2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import, unicode_literals
31 compat_get_terminal_size,
36 compat_tokenize_tokenize,
38 compat_urllib_request,
39 compat_urllib_request_DataHandler,
62 PerRequestProxyHandler,
67 register_socks_protocols,
77 UnavailableVideoError,
82 YoutubeDLCookieProcessor,
85 from .cache import Cache
86 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
87 from .downloader import get_suitable_downloader
88 from .downloader.rtmp import rtmpdump_version
89 from .postprocessor import (
92 FFmpegFixupStretchedPP,
97 from .version import __version__
99 if compat_os_name == 'nt':
103 class YoutubeDL(object):
106 YoutubeDL objects are the ones responsible of downloading the
107 actual video file and writing it to disk if the user has requested
108 it, among some other tasks. In most cases there should be one per
109 program. As, given a video URL, the downloader doesn't know how to
110 extract all the needed information, task that InfoExtractors do, it
111 has to pass the URL to one of them.
113 For this, YoutubeDL objects have a method that allows
114 InfoExtractors to be registered in a given order. When it is passed
115 a URL, the YoutubeDL object handles it to the first InfoExtractor it
116 finds that reports being able to handle it. The InfoExtractor extracts
117 all the information about the video or videos the URL refers to, and
118 YoutubeDL process the extracted information, possibly using a File
119 Downloader to download the video.
121 YoutubeDL objects accept a lot of parameters. In order not to saturate
122 the object constructor with arguments, it receives a dictionary of
123 options instead. These options are available through the params
124 attribute for the InfoExtractors to use. The YoutubeDL also
125 registers itself as the downloader in charge for the InfoExtractors
126 that are added to it, so this is a "mutual registration".
130 username: Username for authentication purposes.
131 password: Password for authentication purposes.
132 videopassword: Password for accessing a video.
133 usenetrc: Use netrc for authentication instead.
134 verbose: Print additional info to stdout.
135 quiet: Do not print messages to stdout.
136 no_warnings: Do not print out anything for warnings.
137 forceurl: Force printing final URL.
138 forcetitle: Force printing title.
139 forceid: Force printing ID.
140 forcethumbnail: Force printing thumbnail URL.
141 forcedescription: Force printing description.
142 forcefilename: Force printing final filename.
143 forceduration: Force printing duration.
144 forcejson: Force printing info_dict as JSON.
145 dump_single_json: Force printing the info_dict of the whole playlist
146 (or video) as a single JSON line.
147 simulate: Do not download the video files.
148 format: Video format code. See options.py for more information.
149 outtmpl: Template for output names.
150 restrictfilenames: Do not allow "&" and spaces in file names
151 ignoreerrors: Do not stop on download errors.
152 force_generic_extractor: Force downloader to use the generic extractor
153 nooverwrites: Prevent overwriting files.
154 playliststart: Playlist item to start at.
155 playlistend: Playlist item to end at.
156 playlist_items: Specific indices of playlist to download.
157 playlistreverse: Download playlist items in reverse order.
158 matchtitle: Download only matching titles.
159 rejecttitle: Reject downloads for matching titles.
160 logger: Log messages to a logging.Logger instance.
161 logtostderr: Log messages to stderr instead of stdout.
162 writedescription: Write the video description to a .description file
163 writeinfojson: Write the video description to a .info.json file
164 writeannotations: Write the video annotations to a .annotations.xml file
165 writethumbnail: Write the thumbnail image to a file
166 write_all_thumbnails: Write all thumbnail formats to files
167 writesubtitles: Write the video subtitles to a file
168 writeautomaticsub: Write the automatically generated subtitles to a file
169 allsubtitles: Downloads all the subtitles of the video
170 (requires writesubtitles or writeautomaticsub)
171 listsubtitles: Lists all available subtitles for the video
172 subtitlesformat: The format code for subtitles
173 subtitleslangs: List of languages of the subtitles to download
174 keepvideo: Keep the video file after post-processing
175 daterange: A DateRange object, download only if the upload_date is in the range.
176 skip_download: Skip the actual download of the video file
177 cachedir: Location of the cache files in the filesystem.
178 False to disable filesystem cache.
179 noplaylist: Download single video instead of a playlist if in doubt.
180 age_limit: An integer representing the user's age in years.
181 Unsuitable videos for the given age are skipped.
182 min_views: An integer representing the minimum view count the video
183 must have in order to not be skipped.
184 Videos without view count information are always
185 downloaded. None for no limit.
186 max_views: An integer representing the maximum view count.
187 Videos that are more popular than that are not
189 Videos without view count information are always
190 downloaded. None for no limit.
191 download_archive: File name of a file where all downloads are recorded.
192 Videos already present in the file are not downloaded
194 cookiefile: File name where cookies should be read from and dumped to.
195 nocheckcertificate:Do not verify SSL certificates
196 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
197 At the moment, this is only supported by YouTube.
198 proxy: URL of the proxy server to use
199 cn_verification_proxy: URL of the proxy to use for IP address verification
200 on Chinese sites. (Experimental)
201 socket_timeout: Time to wait for unresponsive hosts, in seconds
202 bidi_workaround: Work around buggy terminals without bidirectional text
203 support, using fridibi
204 debug_printtraffic:Print out sent and received HTTP traffic
205 include_ads: Download ads as well
206 default_search: Prepend this string if an input url is not valid.
207 'auto' for elaborate guessing
208 encoding: Use this encoding instead of the system-specified.
209 extract_flat: Do not resolve URLs, return the immediate result.
210 Pass in 'in_playlist' to only show this behavior for
212 postprocessors: A list of dictionaries, each with an entry
213 * key: The name of the postprocessor. See
214 youtube_dl/postprocessor/__init__.py for a list.
215 as well as any further keyword arguments for the
217 progress_hooks: A list of functions that get called on download
218 progress, with a dictionary with the entries
219 * status: One of "downloading", "error", or "finished".
220 Check this first and ignore unknown values.
222 If status is one of "downloading", or "finished", the
223 following properties may also be present:
224 * filename: The final filename (always present)
225 * tmpfilename: The filename we're currently writing to
226 * downloaded_bytes: Bytes on disk
227 * total_bytes: Size of the whole file, None if unknown
228 * total_bytes_estimate: Guess of the eventual file size,
230 * elapsed: The number of seconds since download started.
231 * eta: The estimated time in seconds, None if unknown
232 * speed: The download speed in bytes/second, None if
234 * fragment_index: The counter of the currently
235 downloaded video fragment.
236 * fragment_count: The number of fragments (= individual
237 files that will be merged)
239 Progress hooks are guaranteed to be called at least once
240 (with status "finished") if the download is successful.
241 merge_output_format: Extension to use when merging formats.
242 fixup: Automatically correct known faults of the file.
244 - "never": do nothing
245 - "warn": only emit a warning
246 - "detect_or_warn": check whether we can do anything
247 about it, warn otherwise (default)
248 source_address: (Experimental) Client-side IP address to bind to.
249 call_home: Boolean, true iff we are allowed to contact the
250 youtube-dl servers for debugging.
251 sleep_interval: Number of seconds to sleep before each download.
252 listformats: Print an overview of available video formats and exit.
253 list_thumbnails: Print a table of all thumbnails and exit.
254 match_filter: A function that gets called with the info_dict of
256 If it returns a message, the video is ignored.
257 If it returns None, the video is downloaded.
258 match_filter_func in utils.py is one example for this.
259 no_color: Do not emit color codes in output.
261 The following options determine which downloader is picked:
262 external_downloader: Executable of the external downloader to call.
