False):
2044 thumbnails = info_dict.get('thumbnails')
2049 # No thumbnails present, so return immediately
2052 for t in thumbnails:
2053 thumb_ext = determine_ext(t['url'], 'jpg')
2054 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2055 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2056 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2058 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2059 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2060 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2062 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2063 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2065 uf = self.urlopen(t['url'])
2066 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2067 shutil.copyfileobj(uf, thumbf)
2068 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2069 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2070 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2071 self.report_warning('Unable to download thumbnail "%s": %s' %
2072 (t['url'], error_to_compat_str(err)))