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,
60 PerRequestProxyHandler,
71 UnavailableVideoError,
76 YoutubeDLCookieProcessor,
83 from .cache import Cache
84 from .extractor import get_info_extractor, gen_extractors
85 from .downloader import get_suitable_downloader
86 from .downloader.rtmp import rtmpdump_version
87 from .postprocessor import (
89 FFmpegFixupStretchedPP,
94 from .version import __version__
96 if compat_os_name == 'nt':
100 class YoutubeDL(object):
103 YoutubeDL objects are the ones responsible of downloading the
104 actual video file and writing it to disk if the user has requested
105 it, among some other tasks. In most cases there should be one per
106 program. As, given a video URL, the downloader doesn't know how to
107 extract all the needed information, task that InfoExtractors do, it
108 has to pass the URL to one of them.
110 For this, YoutubeDL objects have a method that allows
111 InfoExtractors to be registered in a given order. When it is passed
112 a URL, the YoutubeDL object handles it to the first InfoExtractor it
113 finds that reports being able to handle it. The InfoExtractor extracts
114 all the information about the video or videos the URL refers to, and
115 YoutubeDL process the extracted information, possibly using a File
116 Downloader to download the video.
118 YoutubeDL objects accept a lot of parameters. In order not to saturate
119 the object constructor with arguments, it receives a dictionary of
120 options instead. These options are available through the params
121 attribute for the InfoExtractors to use. The YoutubeDL also
122 registers itself as the downloader in charge for the InfoExtractors
123 that are added to it, so this is a "mutual registration".
127 username: Username for authentication purposes.
128 password: Password for authentication purposes.
129 videopassword: Password for accessing a video.
130 usenetrc: Use netrc for authentication instead.
131 verbose: Print additional info to stdout.
132 quiet: Do not print messages to stdout.
133 no_warnings: Do not print out anything for warnings.
134 forceurl: Force printing final URL.
135 forcetitle: Force printing title.
136 forceid: Force printing ID.
137 forcethumbnail: Force printing thumbnail URL.
138 forcedescription: Force printing description.
139 forcefilename: Force printing final filename.
140 forceduration: Force printing duration.
141 forcejson: Force printing info_dict as JSON.
142 dump_single_json: Force printing the info_dict of the whole playlist
143 (or video) as a single JSON line.
144 simulate: Do not download the video files.
145 format: Video format code. See options.py for more information.
146 outtmpl: Template for output names.
147 restrictfilenames: Do not allow "&" and spaces in file names
148 ignoreerrors: Do not stop on download errors.
149 force_generic_extractor: Force downloader to use the generic extractor
150 nooverwrites: Prevent overwriting files.
151 playliststart: Playlist item to start at.
152 playlistend: Playlist item to end at.
153 playlist_items: Specific indices of playlist to download.
154 playlistreverse: Download playlist items in reverse order.
155 matchtitle: Download only matching titles.
156 rejecttitle: Reject downloads for matching titles.
157 logger: Log messages to a logging.Logger instance.
158 logtostderr: Log messages to stderr instead of stdout.
159 writedescription: Write the video description to a .description file
160 writeinfojson: Write the video description to a .info.json file
161 writeannotations: Write the video annotations to a .annotations.xml file
162 writethumbnail: Write the thumbnail image to a file
163 write_all_thumbnails: Write all thumbnail formats to files
164 writesubtitles: Write the video subtitles to a file
165 writeautomaticsub: Write the automatically generated subtitles to a file
166 allsubtitles: Downloads all the subtitles of the video
167 (requires writesubtitles or writeautomaticsub)
168 listsubtitles: Lists all available subtitles for the video
169 subtitlesformat: The format code for subtitles
170 subtitleslangs: List of languages of the subtitles to download
171 keepvideo: Keep the video file after post-processing
172 daterange: A DateRange object, download only if the upload_date is in the range.
173 skip_download: Skip the actual download of the video file
174 cachedir: Location of the cache files in the filesystem.
175 False to disable filesystem cache.
176 noplaylist: Download single video instead of a playlist if in doubt.
177 age_limit: An integer representing the user's age in years.
178 Unsuitable videos for the given age are skipped.
179 min_views: An integer representing the minimum view count the video
180 must have in order to not be skipped.
181 Videos without view count information are always
182 downloaded. None for no limit.
183 max_views: An integer representing the maximum view count.
184 Videos that are more popular than that are not
186 Videos without view count information are always
187 downloaded. None for no limit.
188 download_archive: File name of a file where all downloads are recorded.
189 Videos already present in the file are not downloaded
191 cookiefile: File name where cookies should be read from and dumped to.
192 nocheckcertificate:Do not verify SSL certificates
193 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
194 At the moment, this is only supported by YouTube.
195 proxy: URL of the proxy server to use
196 cn_verification_proxy: URL of the proxy to use for IP address verification
197 on Chinese sites. (Experimental)
198 socket_timeout: Time to wait for unresponsive hosts, in seconds
199 bidi_workaround: Work around buggy terminals without bidirectional text
200 support, using fridibi
201 debug_printtraffic:Print out sent and received HTTP traffic
202 include_ads: Download ads as well
203 default_search: Prepend this string if an input url is not valid.
204 'auto' for elaborate guessing
205 encoding: Use this encoding instead of the system-specified.
206 extract_flat: Do not resolve URLs, return the immediate result.
207 Pass in 'in_playlist' to only show this behavior for
209 postprocessors: A list of dictionaries, each with an entry
210 * key: The name of the postprocessor. See
211 youtube_dl/postprocessor/__init__.py for a list.
212 as well as any further keyword arguments for the
214 progress_hooks: A list of functions that get called on download
215 progress, with a dictionary with the entries
216 * status: One of "downloading", "error", or "finished".
217 Check this first and ignore unknown values.
219 If status is one of "downloading", or "finished", the
220 following properties may also be present:
221 * filename: The final filename (always present)
222 * tmpfilename: The filename we're currently writing to
223 * downloaded_bytes: Bytes on disk
224 * total_bytes: Size of the whole file, None if unknown
225 * total_bytes_estimate: Guess of the eventual file size,
227 * elapsed: The number of seconds since download started.
228 * eta: The estimated time in seconds, None if unknown
229 * speed: The download speed in bytes/second, None if
231 * fragment_index: The counter of the currently
232 downloaded video fragment.
233 * fragment_count: The number of fragments (= individual
234 files that will be merged)
236 Progress hooks are guaranteed to be called at least once
237 (with status "finished") if the download is successful.
238 merge_output_format: Extension to use when merging formats.
239 fixup: Automatically correct known faults of the file.
241 - "never": do nothing
242 - "warn": only emit a warning
243 - "detect_or_warn": check whether we can do anything
244 about it, warn otherwise (default)
245 source_address: (Experimental) Client-side IP address to bind to.
246 call_home: Boolean, true iff we are allowed to contact the
247 youtube-dl servers for debugging.
248 sleep_interval: Number of seconds to sleep before each download.
249 listformats: Print an overview of available video formats and exit.
250 list_thumbnails: Print a table of all thumbnails and exit.
251 match_filter: A function that gets called with the info_dict of
253 If it returns a message, the video is ignored.
254 If it returns None, the video is downloaded.
255 match_filter_func in utils.py is one example for this.
256 no_color: Do not emit color codes in output.
258 The following options determine which downloader is picked:
259 external_downloader: Executable of the external downloader to call.
260 None or unset for standard (built-in) downloader.
261 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv.