263 None or unset for standard (built-in) downloader.
264 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
265 if True, otherwise use ffmpeg/avconv if False, otherwise
266 use downloader suggested by extractor if None.
268 The following parameters are not used by YoutubeDL itself, they are used by
269 the downloader (see youtube_dl/downloader/common.py):
270 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
271 noresizebuffer, retries, continuedl, noprogress, consoletitle,
272 xattr_set_filesize, external_downloader_args, hls_use_mpegts.
274 The following options are used by the post processors:
275 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
276 otherwise prefer avconv.
277 postprocessor_args: A list of additional command-line arguments for the
284 _download_retcode = None
285 _num_downloads = None
288 def __init__(self, params=None, auto_init=True):
289 """Create a FileDownloader object with the given options."""
293 self._ies_instances = {}
295 self._progress_hooks = []
296 self._download_retcode = 0
297 self._num_downloads = 0
298 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
299 self._err_file = sys.stderr
302 'nocheckcertificate': False,
304 self.params.update(params)
305 self.cache = Cache(self)
307 if params.get('bidi_workaround', False):
310 master, slave = pty.openpty()
311 width = compat_get_terminal_size().columns
315 width_args = ['-w', str(width)]
317 stdin=subprocess.PIPE,
319 stderr=self._err_file)
321 self._output_process = subprocess.Popen(
322 ['bidiv'] + width_args, **sp_kwargs
325 self._output_process = subprocess.Popen(
326 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
327 self._output_channel = os.fdopen(master, 'rb')
328 except OSError as ose:
330 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 $PATH.')
334 if (sys.version_info >= (3,) and sys.platform != 'win32' and
335 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
336 not params.get('restrictfilenames', False)):
337 # On Python 3, the Unicode filesystem API will throw errors (#1474)
339 'Assuming --restrict-filenames since file system encoding '
340 'cannot encode all characters. '
341 'Set the LC_ALL environment variable to fix this.')
342 self.params['restrictfilenames'] = True
344 if isinstance(params.get('outtmpl'), bytes):
346 'Parameter outtmpl is bytes, but should be a unicode string. '
347 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
352 self.print_debug_header()
353 self.add_default_info_extractors()
355 for pp_def_raw in self.params.get('postprocessors', []):
356 pp_class = get_postprocessor(pp_def_raw['key'])
357 pp_def = dict(pp_def_raw)
359 pp = pp_class(self, **compat_kwargs(pp_def))
360 self.add_post_processor(pp)
362 for ph in self.params.get('progress_hooks', []):
363 self.add_progress_hook(ph)
365 register_socks_protocols()
367 def warn_if_short_id(self, argv):
368 # short YouTube ID starting with dash?
370 i for i, a in enumerate(argv)
371 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
375 [a for i, a in enumerate(argv) if i not in idxs] +
376 ['--'] + [argv[i] for i in idxs]
379 'Long argument string detected. '
380 'Use -- to separate parameters and URLs, like this:\n%s\n' %
381 args_to_str(correct_argv))
383 def add_info_extractor(self, ie):
384 """Add an InfoExtractor object to the end of the list."""
386 if not isinstance(ie, type):
387 self._ies_instances[ie.ie_key()] = ie
388 ie.set_downloader(self)
390 def get_info_extractor(self, ie_key):
392 Get an instance of an IE with name ie_key, it will try to get one from
393 the _ies list, if there's no instance it will create a new one and add
394 it to the extractor list.
396 ie = self._ies_instances.get(ie_key)
398 ie = get_info_extractor(ie_key)()
399 self.add_info_extractor(ie)
402 def add_default_info_extractors(self):
404 Add the InfoExtractors returned by gen_extractors to the end of the list
406 for ie in gen_extractor_classes():
407 self.add_info_extractor(ie)
409 def add_post_processor(self, pp):
410 """Add a PostProcessor object to the end of the chain."""
412 pp.set_downloader(self)
414 def add_progress_hook(self, ph):
415 """Add the progress hook (currently only for the file downloader)"""
416 self._progress_hooks.append(ph)
418 def _bidi_workaround(self, message):
419 if not hasattr(self, '_output_channel'):
422 assert hasattr(self, '_output_process')
423 assert isinstance(message, compat_str)
424 line_count = message.count('\n') + 1
425 self._output_process.stdin.write((message + '\n').encode('utf-8'))
426 self._output_process.stdin.flush()
427 res = ''.join(self._output_channel.readline().decode('utf-8')
428 for _ in range(line_count))
429 return res[:-len('\n')]
431 def to_screen(self, message, skip_eol=False):
432 """Print message to stdout if not in quiet mode."""
433 return self.to_stdout(message, skip_eol, check_quiet=True)
435 def _write_string(self, s, out=None):
436 write_string(s, out=out, encoding=self.params.get('encoding'))
438 def to_stdout(self, message, skip_eol=False, check_quiet=False):
439 """Print message to stdout if not in quiet mode."""
440 if self.params.get('logger'):
441 self.params['logger'].debug(message)
442 elif not check_quiet or not self.params.get('quiet', False):
443 message = self._bidi_workaround(message)
444 terminator = ['\n', ''][skip_eol]
445 output = message + terminator
447 self._write_string(output, self._screen_file)
449 def to_stderr(self, message):
450 """Print message to stderr."""
451 assert isinstance(message, compat_str)
452 if self.params.get('logger'):
453 self.params['logger'].error(message)
455 message = self._bidi_workaround(message)
456 output = message + '\n'
457 self._write_string(output, self._err_file)
459 def to_console_title(self, message):
460 if not self.params.get('consoletitle', False):
462 if compat_os_name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
463 # c_wchar_p() might not be necessary if `message` is
464 # already of type unicode()
465 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
466 elif 'TERM' in os.environ:
467 self._write_string('\033]0;%s\007' % message, self._screen_file)
469 def save_console_title(self):
470 if not self.params.get('consoletitle', False):
472 if 'TERM' in os.environ:
473 # Save the title on stack
474 self._write_string('\033[22;0t', self._screen_file)
476 def restore_console_title(self):
477 if not self.params.get('consoletitle', False):
479 if 'TERM' in os.environ:
480 # Restore the title from stack
481 self._write_string('\033[23;0t', self._screen_file)
484 self.save_console_title()
487 def __exit__(self, *args):
488 self.restore_console_title()
490 if self.params.get('cookiefile') is not None:
491 self.cookiejar.save()
493 def trouble(self, message=None, tb=None):
494 """Determine action to take when a download problem appears.
496 Depending on if the downloader has been configured to ignore
497 download errors or not, this method may throw an exception or
498 not when errors are found, after printing the message.
500 tb, if given, is additional traceback information.
502 if message is not None:
503 self.to_stderr(message)
504 if self.params.get('verbose'):
506 if sys.exc_info()[0]: # if .trouble has been called from an except block
508 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
509 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
510 tb += encode_compat_str(traceback.format_exc())
512 tb_data = traceback.format_list(traceback.extract_stack())
513 tb = ''.join(tb_data)
515 if not self.params.get('ignoreerrors', False):
516 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
517 exc_info = sys.exc_info()[1].exc_info
519 exc_info = sys.exc_info()
520 raise DownloadError(message, exc_info)
521 self._download_retcode = 1
523 def report_warning(self, message):
525 Print the message to stderr, it will be prefixed with 'WARNING:'
526 If stderr is a tty file the 'WARNING:' will be colored
528 if self.params.get('logger') is not None:
529 self.params['logger'].warning(message)
531 if self.params.get('no_warnings'):
533 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
534 _msg_header = '\033[0;33mWARNING:\033[0m'
536 _msg_header = 'WARNING:'
537 warning_message = '%s %s' % (_msg_header, message)
538 self.to_stderr(warning_message)
540 def report_error(self, message, tb=None):
542 Do the same as trouble, but prefixes the message with 'ERROR:', colored
543 in red if stderr is a tty file.
545 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
546 _msg_header = '\033[0;31mERROR:\033[0m'
548 _msg_header = 'ERROR:'
549 error_message = '%s %s' % (_msg_header, message)
550 self.trouble(error_message, tb)
552 def report_file_already_downloaded(self, file_name):
553 """Report file has already been fully downloaded."""
555 self.to_screen('[download] %s has already been downloaded' % file_name)
556 except UnicodeEncodeError:
557 self.to_screen('[download] The file has already been downloaded')
559 def prepare_filename(self, info_dict):
560 """Generate the output filename."""
562 template_dict = dict(info_dict)
564 template_dict['epoch'] = int(time.time())
565 autonumber_size = self.params.get('autonumber_size')
566 if autonumber_size is None:
568 autonumber_templ = '%0' + str(autonumber_size) + 'd'
569 template_dict['autonumber'] = autonumber_templ % self._num_downloads
570 if template_dict.get('playlist_index') is not None:
571 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
572 if template_dict.get('resolution') is None:
573 if template_dict.get('width') and template_dict.get('height'):
574 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
575 elif template_dict.get('height'):
576 template_dict['resolution'] = '%sp' % template_dict['height']
577 elif template_dict.get('width'):
578 template_dict['resolution'] = '%dx?' % template_dict['width']
580 sanitize = lambda k, v: sanitize_filename(
582 restricted=self.params.get('restrictfilenames'),
584 template_dict = dict((k, sanitize(k, v))
585 for k, v in template_dict.items()
586 if v is not None and not isinstance(v, (list, tuple, dict)))
587 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
589 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
590 tmpl = compat_expanduser(outtmpl)
591 filename = tmpl % template_dict
592 # Temporary fix for #4787
593 # 'Treat' all problem characters by passing filename through preferredencoding
594 # to workaround encoding issues with subprocess on python2 @ Windows
595 if sys.version_info < (3, 0) and sys.platform == 'win32':
596 filename = encodeFilename(filename, True).decode(preferredencoding())
597 return sanitize_path(filename)
598 except ValueError as err:
599 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
602 def _match_entry(self, info_dict, incomplete):
603 """ Returns None iff the file should be downloaded """
605 video_title = info_dict.get('title', info_dict.get('id', 'video'))
606 if 'title' in info_dict:
607 # This can happen when we're just evaluating the playlist
608 title = info_dict['title']
609 matchtitle = self.params.get('matchtitle', False)
611 if not re.search(matchtitle, title, re.IGNORECASE):
612 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
613 rejecttitle = self.params.get('rejecttitle', False)
615 if re.search(rejecttitle, title, re.IGNORECASE):
616 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
617 date = info_dict.get('upload_date')
619 dateRange = self.params.get('daterange', DateRange())
620 if date not in dateRange:
621 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
622 view_count = info_dict.get('view_count')
623 if view_count is not None:
624 min_views = self.params.get('min_views')
625 if min_views is not None and view_count < min_views:
626 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
627 max_views = self.params.get('max_views')
628 if max_views is not None and view_count > max_views:
629 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
630 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
631 return 'Skipping "%s" because it is age restricted' % video_title
632 if self.in_download_archive(info_dict):
633 return '%s has already been recorded in archive' % video_title
636 match_filter = self.params.get('match_filter')
637 if match_filter is not None:
638 ret = match_filter(info_dict)
645 def add_extra_info(info_dict, extra_info):
646 '''Set the keys from extra_info in info dict if they are missing'''
647 for key, value in extra_info.items():
648 info_dict.setdefault(key, value)
650 def extract_info(self, url, download=True, ie_key=None, extra_info={},
651 process=True, force_generic_extractor=False):
653 Returns a list with a dictionary for each video we find.
654 If 'download', also downloads the videos.
655 extra_info is a dict containing the extra values to add to each result
658 if not ie_key and force_generic_extractor:
662 ies = [self.get_info_extractor(ie_key)]
667 if not ie.suitable(url):
670 ie = self.get_info_extractor(ie.ie_key())
672 self.report_warning('The program functionality for this site has been marked as broken, '
673 'and will probably not work.')
676 ie_result = ie.extract(url)
677 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
679 if isinstance(ie_result, list):
680 # Backwards compatibility: old IE result format
682 '_type': 'compat_list',
683 'entries': ie_result,
685 self.add_default_extra_info(ie_result, ie, url)
687 return self.process_ie_result(ie_result, download, extra_info)
690 except ExtractorError as e: # An error we somewhat expected
691 self.report_error(compat_str(e), e.format_traceback())
693 except MaxDownloadsReached:
695 except Exception as e:
696 if self.params.get('ignoreerrors', False):
697 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
702 self.report_error('no suitable InfoExtractor for URL %s' % url)
704 def add_default_extra_info(self, ie_result, ie, url):
705 self.add_extra_info(ie_result, {
706 'extractor': ie.IE_NAME,
708 'webpage_url_basename': url_basename(url),
709 'extractor_key': ie.ie_key(),
712 def process_ie_result(self, ie_result, download=True, extra_info={}):
714 Take the result of the ie(may be modified) and resolve all unresolved
715 references (URLs, playlist items).