263 The following parameters are not used by YoutubeDL itself, they are used by
264 the downloader (see youtube_dl/downloader/common.py):
265 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
266 noresizebuffer, retries, continuedl, noprogress, consoletitle,
267 xattr_set_filesize, external_downloader_args, hls_use_mpegts.
269 The following options are used by the post processors:
270 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
271 otherwise prefer avconv.
272 postprocessor_args: A list of additional command-line arguments for the
279 _download_retcode = None
280 _num_downloads = None
283 def __init__(self, params=None, auto_init=True):
284 """Create a FileDownloader object with the given options."""
288 self._ies_instances = {}
290 self._progress_hooks = []
291 self._download_retcode = 0
292 self._num_downloads = 0
293 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
294 self._err_file = sys.stderr
297 'nocheckcertificate': False,
299 self.params.update(params)
300 self.cache = Cache(self)
302 if params.get('bidi_workaround', False):
305 master, slave = pty.openpty()
306 width = compat_get_terminal_size().columns
310 width_args = ['-w', str(width)]
312 stdin=subprocess.PIPE,
314 stderr=self._err_file)
316 self._output_process = subprocess.Popen(
317 ['bidiv'] + width_args, **sp_kwargs
320 self._output_process = subprocess.Popen(
321 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
322 self._output_channel = os.fdopen(master, 'rb')
323 except OSError as ose:
325 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 $PATH.')
329 if (sys.version_info >= (3,) and sys.platform != 'win32' and
330 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
331 not params.get('restrictfilenames', False)):
332 # On Python 3, the Unicode filesystem API will throw errors (#1474)
334 'Assuming --restrict-filenames since file system encoding '
335 'cannot encode all characters. '
336 'Set the LC_ALL environment variable to fix this.')
337 self.params['restrictfilenames'] = True
339 if isinstance(params.get('outtmpl'), bytes):
341 'Parameter outtmpl is bytes, but should be a unicode string. '
342 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
347 self.print_debug_header()
348 self.add_default_info_extractors()
350 for pp_def_raw in self.params.get('postprocessors', []):
351 pp_class = get_postprocessor(pp_def_raw['key'])
352 pp_def = dict(pp_def_raw)
354 pp = pp_class(self, **compat_kwargs(pp_def))
355 self.add_post_processor(pp)
357 for ph in self.params.get('progress_hooks', []):
358 self.add_progress_hook(ph)
360 def warn_if_short_id(self, argv):
361 # short YouTube ID starting with dash?
363 i for i, a in enumerate(argv)
364 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
368 [a for i, a in enumerate(argv) if i not in idxs] +
369 ['--'] + [argv[i] for i in idxs]
372 'Long argument string detected. '
373 'Use -- to separate parameters and URLs, like this:\n%s\n' %
374 args_to_str(correct_argv))
376 def add_info_extractor(self, ie):
377 """Add an InfoExtractor object to the end of the list."""
379 self._ies_instances[ie.ie_key()] = ie
380 ie.set_downloader(self)
382 def get_info_extractor(self, ie_key):
384 Get an instance of an IE with name ie_key, it will try to get one from
385 the _ies list, if there's no instance it will create a new one and add
386 it to the extractor list.
388 ie = self._ies_instances.get(ie_key)
390 ie = get_info_extractor(ie_key)()
391 self.add_info_extractor(ie)
394 def add_default_info_extractors(self):
396 Add the InfoExtractors returned by gen_extractors to the end of the list
398 for ie in gen_extractors():
399 self.add_info_extractor(ie)
401 def add_post_processor(self, pp):
402 """Add a PostProcessor object to the end of the chain."""
404 pp.set_downloader(self)
406 def add_progress_hook(self, ph):
407 """Add the progress hook (currently only for the file downloader)"""
408 self._progress_hooks.append(ph)
410 def _bidi_workaround(self, message):
411 if not hasattr(self, '_output_channel'):
414 assert hasattr(self, '_output_process')
415 assert isinstance(message, compat_str)
416 line_count = message.count('\n') + 1
417 self._output_process.stdin.write((message + '\n').encode('utf-8'))
418 self._output_process.stdin.flush()
419 res = ''.join(self._output_channel.readline().decode('utf-8')
420 for _ in range(line_count))
421 return res[:-len('\n')]
423 def to_screen(self, message, skip_eol=False):
424 """Print message to stdout if not in quiet mode."""
425 return self.to_stdout(message, skip_eol, check_quiet=True)
427 def _write_string(self, s, out=None):
428 write_string(s, out=out, encoding=self.params.get('encoding'))
430 def to_stdout(self, message, skip_eol=False, check_quiet=False):
431 """Print message to stdout if not in quiet mode."""
432 if self.params.get('logger'):
433 self.params['logger'].debug(message)
434 elif not check_quiet or not self.params.get('quiet', False):
435 message = self._bidi_workaround(message)
436 terminator = ['\n', ''][skip_eol]
437 output = message + terminator
439 self._write_string(output, self._screen_file)
441 def to_stderr(self, message):
442 """Print message to stderr."""
443 assert isinstance(message, compat_str)
444 if self.params.get('logger'):
445 self.params['logger'].error(message)
447 message = self._bidi_workaround(message)
448 output = message + '\n'
449 self._write_string(output, self._err_file)
451 def to_console_title(self, message):
452 if not self.params.get('consoletitle', False):
454 if compat_os_name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
455 # c_wchar_p() might not be necessary if `message` is
456 # already of type unicode()
457 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
458 elif 'TERM' in os.environ:
459 self._write_string('\033]0;%s\007' % message, self._screen_file)
461 def save_console_title(self):
462 if not self.params.get('consoletitle', False):
464 if 'TERM' in os.environ:
465 # Save the title on stack
466 self._write_string('\033[22;0t', self._screen_file)
468 def restore_console_title(self):
469 if not self.params.get('consoletitle', False):
471 if 'TERM' in os.environ:
472 # Restore the title from stack
473 self._write_string('\033[23;0t', self._screen_file)
476 self.save_console_title()
479 def __exit__(self, *args):
480 self.restore_console_title()
482 if self.params.get('cookiefile') is not None:
483 self.cookiejar.save()
485 def trouble(self, message=None, tb=None):
486 """Determine action to take when a download problem appears.
488 Depending on if the downloader has been configured to ignore
489 download errors or not, this method may throw an exception or
490 not when errors are found, after printing the message.
492 tb, if given, is additional traceback information.
494 if message is not None:
495 self.to_stderr(message)
496 if self.params.get('verbose'):
498 if sys.exc_info()[0]: # if .trouble has been called from an except block
500 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
501 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
502 tb += encode_compat_str(traceback.format_exc())
504 tb_data = traceback.format_list(traceback.extract_stack())
505 tb = ''.join(tb_data)
507 if not self.params.get('ignoreerrors', False):
508 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
509 exc_info = sys.exc_info()[1].exc_info
511 exc_info = sys.exc_info()
512 raise DownloadError(message, exc_info)
513 self._download_retcode = 1
515 def report_warning(self, message):
517 Print the message to stderr, it will be prefixed with 'WARNING:'
518 If stderr is a tty file the 'WARNING:' will be colored
520 if self.params.get('logger') is not None:
521 self.params['logger'].warning(message)
523 if self.params.get('no_warnings'):
525 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
526 _msg_header = '\033[0;33mWARNING:\033[0m'
528 _msg_header = 'WARNING:'
529 warning_message = '%s %s' % (_msg_header, message)
530 self.to_stderr(warning_message)
532 def report_error(self, message, tb=None):
534 Do the same as trouble, but prefixes the message with 'ERROR:', colored
535 in red if stderr is a tty file.