717 It will also download the videos if 'download'.
718 Returns the resolved ie_result.
720 result_type = ie_result.get('_type', 'video')
722 if result_type in ('url', 'url_transparent'):
723 extract_flat = self.params.get('extract_flat', False)
724 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
725 extract_flat is True):
726 if self.params.get('forcejson', False):
727 self.to_stdout(json.dumps(ie_result))
730 if result_type == 'video':
731 self.add_extra_info(ie_result, extra_info)
732 return self.process_video_result(ie_result, download=download)
733 elif result_type == 'url':
734 # We have to add extra_info to the results because it may be
735 # contained in a playlist
736 return self.extract_info(ie_result['url'],
738 ie_key=ie_result.get('ie_key'),
739 extra_info=extra_info)
740 elif result_type == 'url_transparent':
741 # Use the information from the embedding page
742 info = self.extract_info(
743 ie_result['url'], ie_key=ie_result.get('ie_key'),
744 extra_info=extra_info, download=False, process=False)
746 force_properties = dict(
747 (k, v) for k, v in ie_result.items() if v is not None)
748 for f in ('_type', 'url', 'ie_key'):
749 if f in force_properties:
750 del force_properties[f]
751 new_result = info.copy()
752 new_result.update(force_properties)
754 assert new_result.get('_type') != 'url_transparent'
756 return self.process_ie_result(
757 new_result, download=download, extra_info=extra_info)
758 elif result_type == 'playlist' or result_type == 'multi_video':
759 # We process each entry in the playlist
760 playlist = ie_result.get('title') or ie_result.get('id')
761 self.to_screen('[download] Downloading playlist: %s' % playlist)
763 playlist_results = []
765 playliststart = self.params.get('playliststart', 1) - 1
766 playlistend = self.params.get('playlistend')
767 # For backwards compatibility, interpret -1 as whole list
768 if playlistend == -1:
771 playlistitems_str = self.params.get('playlist_items')
773 if playlistitems_str is not None:
774 def iter_playlistitems(format):
775 for string_segment in format.split(','):
776 if '-' in string_segment:
777 start, end = string_segment.split('-')
778 for item in range(int(start), int(end) + 1):
781 yield int(string_segment)
782 playlistitems = iter_playlistitems(playlistitems_str)
784 ie_entries = ie_result['entries']
785 if isinstance(ie_entries, list):
786 n_all_entries = len(ie_entries)
789 ie_entries[i - 1] for i in playlistitems
790 if -n_all_entries <= i - 1 < n_all_entries]
792 entries = ie_entries[playliststart:playlistend]
793 n_entries = len(entries)
795 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
796 (ie_result['extractor'], playlist, n_all_entries, n_entries))
797 elif isinstance(ie_entries, PagedList):
800 for item in playlistitems:
801 entries.extend(ie_entries.getslice(
805 entries = ie_entries.getslice(
806 playliststart, playlistend)
807 n_entries = len(entries)
809 '[%s] playlist %s: Downloading %d videos' %
810 (ie_result['extractor'], playlist, n_entries))
813 entry_list = list(ie_entries)
814 entries = [entry_list[i - 1] for i in playlistitems]
816 entries = list(itertools.islice(
817 ie_entries, playliststart, playlistend))
818 n_entries = len(entries)
820 '[%s] playlist %s: Downloading %d videos' %
821 (ie_result['extractor'], playlist, n_entries))
823 if self.params.get('playlistreverse', False):
824 entries = entries[::-1]
826 for i, entry in enumerate(entries, 1):
827 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
829 'n_entries': n_entries,
830 'playlist': playlist,
831 'playlist_id': ie_result.get('id'),
832 'playlist_title': ie_result.get('title'),
833 'playlist_index': i + playliststart,
834 'extractor': ie_result['extractor'],
835 'webpage_url': ie_result['webpage_url'],
836 'webpage_url_basename': url_basename(ie_result['webpage_url']),
837 'extractor_key': ie_result['extractor_key'],
840 reason = self._match_entry(entry, incomplete=True)
841 if reason is not None:
842 self.to_screen('[download] ' + reason)
845 entry_result = self.process_ie_result(entry,
848 playlist_results.append(entry_result)
849 ie_result['entries'] = playlist_results
850 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
852 elif result_type == 'compat_list':
854 'Extractor %s returned a compat_list result. '
855 'It needs to be updated.' % ie_result.get('extractor'))
861 'extractor': ie_result['extractor'],
862 'webpage_url': ie_result['webpage_url'],
863 'webpage_url_basename': url_basename(ie_result['webpage_url']),
864 'extractor_key': ie_result['extractor_key'],
868 ie_result['entries'] = [
869 self.process_ie_result(_fixup(r), download, extra_info)
870 for r in ie_result['entries']
874 raise Exception('Invalid result type: %s' % result_type)
876 def _build_format_filter(self, filter_spec):
877 " Returns a function to filter the formats according to the filter_spec "
887 operator_rex = re.compile(r'''(?x)\s*
888 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
889 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
890 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
892 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
893 m = operator_rex.search(filter_spec)
896 comparison_value = int(m.group('value'))
898 comparison_value = parse_filesize(m.group('value'))
899 if comparison_value is None:
900 comparison_value = parse_filesize(m.group('value') + 'B')
901 if comparison_value is None:
903 'Invalid value %r in format specification %r' % (
904 m.group('value'), filter_spec))
905 op = OPERATORS[m.group('op')]
911 '^=': lambda attr, value: attr.startswith(value),
912 '$=': lambda attr, value: attr.endswith(value),
913 '*=': lambda attr, value: value in attr,
915 str_operator_rex = re.compile(r'''(?x)
916 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
917 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
918 \s*(?P<value>[a-zA-Z0-9._-]+)
920 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
921 m = str_operator_rex.search(filter_spec)
923 comparison_value = m.group('value')
924 op = STR_OPERATORS[m.group('op')]
927 raise ValueError('Invalid filter specification %r' % filter_spec)
930 actual_value = f.get(m.group('key'))
931 if actual_value is None:
932 return m.group('none_inclusive')
933 return op(actual_value, comparison_value)
936 def build_format_selector(self, format_spec):
937 def syntax_error(note, start):
939 'Invalid format specification: '
940 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
941 return SyntaxError(message)
943 PICKFIRST = 'PICKFIRST'
947 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
949 def _parse_filter(tokens):
951 for type, string, start, _, _ in tokens:
952 if type == tokenize.OP and string == ']':
953 return ''.join(filter_parts)
955 filter_parts.append(string)
957 def _remove_unused_ops(tokens):
958 # Remove operators that we don't use and join them with the surrounding strings
959 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
960 ALLOWED_OPS = ('/', '+', ',', '(', ')')
961 last_string, last_start, last_end, last_line = None, None, None, None
962 for type, string, start, end, line in tokens:
963 if type == tokenize.OP and string == '[':
965 yield tokenize.NAME, last_string, last_start, last_end, last_line
967 yield type, string, start, end, line
968 # everything inside brackets will be handled by _parse_filter
969 for type, string, start, end, line in tokens:
970 yield type, string, start, end, line
971 if type == tokenize.OP and string == ']':
973 elif type == tokenize.OP and string in ALLOWED_OPS:
975 yield tokenize.NAME, last_string, last_start, last_end, last_line
977 yield type, string, start, end, line
978 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
984 last_string += string
986 yield tokenize.NAME, last_string, last_start, last_end, last_line
988 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
990 current_selector = None
991 for type, string, start, _, _ in tokens:
992 # ENCODING is only defined in python 3.x
993 if type == getattr(tokenize, 'ENCODING', None):
995 elif type in [tokenize.NAME, tokenize.NUMBER]:
996 current_selector = FormatSelector(SINGLE, string, [])
997 elif type == tokenize.OP:
1000 # ')' will be handled by the parentheses group
1001 tokens.restore_last_token()
1003 elif inside_merge and string in ['/', ',']:
1004 tokens.restore_last_token()
1006 elif inside_choice and string == ',':
1007 tokens.restore_last_token()
1010 if not current_selector:
1011 raise syntax_error('"," must follow a format selector', start)
1012 selectors.append(current_selector)
1013 current_selector = None
1015 if not current_selector:
1016 raise syntax_error('"/" must follow a format selector', start)
1017 first_choice = current_selector
1018 second_choice = _parse_format_selection(tokens, inside_choice=True)
1019 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1021 if not current_selector:
1022 current_selector = FormatSelector(SINGLE, 'best', [])
1023 format_filter = _parse_filter(tokens)
1024 current_selector.filters.append(format_filter)
1026 if current_selector:
1027 raise syntax_error('Unexpected "("', start)
1028 group = _parse_format_selection(tokens, inside_group=True)
1029 current_selector = FormatSelector(GROUP, group, [])
1031 video_selector = current_selector
1032 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1033 if not video_selector or not audio_selector:
1034 raise syntax_error('"+" must be between two format selectors', start)
1035 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1037 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1038 elif type == tokenize.ENDMARKER:
1040 if current_selector:
1041 selectors.append(current_selector)
1044 def _build_selector_function(selector):
1045 if isinstance(selector, list):
1046 fs = [_build_selector_function(s) for s in selector]
1048 def selector_function(formats):
1050 for format in f(formats):
1052 return selector_function
1053 elif selector.type == GROUP:
1054 selector_function = _build_selector_function(selector.selector)
1055 elif selector.type == PICKFIRST:
1056 fs = [_build_selector_function(s) for s in selector.selector]
1058 def selector_function(formats):
1060 picked_formats = list(f(formats))
1062 return picked_formats
1064 elif selector.type == SINGLE:
1065 format_spec = selector.selector
1067 def selector_function(formats):
1068 formats = list(formats)
1071 if format_spec == 'all':
1074 elif format_spec in ['best', 'worst', None]:
1075 format_idx = 0 if format_spec == 'worst' else -1
1076 audiovideo_formats = [
1078 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1079 if audiovideo_formats:
1080 yield audiovideo_formats[format_idx]
1081 # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format
1082 elif (all(f.get('acodec') != 'none' for f in formats) or
1083 all(f.get('vcodec') != 'none' for f in formats)):
1084 yield formats[format_idx]
1085 elif format_spec == 'bestaudio':
1088 if f.get('vcodec') == 'none']
1090 yield audio_formats[-1]
1091 elif format_spec == 'worstaudio':
1094 if f.get('vcodec') == 'none']
1096 yield audio_formats[0]
1097 elif format_spec == 'bestvideo':
1100 if f.get('acodec') == 'none']
1102 yield video_formats[-1]
1103 elif format_spec == 'worstvideo':
1106 if f.get('acodec') == 'none']
1108 yield video_formats[0]
1110 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1111 if format_spec in extensions:
1112 filter_f = lambda f: f['ext'] == format_spec
1114 filter_f = lambda f: f['format_id'] == format_spec
1115 matches = list(filter(filter_f, formats))
1118 elif selector.type == MERGE:
1119 def _merge(formats_info):
1120 format_1, format_2 = [f['format_id'] for f in formats_info]
1121 # The first format must contain the video and the
1123 if formats_info[0].get('vcodec') == 'none':
1124 self.report_error('The first format must '
1125 'contain the video, try using '
1126 '"-f %s+%s"' % (format_2, format_1))
1128 # Formats must be opposite (video+audio)
1129 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1131 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1132 % (format_1, format_2))
1135 formats_info[0]['ext']
1136 if self.params.get('merge_output_format') is None
1137 else self.params['merge_output_format'])
1139 'requested_formats': formats_info,
1140 'format': '%s+%s' % (formats_info[0].get('format'),
1141 formats_info[1].get('format')),
1142 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1143 formats_info[1].get('format_id')),
1144 'width': formats_info[0].get('width'),
1145 'height': formats_info[0].get('height'),
1146 'resolution': formats_info[0].get('resolution'),
1147 'fps': formats_info[0].get('fps'),
1148 'vcodec': formats_info[0].get('vcodec'),
1149 'vbr': formats_info[0].get('vbr'),
1150 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1151 'acodec': formats_info[1].get('acodec'),
1152 'abr': formats_info[1].get('abr'),
1155 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1157 def selector_function(formats):
1158 formats = list(formats)
1159 for pair in itertools.product(video_selector(formats), audio_selector(formats)):
1162 filters = [self._build_format_filter(f) for f in selector.filters]
1164 def final_selector(formats):
1165 for _filter in filters:
1166 formats = list(filter(_filter, formats))
1167 return selector_function(formats)
1168 return final_selector
1170 stream = io.BytesIO(format_spec.encode('utf-8'))
1172 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1173 except tokenize.TokenError:
1174 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1176 class TokenIterator(object):
1177 def __init__(self, tokens):
1178 self.tokens = tokens
1185 if self.counter >= len(self.tokens):
1186 raise StopIteration()
1187 value = self.tokens[self.counter]
1193 def restore_last_token(self):
1196 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1197 return _build_selector_function(parsed_selector)
1199 def _calc_headers(self, info_dict):
1200 res = std_headers.copy()
1202 add_headers = info_dict.get('http_headers')
1204 res.update(add_headers)
1206 cookies = self._calc_cookies(info_dict)
1208 res['Cookie'] = cookies
1212 def _calc_cookies(self, info_dict):
1213 pr = sanitized_Request(info_dict['url'])
1214 self.cookiejar.add_cookie_header(pr)
1215 return pr.get_header('Cookie')
1217 def process_video_result(self, info_dict, download=True):
1218 assert info_dict.get('_type', 'video') == 'video'
1220 if 'id' not in info_dict:
1221 raise ExtractorError('Missing "id" field in extractor result')
1222 if 'title' not in info_dict:
1223 raise ExtractorError('Missing "title" field in extractor result')
1225 if 'playlist' not in info_dict:
1226 # It isn't part of a playlist
1227 info_dict['playlist'] = None
1228 info_dict['playlist_index'] = None
1230 thumbnails = info_dict.get('thumbnails')
1231 if thumbnails is None:
1232 thumbnail = info_dict.get('thumbnail')
1234 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1236 thumbnails.sort(key=lambda t: (
1237 t.get('preference'), t.get('width'), t.get('height'),
1238 t.get('id'), t.get('url')))
1239 for i, t in enumerate(thumbnails):
1240 t['url'] = sanitize_url(t['url'])
1241 if t.get('width') and t.get('height'):
1242 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1243 if t.get('id') is None:
1246 if self.params.get('list_thumbnails'):
1247 self.list_thumbnails(info_dict)
1250 thumbnail = info_dict.get('thumbnail')
1252 info_dict['thumbnail'] = sanitize_url(thumbnail)
1254 info_dict['thumbnail'] = thumbnails[-1]['url']
1256 if 'display_id' not in info_dict and 'id' in info_dict:
1257 info_dict['display_id'] = info_dict['id']
1259 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1260 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1261 # see http://bugs.python.org/issue1646728)
1263 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1264 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1265 except (ValueError, OverflowError, OSError):
1268 # Auto generate title fields corresponding to the *_number fields when missing
1269 # in order to always have clean titles. This is very common for TV series.
1270 for field in ('chapter', 'season', 'episode'):
1271 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1272 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1274 subtitles = info_dict.get('subtitles')
1276 for _, subtitle in subtitles.items():
1277 for subtitle_format in subtitle:
1278 if subtitle_format.get('url'):
1279 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1280 if 'ext' not in subtitle_format:
1281 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1283 if self.params.get('listsubtitles', False):
1284 if 'automatic_captions' in info_dict:
1285 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1286 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1288 info_dict['requested_subtitles'] = self.process_subtitles(
1289 info_dict['id'], subtitles,
1290 info_dict.get('automatic_captions'))
1292 # We now pick which formats have to be downloaded
1293 if info_dict.get('formats') is None:
1294 # There's only one format available
1295 formats = [info_dict]
1297 formats = info_dict['formats']
1300 raise ExtractorError('No video formats found!')