537 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
538 _msg_header = '\033[0;31mERROR:\033[0m'
540 _msg_header = 'ERROR:'
541 error_message = '%s %s' % (_msg_header, message)
542 self.trouble(error_message, tb)
544 def report_file_already_downloaded(self, file_name):
545 """Report file has already been fully downloaded."""
547 self.to_screen('[download] %s has already been downloaded' % file_name)
548 except UnicodeEncodeError:
549 self.to_screen('[download] The file has already been downloaded')
551 def prepare_filename(self, info_dict):
552 """Generate the output filename."""
554 template_dict = dict(info_dict)
556 template_dict['epoch'] = int(time.time())
557 autonumber_size = self.params.get('autonumber_size')
558 if autonumber_size is None:
560 autonumber_templ = '%0' + str(autonumber_size) + 'd'
561 template_dict['autonumber'] = autonumber_templ % self._num_downloads
562 if template_dict.get('playlist_index') is not None:
563 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
564 if template_dict.get('resolution') is None:
565 if template_dict.get('width') and template_dict.get('height'):
566 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
567 elif template_dict.get('height'):
568 template_dict['resolution'] = '%sp' % template_dict['height']
569 elif template_dict.get('width'):
570 template_dict['resolution'] = '?x%d' % template_dict['width']
572 sanitize = lambda k, v: sanitize_filename(
574 restricted=self.params.get('restrictfilenames'),
576 template_dict = dict((k, sanitize(k, v))
577 for k, v in template_dict.items()
579 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
581 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
582 tmpl = compat_expanduser(outtmpl)
583 filename = tmpl % template_dict
584 # Temporary fix for #4787
585 # 'Treat' all problem characters by passing filename through preferredencoding
586 # to workaround encoding issues with subprocess on python2 @ Windows
587 if sys.version_info < (3, 0) and sys.platform == 'win32':
588 filename = encodeFilename(filename, True).decode(preferredencoding())
589 return sanitize_path(filename)
590 except ValueError as err:
591 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
594 def _match_entry(self, info_dict, incomplete):
595 """ Returns None iff the file should be downloaded """
597 video_title = info_dict.get('title', info_dict.get('id', 'video'))
598 if 'title' in info_dict:
599 # This can happen when we're just evaluating the playlist
600 title = info_dict['title']
601 matchtitle = self.params.get('matchtitle', False)
603 if not re.search(matchtitle, title, re.IGNORECASE):
604 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
605 rejecttitle = self.params.get('rejecttitle', False)
607 if re.search(rejecttitle, title, re.IGNORECASE):
608 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
609 date = info_dict.get('upload_date')
611 dateRange = self.params.get('daterange', DateRange())
612 if date not in dateRange:
613 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
614 view_count = info_dict.get('view_count')
615 if view_count is not None:
616 min_views = self.params.get('min_views')
617 if min_views is not None and view_count < min_views:
618 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
619 max_views = self.params.get('max_views')
620 if max_views is not None and view_count > max_views:
621 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
622 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
623 return 'Skipping "%s" because it is age restricted' % video_title
624 if self.in_download_archive(info_dict):
625 return '%s has already been recorded in archive' % video_title
628 match_filter = self.params.get('match_filter')
629 if match_filter is not None:
630 ret = match_filter(info_dict)
637 def add_extra_info(info_dict, extra_info):
638 '''Set the keys from extra_info in info dict if they are missing'''
639 for key, value in extra_info.items():
640 info_dict.setdefault(key, value)
642 def extract_info(self, url, download=True, ie_key=None, extra_info={},
643 process=True, force_generic_extractor=False):
645 Returns a list with a dictionary for each video we find.
646 If 'download', also downloads the videos.
647 extra_info is a dict containing the extra values to add to each result
650 if not ie_key and force_generic_extractor:
654 ies = [self.get_info_extractor(ie_key)]
659 if not ie.suitable(url):
663 self.report_warning('The program functionality for this site has been marked as broken, '
664 'and will probably not work.')
667 ie_result = ie.extract(url)
668 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
670 if isinstance(ie_result, list):
671 # Backwards compatibility: old IE result format
673 '_type': 'compat_list',
674 'entries': ie_result,
676 self.add_default_extra_info(ie_result, ie, url)
678 return self.process_ie_result(ie_result, download, extra_info)
681 except ExtractorError as e: # An error we somewhat expected
682 self.report_error(compat_str(e), e.format_traceback())
684 except MaxDownloadsReached:
686 except Exception as e:
687 if self.params.get('ignoreerrors', False):
688 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
693 self.report_error('no suitable InfoExtractor for URL %s' % url)
695 def add_default_extra_info(self, ie_result, ie, url):
696 self.add_extra_info(ie_result, {
697 'extractor': ie.IE_NAME,
699 'webpage_url_basename': url_basename(url),
700 'extractor_key': ie.ie_key(),
703 def process_ie_result(self, ie_result, download=True, extra_info={}):
705 Take the result of the ie(may be modified) and resolve all unresolved
706 references (URLs, playlist items).