1304 # We check that all the formats have the format and format_id fields
1305 for i, format in enumerate(formats):
1306 if 'url' not in format:
1307 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1309 format['url'] = sanitize_url(format['url'])
1311 if format.get('format_id') is None:
1312 format['format_id'] = compat_str(i)
1314 # Sanitize format_id from characters used in format selector expression
1315 format['format_id'] = re.sub('[\s,/+\[\]()]', '_', format['format_id'])
1316 format_id = format['format_id']
1317 if format_id not in formats_dict:
1318 formats_dict[format_id] = []
1319 formats_dict[format_id].append(format)
1321 # Make sure all formats have unique format_id
1322 for format_id, ambiguous_formats in formats_dict.items():
1323 if len(ambiguous_formats) > 1:
1324 for i, format in enumerate(ambiguous_formats):
1325 format['format_id'] = '%s-%d' % (format_id, i)
1327 for i, format in enumerate(formats):
1328 if format.get('format') is None:
1329 format['format'] = '{id} - {res}{note}'.format(
1330 id=format['format_id'],
1331 res=self.format_resolution(format),
1332 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1334 # Automatically determine file extension if missing
1335 if 'ext' not in format:
1336 format['ext'] = determine_ext(format['url']).lower()
1337 # Automatically determine protocol if missing (useful for format
1338 # selection purposes)
1339 if 'protocol' not in format:
1340 format['protocol'] = determine_protocol(format)
1341 # Add HTTP headers, so that external programs can use them from the
1343 full_format_info = info_dict.copy()
1344 full_format_info.update(format)
1345 format['http_headers'] = self._calc_headers(full_format_info)
1347 # TODO Central sorting goes here
1349 if formats[0] is not info_dict:
1350 # only set the 'formats' fields if the original info_dict list them
1351 # otherwise we end up with a circular reference, the first (and unique)
1352 # element in the 'formats' field in info_dict is info_dict itself,
1353 # which can't be exported to json
1354 info_dict['formats'] = formats
1355 if self.params.get('listformats'):
1356 self.list_formats(info_dict)
1359 req_format = self.params.get('format')
1360 if req_format is None:
1361 req_format_list = []
1362 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1363 not info_dict.get('is_live')):
1364 merger = FFmpegMergerPP(self)
1365 if merger.available and merger.can_merge():
1366 req_format_list.append('bestvideo+bestaudio')
1367 req_format_list.append('best')
1368 req_format = '/'.join(req_format_list)
1369 format_selector = self.build_format_selector(req_format)
1370 formats_to_download = list(format_selector(formats))
1371 if not formats_to_download:
1372 raise ExtractorError('requested format not available',
1376 if len(formats_to_download) > 1:
1377 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1378 for format in formats_to_download:
1379 new_info = dict(info_dict)
1380 new_info.update(format)
1381 self.process_info(new_info)
1382 # We update the info dict with the best quality format (backwards compatibility)
1383 info_dict.update(formats_to_download[-1])
1386 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1387 """Select the requested subtitles and their format"""
1389 if normal_subtitles and self.params.get('writesubtitles'):
1390 available_subs.update(normal_subtitles)
1391 if automatic_captions and self.params.get('writeautomaticsub'):
1392 for lang, cap_info in automatic_captions.items():
1393 if lang not in available_subs:
1394 available_subs[lang] = cap_info
1396 if (not self.params.get('writesubtitles') and not
1397 self.params.get('writeautomaticsub') or not
1401 if self.params.get('allsubtitles', False):
1402 requested_langs = available_subs.keys()
1404 if self.params.get('subtitleslangs', False):
1405 requested_langs = self.params.get('subtitleslangs')
1406 elif 'en' in available_subs:
1407 requested_langs = ['en']
1409 requested_langs = [list(available_subs.keys())[0]]
1411 formats_query = self.params.get('subtitlesformat', 'best')
1412 formats_preference = formats_query.split('/') if formats_query else []
1414 for lang in requested_langs:
1415 formats = available_subs.get(lang)
1417 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1419 for ext in formats_preference:
1423 matches = list(filter(lambda f: f['ext'] == ext, formats))
1429 self.report_warning(
1430 'No subtitle format found matching "%s" for language %s, '
1431 'using %s' % (formats_query, lang, f['ext']))
1435 def process_info(self, info_dict):
1436 """Process a single resolved IE result."""
1438 assert info_dict.get('_type', 'video') == 'video'
1440 max_downloads = self.params.get('max_downloads')
1441 if max_downloads is not None:
1442 if self._num_downloads >= int(max_downloads):
1443 raise MaxDownloadsReached()
1445 info_dict['fulltitle'] = info_dict['title']
1446 if len(info_dict['title']) > 200:
1447 info_dict['title'] = info_dict['title'][:197] + '...'
1449 if 'format' not in info_dict:
1450 info_dict['format'] = info_dict['ext']
1452 reason = self._match_entry(info_dict, incomplete=False)
1453 if reason is not None:
1454 self.to_screen('[download] ' + reason)
1457 self._num_downloads += 1
1459 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1462 if self.params.get('forcetitle', False):
1463 self.to_stdout(info_dict['fulltitle'])
1464 if self.params.get('forceid', False):
1465 self.to_stdout(info_dict['id'])
1466 if self.params.get('forceurl', False):
1467 if info_dict.get('requested_formats') is not None:
1468 for f in info_dict['requested_formats']:
1469 self.to_stdout(f['url'] + f.get('play_path', ''))
1471 # For RTMP URLs, also include the playpath
1472 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1473 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1474 self.to_stdout(info_dict['thumbnail'])
1475 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1476 self.to_stdout(info_dict['description'])
1477 if self.params.get('forcefilename', False) and filename is not None:
1478 self.to_stdout(filename)
1479 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1480 self.to_stdout(formatSeconds(info_dict['duration']))
1481 if self.params.get('forceformat', False):
1482 self.to_stdout(info_dict['format'])
1483 if self.params.get('forcejson', False):
1484 self.to_stdout(json.dumps(info_dict))
1486 # Do nothing else if in simulate mode
1487 if self.params.get('simulate', False):
1490 if filename is None:
1494 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1495 if dn and not os.path.exists(dn):
1497 except (OSError, IOError) as err:
1498 self.report_error('unable to create directory ' + error_to_compat_str(err))
1501 if self.params.get('writedescription', False):
1502 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1503 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1504 self.to_screen('[info] Video description is already present')
1505 elif info_dict.get('description') is None:
1506 self.report_warning('There\'s no description to write.')
1509 self.to_screen('[info] Writing video description to: ' + descfn)
1510 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1511 descfile.write(info_dict['description'])
1512 except (OSError, IOError):
1513 self.report_error('Cannot write description file ' + descfn)
1516 if self.params.get('writeannotations', False):
1517 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1518 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1519 self.to_screen('[info] Video annotations are already present')
1522 self.to_screen('[info] Writing video annotations to: ' + annofn)
1523 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1524 annofile.write(info_dict['annotations'])
1525 except (KeyError, TypeError):
1526 self.report_warning('There are no annotations to write.')
1527 except (OSError, IOError):
1528 self.report_error('Cannot write annotations file: ' + annofn)
1531 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1532 self.params.get('writeautomaticsub')])
1534 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1535 # subtitles download errors are already managed as troubles in relevant IE
1536 # that way it will silently go on when used with unsupporting IE
1537 subtitles = info_dict['requested_subtitles']
1538 ie = self.get_info_extractor(info_dict['extractor_key'])
1539 for sub_lang, sub_info in subtitles.items():
1540 sub_format = sub_info['ext']
1541 if sub_info.get('data') is not None:
1542 sub_data = sub_info['data']
1545 sub_data = ie._download_webpage(
1546 sub_info['url'], info_dict['id'], note=False)
1547 except ExtractorError as err:
1548 self.report_warning('Unable to download subtitle for "%s": %s' %
1549 (sub_lang, error_to_compat_str(err.cause)))
1552 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1553 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1554 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1556 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1557 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1558 subfile.write(sub_data)
1559 except (OSError, IOError):
1560 self.report_error('Cannot write subtitles file ' + sub_filename)
1563 if self.params.get('writeinfojson', False):
1564 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1565 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1566 self.to_screen('[info] Video description metadata is already present')
1568 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1570 write_json_file(self.filter_requested_info(info_dict), infofn)
1571 except (OSError, IOError):
1572 self.report_error('Cannot write metadata to JSON file ' + infofn)
1575 self._write_thumbnails(info_dict, filename)
1577 if not self.params.get('skip_download', False):
1580 fd = get_suitable_downloader(info, self.params)(self, self.params)
1581 for ph in self._progress_hooks:
1582 fd.add_progress_hook(ph)
1583 if self.params.get('verbose'):
1584 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1585 return fd.download(name, info)
1587 if info_dict.get('requested_formats') is not None:
1590 merger = FFmpegMergerPP(self)
1591 if not merger.available:
1593 self.report_warning('You have requested multiple '
1594 'formats but ffmpeg or avconv are not installed.'
1595 ' The formats won\'t be merged.')
1597 postprocessors = [merger]
1599 def compatible_formats(formats):
1600 video, audio = formats
1602 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1603 if video_ext and audio_ext:
1605 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1608 for exts in COMPATIBLE_EXTS:
1609 if video_ext in exts and audio_ext in exts:
1611 # TODO: Check acodec/vcodec
1614 filename_real_ext = os.path.splitext(filename)[1][1:]
1616 os.path.splitext(filename)[0]
1617 if filename_real_ext == info_dict['ext']
1619 requested_formats = info_dict['requested_formats']
1620 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1621 info_dict['ext'] = 'mkv'
1622 self.report_warning(
1623 'Requested formats are incompatible for merge and will be merged into mkv.')
1624 # Ensure filename always has a correct extension for successful merge
1625 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1626 if os.path.exists(encodeFilename(filename)):
1628 '[download] %s has already been downloaded and '
1629 'merged' % filename)
1631 for f in requested_formats:
1632 new_info = dict(info_dict)
1634 fname = self.prepare_filename(new_info)
1635 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1636 downloaded.append(fname)
1637 partial_success = dl(fname, new_info)
1638 success = success and partial_success
1639 info_dict['__postprocessors'] = postprocessors
1640 info_dict['__files_to_merge'] = downloaded
1642 # Just a single file
1643 success = dl(filename, info_dict)
1644 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1645 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1647 except (OSError, IOError) as err:
1648 raise UnavailableVideoError(err)
1649 except (ContentTooShortError, ) as err:
1650 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1653 if success and filename != '-':
1655 fixup_policy = self.params.get('fixup')
1656 if fixup_policy is None:
1657 fixup_policy = 'detect_or_warn'
1659 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1661 stretched_ratio = info_dict.get('stretched_ratio')
1662 if stretched_ratio is not None and stretched_ratio != 1:
1663 if fixup_policy == 'warn':
1664 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1665 info_dict['id'], stretched_ratio))
1666 elif fixup_policy == 'detect_or_warn':
1667 stretched_pp = FFmpegFixupStretchedPP(self)
1668 if stretched_pp.available:
1669 info_dict.setdefault('__postprocessors', [])
1670 info_dict['__postprocessors'].append(stretched_pp)
1672 self.report_warning(
1673 '%s: Non-uniform pixel ratio (%s). %s'
1674 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1676 assert fixup_policy in ('ignore', 'never')
1678 if (info_dict.get('requested_formats') is None and
1679 info_dict.get('container') == 'm4a_dash'):
1680 if fixup_policy == 'warn':
1681 self.report_warning(
1682 '%s: writing DASH m4a. '
1683 'Only some players support this container.'
1685 elif fixup_policy == 'detect_or_warn':
1686 fixup_pp = FFmpegFixupM4aPP(self)
1687 if fixup_pp.available:
1688 info_dict.setdefault('__postprocessors', [])
1689 info_dict['__postprocessors'].append(fixup_pp)
1691 self.report_warning(
1692 '%s: writing DASH m4a. '
1693 'Only some players support this container. %s'
1694 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1696 assert fixup_policy in ('ignore', 'never')
1698 if (info_dict.get('protocol') == 'm3u8_native' or
1699 info_dict.get('protocol') == 'm3u8' and
1700 self.params.get('hls_prefer_native')):
1701 if fixup_policy == 'warn':
1702 self.report_warning('%s: malformated aac bitstream.' % (
1704 elif fixup_policy == 'detect_or_warn':
1705 fixup_pp = FFmpegFixupM3u8PP(self)
1706 if fixup_pp.available:
1707 info_dict.setdefault('__postprocessors', [])
1708 info_dict['__postprocessors'].append(fixup_pp)
1710 self.report_warning(
1711 '%s: malformated aac bitstream. %s'
1712 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1714 assert fixup_policy in ('ignore', 'never')
1717 self.post_process(filename, info_dict)
1718 except (PostProcessingError) as err:
1719 self.report_error('postprocessing: %s' % str(err))
1721 self.record_download_archive(info_dict)
1723 def download(self, url_list):
1724 """Download a given list of URLs."""
1725 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1726 if (len(url_list) > 1 and
1727 '%' not in outtmpl and
1728 self.params.get('max_downloads') != 1):
1729 raise SameFileError(outtmpl)
1731 for url in url_list:
1733 # It also downloads the videos
1734 res = self.extract_info(
1735 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1736 except UnavailableVideoError:
1737 self.report_error('unable to download video')
1738 except MaxDownloadsReached:
1739 self.to_screen('[info] Maximum number of downloaded files reached.')
1742 if self.params.get('dump_single_json', False):
1743 self.to_stdout(json.dumps(res))
1745 return self._download_retcode
1747 def download_with_info_file(self, info_filename):
1748 with contextlib.closing(fileinput.FileInput(
1749 [info_filename], mode='r',
1750 openhook=fileinput.hook_encoded('utf-8'))) as f:
1751 # FileInput doesn't have a read method, we can't call json.load
1752 info = self.filter_requested_info(json.loads('\n'.join(f)))
1754 self.process_ie_result(info, download=True)
1755 except DownloadError:
1756 webpage_url = info.get('webpage_url')
1757 if webpage_url is not None:
1758 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1759 return self.download([webpage_url])
1762 return self._download_retcode
1765 def filter_requested_info(info_dict):
1767 (k, v) for k, v in info_dict.items()
1768 if k not in ['requested_formats', 'requested_subtitles'])
1770 def post_process(self, filename, ie_info):
1771 """Run all the postprocessors on the given file."""