708 It will also download the videos if 'download'.
709 Returns the resolved ie_result.
711 result_type = ie_result.get('_type', 'video')
713 if result_type in ('url', 'url_transparent'):
714 extract_flat = self.params.get('extract_flat', False)
715 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
716 extract_flat is True):
717 if self.params.get('forcejson', False):
718 self.to_stdout(json.dumps(ie_result))
721 if result_type == 'video':
722 self.add_extra_info(ie_result, extra_info)
723 return self.process_video_result(ie_result, download=download)
724 elif result_type == 'url':
725 # We have to add extra_info to the results because it may be
726 # contained in a playlist
727 return self.extract_info(ie_result['url'],
729 ie_key=ie_result.get('ie_key'),
730 extra_info=extra_info)
731 elif result_type == 'url_transparent':
732 # Use the information from the embedding page
733 info = self.extract_info(
734 ie_result['url'], ie_key=ie_result.get('ie_key'),
735 extra_info=extra_info, download=False, process=False)
737 force_properties = dict(
738 (k, v) for k, v in ie_result.items() if v is not None)
739 for f in ('_type', 'url', 'ie_key'):
740 if f in force_properties:
741 del force_properties[f]
742 new_result = info.copy()
743 new_result.update(force_properties)
745 assert new_result.get('_type') != 'url_transparent'
747 return self.process_ie_result(
748 new_result, download=download, extra_info=extra_info)
749 elif result_type == 'playlist' or result_type == 'multi_video':
750 # We process each entry in the playlist
751 playlist = ie_result.get('title') or ie_result.get('id')
752 self.to_screen('[download] Downloading playlist: %s' % playlist)
754 playlist_results = []
756 playliststart = self.params.get('playliststart', 1) - 1
757 playlistend = self.params.get('playlistend')
758 # For backwards compatibility, interpret -1 as whole list
759 if playlistend == -1:
762 playlistitems_str = self.params.get('playlist_items')
764 if playlistitems_str is not None:
765 def iter_playlistitems(format):
766 for string_segment in format.split(','):
767 if '-' in string_segment:
768 start, end = string_segment.split('-')
769 for item in range(int(start), int(end) + 1):
772 yield int(string_segment)
773 playlistitems = iter_playlistitems(playlistitems_str)
775 ie_entries = ie_result['entries']
776 if isinstance(ie_entries, list):
777 n_all_entries = len(ie_entries)
780 ie_entries[i - 1] for i in playlistitems
781 if -n_all_entries <= i - 1 < n_all_entries]
783 entries = ie_entries[playliststart:playlistend]
784 n_entries = len(entries)
786 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
787 (ie_result['extractor'], playlist, n_all_entries, n_entries))
788 elif isinstance(ie_entries, PagedList):
791 for item in playlistitems:
792 entries.extend(ie_entries.getslice(
796 entries = ie_entries.getslice(
797 playliststart, playlistend)
798 n_entries = len(entries)
800 '[%s] playlist %s: Downloading %d videos' %
801 (ie_result['extractor'], playlist, n_entries))
804 entry_list = list(ie_entries)
805 entries = [entry_list[i - 1] for i in playlistitems]
807 entries = list(itertools.islice(
808 ie_entries, playliststart, playlistend))
809 n_entries = len(entries)
811 '[%s] playlist %s: Downloading %d videos' %
812 (ie_result['extractor'], playlist, n_entries))
814 if self.params.get('playlistreverse', False):
815 entries = entries[::-1]
817 for i, entry in enumerate(entries, 1):
818 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
820 'n_entries': n_entries,
821 'playlist': playlist,
822 'playlist_id': ie_result.get('id'),
823 'playlist_title': ie_result.get('title'),
824 'playlist_index': i + playliststart,
825 'extractor': ie_result['extractor'],
826 'webpage_url': ie_result['webpage_url'],
827 'webpage_url_basename': url_basename(ie_result['webpage_url']),
828 'extractor_key': ie_result['extractor_key'],
831 reason = self._match_entry(entry, incomplete=True)
832 if reason is not None:
833 self.to_screen('[download] ' + reason)
836 entry_result = self.process_ie_result(entry,
839 playlist_results.append(entry_result)
840 ie_result['entries'] = playlist_results
841 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
843 elif result_type == 'compat_list':
845 'Extractor %s returned a compat_list result. '
846 'It needs to be updated.' % ie_result.get('extractor'))
852 'extractor': ie_result['extractor'],
853 'webpage_url': ie_result['webpage_url'],
854 'webpage_url_basename': url_basename(ie_result['webpage_url']),
855 'extractor_key': ie_result['extractor_key'],
859 ie_result['entries'] = [
860 self.process_ie_result(_fixup(r), download, extra_info)
861 for r in ie_result['entries']
865 raise Exception('Invalid result type: %s' % result_type)
867 def _build_format_filter(self, filter_spec):
868 " Returns a function to filter the formats according to the filter_spec "
878 operator_rex = re.compile(r'''(?x)\s*
879 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
880 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
881 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
883 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
884 m = operator_rex.search(filter_spec)
887 comparison_value = int(m.group('value'))
889 comparison_value = parse_filesize(m.group('value'))
890 if comparison_value is None:
891 comparison_value = parse_filesize(m.group('value') + 'B')
892 if comparison_value is None:
894 'Invalid value %r in format specification %r' % (
895 m.group('value'), filter_spec))
896 op = OPERATORS[m.group('op')]
902 '^=': lambda attr, value: attr.startswith(value),
903 '$=': lambda attr, value: attr.endswith(value),
904 '*=': lambda attr, value: value in attr,
906 str_operator_rex = re.compile(r'''(?x)
907 \s*(?P<key>ext|acodec|vcodec|container|protocol)
908 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
909 \s*(?P<value>[a-zA-Z0-9._-]+)
911 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
912 m = str_operator_rex.search(filter_spec)
914 comparison_value = m.group('value')
915 op = STR_OPERATORS[m.group('op')]
918 raise ValueError('Invalid filter specification %r' % filter_spec)
921 actual_value = f.get(m.group('key'))
922 if actual_value is None:
923 return m.group('none_inclusive')
924 return op(actual_value, comparison_value)
927 def build_format_selector(self, format_spec):
928 def syntax_error(note, start):
930 'Invalid format specification: '
931 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
932 return SyntaxError(message)
934 PICKFIRST = 'PICKFIRST'
938 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
940 def _parse_filter(tokens):
942 for type, string, start, _, _ in tokens:
943 if type == tokenize.OP and string == ']':
944 return ''.join(filter_parts)
946 filter_parts.append(string)
948 def _remove_unused_ops(tokens):
949 # Remove operators that we don't use and join them with the surrounding strings
950 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
951 ALLOWED_OPS = ('/', '+', ',', '(', ')')
952 last_string, last_start, last_end, last_line = None, None, None, None
953 for type, string, start, end, line in tokens:
954 if type == tokenize.OP and string == '[':
956 yield tokenize.NAME, last_string, last_start, last_end, last_line
958 yield type, string, start, end, line
959 # everything inside brackets will be handled by _parse_filter
960 for type, string, start, end, line in tokens:
961 yield type, string, start, end, line
962 if type == tokenize.OP and string == ']':
964 elif type == tokenize.OP and string in ALLOWED_OPS:
966 yield tokenize.NAME, last_string, last_start, last_end, last_line
968 yield type, string, start, end, line
969 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
975 last_string += string
977 yield tokenize.NAME, last_string, last_start, last_end, last_line
979 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
981 current_selector = None
982 for type, string, start, _, _ in tokens:
983 # ENCODING is only defined in python 3.x
984 if type == getattr(tokenize, 'ENCODING', None):
986 elif type in [tokenize.NAME, tokenize.NUMBER]:
987 current_selector = FormatSelector(SINGLE, string, [])
988 elif type == tokenize.OP:
991 # ')' will be handled by the parentheses group
992 tokens.restore_last_token()
994 elif inside_merge and string in ['/', ',']:
995 tokens.restore_last_token()
997 elif inside_choice and string == ',':
998 tokens.restore_last_token()
1001 if not current_selector:
1002 raise syntax_error('"," must follow a format selector', start)
1003 selectors.append(current_selector)
1004 current_selector = None
1006 if not current_selector:
1007 raise syntax_error('"/" must follow a format selector', start)
1008 first_choice = current_selector
1009 second_choice = _parse_format_selection(tokens, inside_choice=True)
1010 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1012 if not current_selector:
1013 current_selector = FormatSelector(SINGLE, 'best', [])
1014 format_filter = _parse_filter(tokens)
1015 current_selector.filters.append(format_filter)
1017 if current_selector:
1018 raise syntax_error('Unexpected "("', start)
1019 group = _parse_format_selection(tokens, inside_group=True)
1020 current_selector = FormatSelector(GROUP, group, [])
1022 video_selector = current_selector
1023 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1024 if not video_selector or not audio_selector:
1025 raise syntax_error('"+" must be between two format selectors', start)
1026 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1028 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1029 elif type == tokenize.ENDMARKER:
1031 if current_selector:
1032 selectors.append(current_selector)
1035 def _build_selector_function(selector):
1036 if isinstance(selector, list):
1037 fs = [_build_selector_function(s) for s in selector]
1039 def selector_function(formats):
1041 for format in f(formats):
1043 return selector_function
1044 elif selector.type == GROUP:
1045 selector_function = _build_selector_function(selector.selector)
1046 elif selector.type == PICKFIRST:
1047 fs = [_build_selector_function(s) for s in selector.selector]
1049 def selector_function(formats):
1051 picked_formats = list(f(formats))
1053 return picked_formats
1055 elif selector.type == SINGLE:
1056 format_spec = selector.selector
1058 def selector_function(formats):
1059 formats = list(formats)
1062 if format_spec == 'all':
1065 elif format_spec in ['best', 'worst', None]:
1066 format_idx = 0 if format_spec == 'worst' else -1
1067 audiovideo_formats = [
1069 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1070 if audiovideo_formats:
1071 yield audiovideo_formats[format_idx]
1072 # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format
1073 elif (all(f.get('acodec') != 'none' for f in formats) or
1074 all(f.get('vcodec') != 'none' for f in formats)):
1075 yield formats[format_idx]
1076 elif format_spec == 'bestaudio':
1079 if f.get('vcodec') == 'none']
1081 yield audio_formats[-1]
1082 elif format_spec == 'worstaudio':
1085 if f.get('vcodec') == 'none']
1087 yield audio_formats[0]
1088 elif format_spec == 'bestvideo':
1091 if f.get('acodec') == 'none']
1093 yield video_formats[-1]
1094 elif format_spec == 'worstvideo':
1097 if f.get('acodec') == 'none']
1099 yield video_formats[0]
1101 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1102 if format_spec in extensions:
1103 filter_f = lambda f: f['ext'] == format_spec
1105 filter_f = lambda f: f['format_id'] == format_spec
1106 matches = list(filter(filter_f, formats))
1109 elif selector.type == MERGE:
1110 def _merge(formats_info):
1111 format_1, format_2 = [f['format_id'] for f in formats_info]
1112 # The first format must contain the video and the
1114 if formats_info[0].get('vcodec') == 'none':
1115 self.report_error('The first format must '
1116 'contain the video, try using '
1117 '"-f %s+%s"' % (format_2, format_1))
1119 # Formats must be opposite (video+audio)
1120 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1122 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1123 % (format_1, format_2))
1126 formats_info[0]['ext']
1127 if self.params.get('merge_output_format') is None
1128 else self.params['merge_output_format'])
1130 'requested_formats': formats_info,
1131 'format': '%s+%s' % (formats_info[0].get('format'),
1132 formats_info[1].get('format')),
1133 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1134 formats_info[1].get('format_id')),
1135 'width': formats_info[0].get('width'),
1136 'height': formats_info[0].get('height'),
1137 'resolution': formats_info[0].get('resolution'),
1138 'fps': formats_info[0].get('fps'),
1139 'vcodec': formats_info[0].get('vcodec'),
1140 'vbr': formats_info[0].get('vbr'),
1141 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1142 'acodec': formats_info[1].get('acodec'),
1143 'abr': formats_info[1].get('abr'),
1146 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1148 def selector_function(formats):
1149 formats = list(formats)
1150 for pair in itertools.product(video_selector(formats), audio_selector(formats)):
1153 filters = [self._build_format_filter(f) for f in selector.filters]
1155 def final_selector(formats):
1156 for _filter in filters:
1157 formats = list(filter(_filter, formats))
1158 return selector_function(formats)
1159 return final_selector
1161 stream = io.BytesIO(format_spec.encode('utf-8'))
1163 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1164 except tokenize.TokenError:
1165 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1167 class TokenIterator(object):
1168 def __init__(self, tokens):
1169 self.tokens = tokens
1176 if self.counter >= len(self.tokens):
1177 raise StopIteration()
1178 value = self.tokens[self.counter]
1184 def restore_last_token(self):
1187 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1188 return _build_selector_function(parsed_selector)
1190 def _calc_headers(self, info_dict):
1191 res = std_headers.copy()
1193 add_headers = info_dict.get('http_headers')
1195 res.update(add_headers)
1197 cookies = self._calc_cookies(info_dict)
1199 res['Cookie'] = cookies
1203 def _calc_cookies(self, info_dict):
1204 pr = sanitized_Request(info_dict['url'])
1205 self.cookiejar.add_cookie_header(pr)
1206 return pr.get_header('Cookie')
1208 def process_video_result(self, info_dict, download=True):
1209 assert info_dict.get('_type', 'video') == 'video'
1211 if 'id' not in info_dict:
1212 raise ExtractorError('Missing "id" field in extractor result')
1213 if 'title' not in info_dict:
1214 raise ExtractorError('Missing "title" field in extractor result')
1216 if 'playlist' not in info_dict:
1217 # It isn't part of a playlist
1218 info_dict['playlist'] = None
1219 info_dict['playlist_index'] = None
1221 thumbnails = info_dict.get('thumbnails')
1222 if thumbnails is None:
1223 thumbnail = info_dict.get('thumbnail')
1225 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1227 thumbnails.sort(key=lambda t: (
1228 t.get('preference'), t.get('width'), t.get('height'),
1229 t.get('id'), t.get('url')))
1230 for i, t in enumerate(thumbnails):
1231 if t.get('width') and t.get('height'):
1232 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1233 if t.get('id') is None:
1236 if thumbnails and 'thumbnail' not in info_dict:
1237 info_dict['thumbnail'] = thumbnails[-1]['url']
1239 if 'display_id' not in info_dict and 'id' in info_dict:
1240 info_dict['display_id'] = info_dict['id']
1242 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1243 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1244 # see http://bugs.python.org/issue1646728)
1246 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1247 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1248 except (ValueError, OverflowError, OSError):
1251 # Auto generate title fields corresponding to the *_number fields when missing
1252 # in order to always have clean titles. This is very common for TV series.
1253 for field in ('chapter', 'season', 'episode'):
1254 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1255 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1257 subtitles = info_dict.get('subtitles')
1259 for _, subtitle in subtitles.items():
1260 for subtitle_format in subtitle:
1261 if 'ext' not in subtitle_format:
1262 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1264 if self.params.get('listsubtitles', False):
1265 if 'automatic_captions' in info_dict:
1266 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1267 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1269 info_dict['requested_subtitles'] = self.process_subtitles(
1270 info_dict['id'], subtitles,
1271 info_dict.get('automatic_captions'))
1273 # We now pick which formats have to be downloaded
1274 if info_dict.get('formats') is None:
1275 # There's only one format available
1276 formats = [info_dict]
1278 formats = info_dict['formats']
1281 raise ExtractorError('No video formats found!')