1772 info = dict(ie_info)
1773 info['filepath'] = filename
1775 if ie_info.get('__postprocessors') is not None:
1776 pps_chain.extend(ie_info['__postprocessors'])
1777 pps_chain.extend(self._pps)
1778 for pp in pps_chain:
1779 files_to_delete = []
1781 files_to_delete, info = pp.run(info)
1782 except PostProcessingError as e:
1783 self.report_error(e.msg)
1784 if files_to_delete and not self.params.get('keepvideo', False):
1785 for old_filename in files_to_delete:
1786 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1788 os.remove(encodeFilename(old_filename))
1789 except (IOError, OSError):
1790 self.report_warning('Unable to remove downloaded original file')
1792 def _make_archive_id(self, info_dict):
1793 # Future-proof against any change in case
1794 # and backwards compatibility with prior versions
1795 extractor = info_dict.get('extractor_key')
1796 if extractor is None:
1797 if 'id' in info_dict:
1798 extractor = info_dict.get('ie_key') # key in a playlist
1799 if extractor is None:
1800 return None # Incomplete video information
1801 return extractor.lower() + ' ' + info_dict['id']
1803 def in_download_archive(self, info_dict):
1804 fn = self.params.get('download_archive')
1808 vid_id = self._make_archive_id(info_dict)
1810 return False # Incomplete video information
1813 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1814 for line in archive_file:
1815 if line.strip() == vid_id:
1817 except IOError as ioe:
1818 if ioe.errno != errno.ENOENT:
1822 def record_download_archive(self, info_dict):
1823 fn = self.params.get('download_archive')
1826 vid_id = self._make_archive_id(info_dict)
1828 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1829 archive_file.write(vid_id + '\n')
1832 def format_resolution(format, default='unknown'):
1833 if format.get('vcodec') == 'none':
1835 if format.get('resolution') is not None:
1836 return format['resolution']
1837 if format.get('height') is not None:
1838 if format.get('width') is not None:
1839 res = '%sx%s' % (format['width'], format['height'])
1841 res = '%sp' % format['height']
1842 elif format.get('width') is not None:
1843 res = '%dx?' % format['width']
1848 def _format_note(self, fdict):
1850 if fdict.get('ext') in ['f4f', 'f4m']:
1851 res += '(unsupported) '
1852 if fdict.get('language'):
1855 res += '[%s] ' % fdict['language']
1856 if fdict.get('format_note') is not None:
1857 res += fdict['format_note'] + ' '
1858 if fdict.get('tbr') is not None:
1859 res += '%4dk ' % fdict['tbr']
1860 if fdict.get('container') is not None:
1863 res += '%s container' % fdict['container']
1864 if (fdict.get('vcodec') is not None and
1865 fdict.get('vcodec') != 'none'):
1868 res += fdict['vcodec']
1869 if fdict.get('vbr') is not None:
1871 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1873 if fdict.get('vbr') is not None:
1874 res += '%4dk' % fdict['vbr']
1875 if fdict.get('fps') is not None:
1878 res += '%sfps' % fdict['fps']
1879 if fdict.get('acodec') is not None:
1882 if fdict['acodec'] == 'none':
1885 res += '%-5s' % fdict['acodec']
1886 elif fdict.get('abr') is not None:
1890 if fdict.get('abr') is not None:
1891 res += '@%3dk' % fdict['abr']
1892 if fdict.get('asr') is not None:
1893 res += ' (%5dHz)' % fdict['asr']
1894 if fdict.get('filesize') is not None:
1897 res += format_bytes(fdict['filesize'])
1898 elif fdict.get('filesize_approx') is not None:
1901 res += '~' + format_bytes(fdict['filesize_approx'])
1904 def list_formats(self, info_dict):
1905 formats = info_dict.get('formats', [info_dict])
1907 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1909 if f.get('preference') is None or f['preference'] >= -1000]
1910 if len(formats) > 1:
1911 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1913 header_line = ['format code', 'extension', 'resolution', 'note']
1915 '[info] Available formats for %s:\n%s' %
1916 (info_dict['id'], render_table(header_line, table)))
1918 def list_thumbnails(self, info_dict):
1919 thumbnails = info_dict.get('thumbnails')
1921 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
1925 '[info] Thumbnails for %s:' % info_dict['id'])
1926 self.to_screen(render_table(
1927 ['ID', 'width', 'height', 'URL'],
1928 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1930 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1932 self.to_screen('%s has no %s' % (video_id, name))
1935 'Available %s for %s:' % (name, video_id))
1936 self.to_screen(render_table(
1937 ['Language', 'formats'],
1938 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1939 for lang, formats in subtitles.items()]))
1941 def urlopen(self, req):
1942 """ Start an HTTP download """
1943 if isinstance(req, compat_basestring):
1944 req = sanitized_Request(req)
1945 return self._opener.open(req, timeout=self._socket_timeout)
1947 def print_debug_header(self):
1948 if not self.params.get('verbose'):
1951 if type('') is not compat_str:
1952 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1953 self.report_warning(
1954 'Your Python is broken! Update to a newer and supported version')
1956 stdout_encoding = getattr(
1957 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1959 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1960 locale.getpreferredencoding(),
1961 sys.getfilesystemencoding(),
1963 self.get_encoding()))
1964 write_string(encoding_str, encoding=None)
1966 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1968 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
1970 sp = subprocess.Popen(
1971 ['git', 'rev-parse', '--short', 'HEAD'],
1972 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1973 cwd=os.path.dirname(os.path.abspath(__file__)))
1974 out, err = sp.communicate()
1975 out = out.decode().strip()
1976 if re.match('[0-9a-f]+', out):
1977 self._write_string('[debug] Git HEAD: ' + out + '\n')
1983 self._write_string('[debug] Python version %s - %s\n' % (
1984 platform.python_version(), platform_name()))
1986 exe_versions = FFmpegPostProcessor.get_versions(self)
1987 exe_versions['rtmpdump'] = rtmpdump_version()
1988 exe_str = ', '.join(
1990 for exe, v in sorted(exe_versions.items())
1995 self._write_string('[debug] exe versions: %s\n' % exe_str)
1998 for handler in self._opener.handlers:
1999 if hasattr(handler, 'proxies'):
2000 proxy_map.update(handler.proxies)
2001 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2003 if self.params.get('call_home', False):
2004 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2005 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2006 latest_version = self.urlopen(
2007 'https://yt-dl.org/latest/version').read().decode('utf-8')
2008 if version_tuple(latest_version) > version_tuple(__version__):
2009 self.report_warning(
2010 'You are using an outdated version (newest version: %s)! '
2011 'See https://yt-dl.org/update if you need help updating.' %
2014 def _setup_opener(self):
2015 timeout_val = self.params.get('socket_timeout')
2016 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2018 opts_cookiefile = self.params.get('cookiefile')
2019 opts_proxy = self.params.get('proxy')
2021 if opts_cookiefile is None:
2022 self.cookiejar = compat_cookiejar.CookieJar()
2024 opts_cookiefile = compat_expanduser(opts_cookiefile)
2025 self.cookiejar = compat_cookiejar.MozillaCookieJar(
2027 if os.access(opts_cookiefile, os.R_OK):
2028 self.cookiejar.load()
2030 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2031 if opts_proxy is not None:
2032 if opts_proxy == '':
2035 proxies = {'http': opts_proxy, 'https': opts_proxy}
2037 proxies = compat_urllib_request.getproxies()
2038 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2039 if 'http' in proxies and 'https' not in proxies:
2040 proxies['https'] = proxies['http']
2041 proxy_handler = PerRequestProxyHandler(proxies)
2043 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2044 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2045 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2046 data_handler = compat_urllib_request_DataHandler()
2048 # When passing our own FileHandler instance, build_opener won't add the
2049 # default FileHandler and allows us to disable the file protocol, which
2050 # can be used for malicious purposes (see
2051 # https://github.com/rg3/youtube-dl/issues/8227)
2052 file_handler = compat_urllib_request.FileHandler()
2054 def file_open(*args, **kwargs):
2055 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2056 file_handler.file_open = file_open
2058 opener = compat_urllib_request.build_opener(
2059 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2061 # Delete the default user-agent header, which would otherwise apply in
2062 # cases where our custom HTTP handler doesn't come into play
2063 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2064 opener.addheaders = []
2065 self._opener = opener
2067 def encode(self, s):
2068 if isinstance(s, bytes):
2069 return s # Already encoded
2072 return s.encode(self.get_encoding())
2073 except UnicodeEncodeError as err:
2074 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2077 def get_encoding(self):
2078 encoding = self.params.get('encoding')
2079 if encoding is None:
2080 encoding = preferredencoding()
2083 def _write_thumbnails(self, info_dict, filename):
2084 if self.params.get('writethumbnail', False):
2085 thumbnails = info_dict.get('thumbnails')
2087 thumbnails = [thumbnails[-1]]
2088 elif self.params.get('write_all_thumbnails', False):
2089 thumbnails = info_dict.get('thumbnails')
2094 # No thumbnails present, so return immediately
2097 for t in thumbnails:
2098 thumb_ext = determine_ext(t['url'], 'jpg')
2099 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2100 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2101 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2103 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2104 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2105 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2107 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2108 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2110 uf = self.urlopen(t['url'])
2111 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2112 shutil.copyfileobj(uf, thumbf)
2113 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2114 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2115 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2116 self.report_warning('Unable to download thumbnail "%s": %s' %
2117 (t['url'], error_to_compat_str(err)))