1285 # We check that all the formats have the format and format_id fields
1286 for i, format in enumerate(formats):
1287 if 'url' not in format:
1288 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1290 if format.get('format_id') is None:
1291 format['format_id'] = compat_str(i)
1293 # Sanitize format_id from characters used in format selector expression
1294 format['format_id'] = re.sub('[\s,/+\[\]()]', '_', format['format_id'])
1295 format_id = format['format_id']
1296 if format_id not in formats_dict:
1297 formats_dict[format_id] = []
1298 formats_dict[format_id].append(format)
1300 # Make sure all formats have unique format_id
1301 for format_id, ambiguous_formats in formats_dict.items():
1302 if len(ambiguous_formats) > 1:
1303 for i, format in enumerate(ambiguous_formats):
1304 format['format_id'] = '%s-%d' % (format_id, i)
1306 for i, format in enumerate(formats):
1307 if format.get('format') is None:
1308 format['format'] = '{id} - {res}{note}'.format(
1309 id=format['format_id'],
1310 res=self.format_resolution(format),
1311 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1313 # Automatically determine file extension if missing
1314 if 'ext' not in format:
1315 format['ext'] = determine_ext(format['url']).lower()
1316 # Automatically determine protocol if missing (useful for format
1317 # selection purposes)
1318 if 'protocol' not in format:
1319 format['protocol'] = determine_protocol(format)
1320 # Add HTTP headers, so that external programs can use them from the
1322 full_format_info = info_dict.copy()
1323 full_format_info.update(format)
1324 format['http_headers'] = self._calc_headers(full_format_info)
1326 # TODO Central sorting goes here
1328 if formats[0] is not info_dict:
1329 # only set the 'formats' fields if the original info_dict list them
1330 # otherwise we end up with a circular reference, the first (and unique)
1331 # element in the 'formats' field in info_dict is info_dict itself,
1332 # which can't be exported to json
1333 info_dict['formats'] = formats
1334 if self.params.get('listformats'):
1335 self.list_formats(info_dict)
1337 if self.params.get('list_thumbnails'):
1338 self.list_thumbnails(info_dict)
1341 req_format = self.params.get('format')
1342 if req_format is None:
1343 req_format_list = []
1344 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1345 not info_dict.get('is_live')):
1346 merger = FFmpegMergerPP(self)
1347 if merger.available and merger.can_merge():
1348 req_format_list.append('bestvideo+bestaudio')
1349 req_format_list.append('best')
1350 req_format = '/'.join(req_format_list)
1351 format_selector = self.build_format_selector(req_format)
1352 formats_to_download = list(format_selector(formats))
1353 if not formats_to_download:
1354 raise ExtractorError('requested format not available',
1358 if len(formats_to_download) > 1:
1359 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1360 for format in formats_to_download:
1361 new_info = dict(info_dict)
1362 new_info.update(format)
1363 self.process_info(new_info)
1364 # We update the info dict with the best quality format (backwards compatibility)
1365 info_dict.update(formats_to_download[-1])
1368 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1369 """Select the requested subtitles and their format"""
1371 if normal_subtitles and self.params.get('writesubtitles'):
1372 available_subs.update(normal_subtitles)
1373 if automatic_captions and self.params.get('writeautomaticsub'):
1374 for lang, cap_info in automatic_captions.items():
1375 if lang not in available_subs:
1376 available_subs[lang] = cap_info
1378 if (not self.params.get('writesubtitles') and not
1379 self.params.get('writeautomaticsub') or not
1383 if self.params.get('allsubtitles', False):
1384 requested_langs = available_subs.keys()
1386 if self.params.get('subtitleslangs', False):
1387 requested_langs = self.params.get('subtitleslangs')
1388 elif 'en' in available_subs:
1389 requested_langs = ['en']
1391 requested_langs = [list(available_subs.keys())[0]]
1393 formats_query = self.params.get('subtitlesformat', 'best')
1394 formats_preference = formats_query.split('/') if formats_query else []
1396 for lang in requested_langs:
1397 formats = available_subs.get(lang)
1399 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1401 for ext in formats_preference:
1405 matches = list(filter(lambda f: f['ext'] == ext, formats))
1411 self.report_warning(
1412 'No subtitle format found matching "%s" for language %s, '
1413 'using %s' % (formats_query, lang, f['ext']))
1417 def process_info(self, info_dict):
1418 """Process a single resolved IE result."""
1420 assert info_dict.get('_type', 'video') == 'video'
1422 max_downloads = self.params.get('max_downloads')
1423 if max_downloads is not None:
1424 if self._num_downloads >= int(max_downloads):
1425 raise MaxDownloadsReached()
1427 info_dict['fulltitle'] = info_dict['title']
1428 if len(info_dict['title']) > 200:
1429 info_dict['title'] = info_dict['title'][:197] + '...'
1431 if 'format' not in info_dict:
1432 info_dict['format'] = info_dict['ext']
1434 reason = self._match_entry(info_dict, incomplete=False)
1435 if reason is not None:
1436 self.to_screen('[download] ' + reason)
1439 self._num_downloads += 1
1441 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1444 if self.params.get('forcetitle', False):
1445 self.to_stdout(info_dict['fulltitle'])
1446 if self.params.get('forceid', False):
1447 self.to_stdout(info_dict['id'])
1448 if self.params.get('forceurl', False):
1449 if info_dict.get('requested_formats') is not None:
1450 for f in info_dict['requested_formats']:
1451 self.to_stdout(f['url'] + f.get('play_path', ''))
1453 # For RTMP URLs, also include the playpath
1454 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1455 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1456 self.to_stdout(info_dict['thumbnail'])
1457 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1458 self.to_stdout(info_dict['description'])
1459 if self.params.get('forcefilename', False) and filename is not None:
1460 self.to_stdout(filename)
1461 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1462 self.to_stdout(formatSeconds(info_dict['duration']))
1463 if self.params.get('forceformat', False):
1464 self.to_stdout(info_dict['format'])
1465 if self.params.get('forcejson', False):
1466 self.to_stdout(json.dumps(info_dict))
1468 # Do nothing else if in simulate mode
1469 if self.params.get('simulate', False):
1472 if filename is None:
1476 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1477 if dn and not os.path.exists(dn):
1479 except (OSError, IOError) as err:
1480 self.report_error('unable to create directory ' + error_to_compat_str(err))
1483 if self.params.get('writedescription', False):
1484 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1485 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1486 self.to_screen('[info] Video description is already present')
1487 elif info_dict.get('description') is None:
1488 self.report_warning('There\'s no description to write.')
1491 self.to_screen('[info] Writing video description to: ' + descfn)
1492 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1493 descfile.write(info_dict['description'])
1494 except (OSError, IOError):
1495 self.report_error('Cannot write description file ' + descfn)
1498 if self.params.get('writeannotations', False):
1499 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1500 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1501 self.to_screen('[info] Video annotations are already present')
1504 self.to_screen('[info] Writing video annotations to: ' + annofn)
1505 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1506 annofile.write(info_dict['annotations'])
1507 except (KeyError, TypeError):
1508 self.report_warning('There are no annotations to write.')
1509 except (OSError, IOError):
1510 self.report_error('Cannot write annotations file: ' + annofn)
1513 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1514 self.params.get('writeautomaticsub')])
1516 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1517 # subtitles download errors are already managed as troubles in relevant IE
1518 # that way it will silently go on when used with unsupporting IE
1519 subtitles = info_dict['requested_subtitles']
1520 ie = self.get_info_extractor(info_dict['extractor_key'])
1521 for sub_lang, sub_info in subtitles.items():
1522 sub_format = sub_info['ext']
1523 if sub_info.get('data') is not None:
1524 sub_data = sub_info['data']
1527 sub_data = ie._download_webpage(
1528 sub_info['url'], info_dict['id'], note=False)
1529 except ExtractorError as err:
1530 self.report_warning('Unable to download subtitle for "%s": %s' %
1531 (sub_lang, error_to_compat_str(err.cause)))
1534 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1535 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1536 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1538 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1539 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1540 subfile.write(sub_data)
1541 except (OSError, IOError):
1542 self.report_error('Cannot write subtitles file ' + sub_filename)
1545 if self.params.get('writeinfojson', False):
1546 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1547 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1548 self.to_screen('[info] Video description metadata is already present')
1550 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1552 write_json_file(self.filter_requested_info(info_dict), infofn)
1553 except (OSError, IOError):
1554 self.report_error('Cannot write metadata to JSON file ' + infofn)
1557 self._write_thumbnails(info_dict, filename)
1559 if not self.params.get('skip_download', False):
1562 fd = get_suitable_downloader(info, self.params)(self, self.params)
1563 for ph in self._progress_hooks:
1564 fd.add_progress_hook(ph)
1565 if self.params.get('verbose'):
1566 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1567 return fd.download(name, info)
1569 if info_dict.get('requested_formats') is not None:
1572 merger = FFmpegMergerPP(self)
1573 if not merger.available:
1575 self.report_warning('You have requested multiple '
1576 'formats but ffmpeg or avconv are not installed.'
1577 ' The formats won\'t be merged.')
1579 postprocessors = [merger]
1581 def compatible_formats(formats):
1582 video, audio = formats
1584 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1585 if video_ext and audio_ext:
1587 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1590 for exts in COMPATIBLE_EXTS:
1591 if video_ext in exts and audio_ext in exts:
1593 # TODO: Check acodec/vcodec
1596 filename_real_ext = os.path.splitext(filename)[1][1:]
1598 os.path.splitext(filename)[0]
1599 if filename_real_ext == info_dict['ext']
1601 requested_formats = info_dict['requested_formats']
1602 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1603 info_dict['ext'] = 'mkv'
1604 self.report_warning(
1605 'Requested formats are incompatible for merge and will be merged into mkv.')
1606 # Ensure filename always has a correct extension for successful merge
1607 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1608 if os.path.exists(encodeFilename(filename)):
1610 '[download] %s has already been downloaded and '
1611 'merged' % filename)
1613 for f in requested_formats:
1614 new_info = dict(info_dict)
1616 fname = self.prepare_filename(new_info)
1617 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1618 downloaded.append(fname)
1619 partial_success = dl(fname, new_info)
1620 success = success and partial_success
1621 info_dict['__postprocessors'] = postprocessors
1622 info_dict['__files_to_merge'] = downloaded
1624 # Just a single file
1625 success = dl(filename, info_dict)
1626 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1627 self.report_error('unable to download video data: %s' % str(err))
1629 except (OSError, IOError) as err:
1630 raise UnavailableVideoError(err)
1631 except (ContentTooShortError, ) as err:
1632 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1635 if success and filename != '-':
1637 fixup_policy = self.params.get('fixup')
1638 if fixup_policy is None:
1639 fixup_policy = 'detect_or_warn'
1641 stretched_ratio = info_dict.get('stretched_ratio')
1642 if stretched_ratio is not None and stretched_ratio != 1:
1643 if fixup_policy == 'warn':
1644 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1645 info_dict['id'], stretched_ratio))
1646 elif fixup_policy == 'detect_or_warn':
1647 stretched_pp = FFmpegFixupStretchedPP(self)
1648 if stretched_pp.available:
1649 info_dict.setdefault('__postprocessors', [])
1650 info_dict['__postprocessors'].append(stretched_pp)
1652 self.report_warning(
1653 '%s: Non-uniform pixel ratio (%s). Install ffmpeg or avconv to fix this automatically.' % (
1654 info_dict['id'], stretched_ratio))
1656 assert fixup_policy in ('ignore', 'never')
1658 if info_dict.get('requested_formats') is None and info_dict.get('container') == 'm4a_dash':
1659 if fixup_policy == 'warn':
1660 self.report_warning('%s: writing DASH m4a. Only some players support this container.' % (
1662 elif fixup_policy == 'detect_or_warn':
1663 fixup_pp = FFmpegFixupM4aPP(self)
1664 if fixup_pp.available:
1665 info_dict.setdefault('__postprocessors', [])
1666 info_dict['__postprocessors'].append(fixup_pp)
1668 self.report_warning(
1669 '%s: writing DASH m4a. Only some players support this container. Install ffmpeg or avconv to fix this automatically.' % (
1672 assert fixup_policy in ('ignore', 'never')
1675 self.post_process(filename, info_dict)
1676 except (PostProcessingError) as err:
1677 self.report_error('postprocessing: %s' % str(err))
1679 self.record_download_archive(info_dict)
1681 def download(self, url_list):
1682 """Download a given list of URLs."""
1683 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1684 if (len(url_list) > 1 and
1685 '%' not in outtmpl and
1686 self.params.get('max_downloads') != 1):
1687 raise SameFileError(outtmpl)
1689 for url in url_list:
1691 # It also downloads the videos
1692 res = self.extract_info(
1693 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1694 except UnavailableVideoError:
1695 self.report_error('unable to download video')
1696 except MaxDownloadsReached:
1697 self.to_screen('[info] Maximum number of downloaded files reached.')
1700 if self.params.get('dump_single_json', False):
1701 self.to_stdout(json.dumps(res))
1703 return self._download_retcode
1705 def download_with_info_file(self, info_filename):
1706 with contextlib.closing(fileinput.FileInput(
1707 [info_filename], mode='r',
1708 openhook=fileinput.hook_encoded('utf-8'))) as f:
1709 # FileInput doesn't have a read method, we can't call json.load
1710 info = self.filter_requested_info(json.loads('\n'.join(f)))
1712 self.process_ie_result(info, download=True)
1713 except DownloadError:
1714 webpage_url = info.get('webpage_url')
1715 if webpage_url is not None:
1716 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1717 return self.download([webpage_url])
1720 return self._download_retcode
1723 def filter_requested_info(info_dict):
1725 (k, v) for k, v in info_dict.items()
1726 if k not in ['requested_formats', 'requested_subtitles'])
1728 def post_process(self, filename, ie_info):
1729 """Run all the postprocessors on the given file."""
1730 info = dict(ie_info)
1731 info['filepath'] = filename
1733 if ie_info.get('__postprocessors') is not None:
1734 pps_chain.extend(ie_info['__postprocessors'])
1735 pps_chain.extend(self._pps)
1736 for pp in pps_chain:
1737 files_to_delete = []
1739 files_to_delete, info = pp.run(info)
1740 except PostProcessingError as e:
1741 self.report_error(e.msg)
1742 if files_to_delete and not self.params.get('keepvideo', False):
1743 for old_filename in files_to_delete:
1744 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1746 os.remove(encodeFilename(old_filename))
1747 except (IOError, OSError):
1748 self.report_warning('Unable to remove downloaded original file')
1750 def _make_archive_id(self, info_dict):
1751 # Future-proof against any change in case
1752 # and backwards compatibility with prior versions
1753 extractor = info_dict.get('extractor_key')
1754 if extractor is None:
1755 if 'id' in info_dict:
1756 extractor = info_dict.get('ie_key') # key in a playlist
1757 if extractor is None:
1758 return None # Incomplete video information
1759 return extractor.lower() + ' ' + info_dict['id']
1761 def in_download_archive(self, info_dict):
1762 fn = self.params.get('download_archive')
1766 vid_id = self._make_archive_id(info_dict)
1768 return False # Incomplete video information
1771 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1772 for line in archive_file:
1773 if line.strip() == vid_id:
1775 except IOError as ioe:
1776 if ioe.errno != errno.ENOENT:
1780 def record_download_archive(self, info_dict):
1781 fn = self.params.get('download_archive')
1784 vid_id = self._make_archive_id(info_dict)
1786 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1787 archive_file.write(vid_id + '\n')
1790 def format_resolution(format, default='unknown'):
1791 if format.get('vcodec') == 'none':
1793 if format.get('resolution') is not None:
1794 return format['resolution']
1795 if format.get('height') is not None:
1796 if format.get('width') is not None:
1797 res = '%sx%s' % (format['width'], format['height'])
1799 res = '%sp' % format['height']
1800 elif format.get('width') is not None:
1801 res = '%dx?' % format['width']
1806 def _format_note(self, fdict):
1808 if fdict.get('ext') in ['f4f', 'f4m']:
1809 res += '(unsupported) '
1810 if fdict.get('language'):
1813 res += '[%s]' % fdict['language']
1814 if fdict.get('format_note') is not None:
1815 res += fdict['format_note'] + ' '
1816 if fdict.get('tbr') is not None:
1817 res += '%4dk ' % fdict['tbr']
1818 if fdict.get('container') is not None:
1821 res += '%s container' % fdict['container']
1822 if (fdict.get('vcodec') is not None and
1823 fdict.get('vcodec') != 'none'):
1826 res += fdict['vcodec']
1827 if fdict.get('vbr') is not None:
1829 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1831 if fdict.get('vbr') is not None:
1832 res += '%4dk' % fdict['vbr']
1833 if fdict.get('fps') is not None:
1834 res += ', %sfps' % fdict['fps']
1835 if fdict.get('acodec') is not None:
1838 if fdict['acodec'] == 'none':
1841 res += '%-5s' % fdict['acodec']
1842 elif fdict.get('abr') is not None:
1846 if fdict.get('abr') is not None:
1847 res += '@%3dk' % fdict['abr']
1848 if fdict.get('asr') is not None:
1849 res += ' (%5dHz)' % fdict['asr']
1850 if fdict.get('filesize') is not None:
1853 res += format_bytes(fdict['filesize'])
1854 elif fdict.get('filesize_approx') is not None:
1857 res += '~' + format_bytes(fdict['filesize_approx'])
1860 def list_formats(self, info_dict):
1861 formats = info_dict.get('formats', [info_dict])
1863 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1865 if f.get('preference') is None or f['preference'] >= -1000]
1866 if len(formats) > 1:
1867 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1869 header_line = ['format code', 'extension', 'resolution', 'note']
1871 '[info] Available formats for %s:\n%s' %
1872 (info_dict['id'], render_table(header_line, table)))
1874 def list_thumbnails(self, info_dict):
1875 thumbnails = info_dict.get('thumbnails')
1877 tn_url = info_dict.get('thumbnail')
1879 thumbnails = [{'id': '0', 'url': tn_url}]
1882 '[info] No thumbnails present for %s' % info_dict['id'])
1886 '[info] Thumbnails for %s:' % info_dict['id'])
1887 self.to_screen(render_table(
1888 ['ID', 'width', 'height', 'URL'],
1889 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1891 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1893 self.to_screen('%s has no %s' % (video_id, name))
1896 'Available %s for %s:' % (name, video_id))
1897 self.to_screen(render_table(
1898 ['Language', 'formats'],
1899 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1900 for lang, formats in subtitles.items()]))
1902 def urlopen(self, req):
1903 """ Start an HTTP download """
1904 if isinstance(req, compat_basestring):
1905 req = sanitized_Request(req)
1906 return self._opener.open(req, timeout=self._socket_timeout)
1908 def print_debug_header(self):
1909 if not self.params.get('verbose'):
1912 if type('') is not compat_str:
1913 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1914 self.report_warning(
1915 'Your Python is broken! Update to a newer and supported version')
1917 stdout_encoding = getattr(
1918 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1920 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1921 locale.getpreferredencoding(),
1922 sys.getfilesystemencoding(),
1924 self.get_encoding()))
1925 write_string(encoding_str, encoding=None)
1927 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1929 sp = subprocess.Popen(
1930 ['git', 'rev-parse', '--short', 'HEAD'],
1931 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1932 cwd=os.path.dirname(os.path.abspath(__file__)))
1933 out, err = sp.communicate()
1934 out = out.decode().strip()
1935 if re.match('[0-9a-f]+', out):
1936 self._write_string('[debug] Git HEAD: ' + out + '\n')
1942 self._write_string('[debug] Python version %s - %s\n' % (
1943 platform.python_version(), platform_name()))
1945 exe_versions = FFmpegPostProcessor.get_versions(self)
1946 exe_versions['rtmpdump'] = rtmpdump_version()
1947 exe_str = ', '.join(
1949 for exe, v in sorted(exe_versions.items())
1954 self._write_string('[debug] exe versions: %s\n' % exe_str)
1957 for handler in self._opener.handlers:
1958 if hasattr(handler, 'proxies'):
1959 proxy_map.update(handler.proxies)
1960 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1962 if self.params.get('call_home', False):
1963 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
1964 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
1965 latest_version = self.urlopen(
1966 'https://yt-dl.org/latest/version').read().decode('utf-8')
1967 if version_tuple(latest_version) > version_tuple(__version__):
1968 self.report_warning(
1969 'You are using an outdated version (newest version: %s)! '
1970 'See https://yt-dl.org/update if you need help updating.' %
1973 def _setup_opener(self):
1974 timeout_val = self.params.get('socket_timeout')
1975 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
1977 opts_cookiefile = self.params.get('cookiefile')
1978 opts_proxy = self.params.get('proxy')
1980 if opts_cookiefile is None:
1981 self.cookiejar = compat_cookiejar.CookieJar()
1983 self.cookiejar = compat_cookiejar.MozillaCookieJar(
1985 if os.access(opts_cookiefile, os.R_OK):
1986 self.cookiejar.load()
1988 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
1989 if opts_proxy is not None:
1990 if opts_proxy == '':
1993 proxies = {'http': opts_proxy, 'https': opts_proxy}
1995 proxies = compat_urllib_request.getproxies()
1996 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
1997 if 'http' in proxies and 'https' not in proxies:
1998 proxies['https'] = proxies['http']
1999 proxy_handler = PerRequestProxyHandler(proxies)
2001 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2002 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2003 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2004 data_handler = compat_urllib_request_DataHandler()
2006 # When passing our own FileHandler instance, build_opener won't add the
2007 # default FileHandler and allows us to disable the file protocol, which
2008 # can be used for malicious purposes (see
2009 # https://github.com/rg3/youtube-dl/issues/8227)
2010 file_handler = compat_urllib_request.FileHandler()
2012 def file_open(*args, **kwargs):
2013 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2014 file_handler.file_open = file_open
2016 opener = compat_urllib_request.build_opener(
2017 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2019 # Delete the default user-agent header, which would otherwise apply in
2020 # cases where our custom HTTP handler doesn't come into play
2021 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2022 opener.addheaders = []
2023 self._opener = opener
2025 def encode(self, s):
2026 if isinstance(s, bytes):
2027 return s # Already encoded
2030 return s.encode(self.get_encoding())
2031 except UnicodeEncodeError as err:
2032 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2035 def get_encoding(self):
2036 encoding = self.params.get('encoding')
2037 if encoding is None:
2038 encoding = preferredencoding()
2041 def _write_thumbnails(self, info_dict, filename):
2042 if self.params.get('writethumbnail', False):
2043 thumbnails = info_dict.get('thumbnails')
2045 thumbnails = [thumbnails[-1]]
2046 elif self.params.get('write_all_thumbnails', False):
2047 thumbnails = info_dict.get('thumbnails')
2052 # No thumbnails present, so return immediately
2055 for t in thumbnails:
2056 thumb_ext = determine_ext(t['url'], 'jpg')
2057 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2058 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2059 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2061 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2062 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2063 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2065 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2066 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2068 uf = self.urlopen(t['url'])
2069 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2070 shutil.copyfileobj(uf, thumbf)
2071 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2072 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2073 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2074 self.report_warning('Unable to download thumbnail "%s": %s' %
2075 (t['url'], error_to_compat_str(err)))