2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import, unicode_literals
34 compat_get_terminal_size,
38 compat_tokenize_tokenize,
40 compat_urllib_request,
41 compat_urllib_request_DataHandler,
61 PerRequestProxyHandler,
72 UnavailableVideoError,
77 YoutubeDLCookieProcessor,
84 from .cache import Cache
85 from .extractor import get_info_extractor, gen_extractors
86 from .downloader import get_suitable_downloader
87 from .downloader.rtmp import rtmpdump_version
88 from .postprocessor import (
90 FFmpegFixupStretchedPP,
95 from .version import __version__
98 class YoutubeDL(object):
101 YoutubeDL objects are the ones responsible of downloading the
102 actual video file and writing it to disk if the user has requested
103 it, among some other tasks. In most cases there should be one per
104 program. As, given a video URL, the downloader doesn't know how to
105 extract all the needed information, task that InfoExtractors do, it
106 has to pass the URL to one of them.
108 For this, YoutubeDL objects have a method that allows
109 InfoExtractors to be registered in a given order. When it is passed
110 a URL, the YoutubeDL object handles it to the first InfoExtractor it
111 finds that reports being able to handle it. The InfoExtractor extracts
112 all the information about the video or videos the URL refers to, and
113 YoutubeDL process the extracted information, possibly using a File
114 Downloader to download the video.
116 YoutubeDL objects accept a lot of parameters. In order not to saturate
117 the object constructor with arguments, it receives a dictionary of
118 options instead. These options are available through the params
119 attribute for the InfoExtractors to use. The YoutubeDL also
120 registers itself as the downloader in charge for the InfoExtractors
121 that are added to it, so this is a "mutual registration".
125 username: Username for authentication purposes.
126 password: Password for authentication purposes.
127 videopassword: Password for accessing a video.
128 usenetrc: Use netrc for authentication instead.
129 verbose: Print additional info to stdout.
130 quiet: Do not print messages to stdout.
131 no_warnings: Do not print out anything for warnings.
132 forceurl: Force printing final URL.
133 forcetitle: Force printing title.
134 forceid: Force printing ID.
135 forcethumbnail: Force printing thumbnail URL.
136 forcedescription: Force printing description.
137 forcefilename: Force printing final filename.
138 forceduration: Force printing duration.
139 forcejson: Force printing info_dict as JSON.
140 dump_single_json: Force printing the info_dict of the whole playlist
141 (or video) as a single JSON line.
142 simulate: Do not download the video files.
143 format: Video format code. See options.py for more information.
144 outtmpl: Template for output names.
145 restrictfilenames: Do not allow "&" and spaces in file names
146 ignoreerrors: Do not stop on download errors.
147 force_generic_extractor: Force downloader to use the generic extractor
148 nooverwrites: Prevent overwriting files.
149 playliststart: Playlist item to start at.
150 playlistend: Playlist item to end at.
151 playlist_items: Specific indices of playlist to download.
152 playlistreverse: Download playlist items in reverse order.
153 matchtitle: Download only matching titles.
154 rejecttitle: Reject downloads for matching titles.
155 logger: Log messages to a logging.Logger instance.
156 logtostderr: Log messages to stderr instead of stdout.
157 writedescription: Write the video description to a .description file
158 writeinfojson: Write the video description to a .info.json file
159 writeannotations: Write the video annotations to a .annotations.xml file
160 writethumbnail: Write the thumbnail image to a file
161 write_all_thumbnails: Write all thumbnail formats to files
162 writesubtitles: Write the video subtitles to a file
163 writeautomaticsub: Write the automatically generated subtitles to a file
164 allsubtitles: Downloads all the subtitles of the video
165 (requires writesubtitles or writeautomaticsub)
166 listsubtitles: Lists all available subtitles for the video
167 subtitlesformat: The format code for subtitles
168 subtitleslangs: List of languages of the subtitles to download
169 keepvideo: Keep the video file after post-processing
170 daterange: A DateRange object, download only if the upload_date is in the range.
171 skip_download: Skip the actual download of the video file
172 cachedir: Location of the cache files in the filesystem.
173 False to disable filesystem cache.
174 noplaylist: Download single video instead of a playlist if in doubt.
175 age_limit: An integer representing the user's age in years.
176 Unsuitable videos for the given age are skipped.
177 min_views: An integer representing the minimum view count the video
178 must have in order to not be skipped.
179 Videos without view count information are always
180 downloaded. None for no limit.
181 max_views: An integer representing the maximum view count.
182 Videos that are more popular than that are not
184 Videos without view count information are always
185 downloaded. None for no limit.
186 download_archive: File name of a file where all downloads are recorded.
187 Videos already present in the file are not downloaded
189 cookiefile: File name where cookies should be read from and dumped to.
190 nocheckcertificate:Do not verify SSL certificates
191 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
192 At the moment, this is only supported by YouTube.
193 proxy: URL of the proxy server to use
194 cn_verification_proxy: URL of the proxy to use for IP address verification
195 on Chinese sites. (Experimental)
196 socket_timeout: Time to wait for unresponsive hosts, in seconds
197 bidi_workaround: Work around buggy terminals without bidirectional text
198 support, using fridibi
199 debug_printtraffic:Print out sent and received HTTP traffic
200 include_ads: Download ads as well
201 default_search: Prepend this string if an input url is not valid.
202 'auto' for elaborate guessing
203 encoding: Use this encoding instead of the system-specified.
204 extract_flat: Do not resolve URLs, return the immediate result.
205 Pass in 'in_playlist' to only show this behavior for
207 postprocessors: A list of dictionaries, each with an entry
208 * key: The name of the postprocessor. See
209 youtube_dl/postprocessor/__init__.py for a list.
210 as well as any further keyword arguments for the
212 progress_hooks: A list of functions that get called on download
213 progress, with a dictionary with the entries
214 * status: One of "downloading", "error", or "finished".
215 Check this first and ignore unknown values.
217 If status is one of "downloading", or "finished", the
218 following properties may also be present:
219 * filename: The final filename (always present)
220 * tmpfilename: The filename we're currently writing to
221 * downloaded_bytes: Bytes on disk
222 * total_bytes: Size of the whole file, None if unknown
223 * total_bytes_estimate: Guess of the eventual file size,
225 * elapsed: The number of seconds since download started.
226 * eta: The estimated time in seconds, None if unknown
227 * speed: The download speed in bytes/second, None if
229 * fragment_index: The counter of the currently
230 downloaded video fragment.
231 * fragment_count: The number of fragments (= individual
232 files that will be merged)
234 Progress hooks are guaranteed to be called at least once
235 (with status "finished") if the download is successful.
236 merge_output_format: Extension to use when merging formats.
237 fixup: Automatically correct known faults of the file.
239 - "never": do nothing
240 - "warn": only emit a warning
241 - "detect_or_warn": check whether we can do anything
242 about it, warn otherwise (default)
243 source_address: (Experimental) Client-side IP address to bind to.
244 call_home: Boolean, true iff we are allowed to contact the
245 youtube-dl servers for debugging.
246 sleep_interval: Number of seconds to sleep before each download.
247 listformats: Print an overview of available video formats and exit.
248 list_thumbnails: Print a table of all thumbnails and exit.
249 match_filter: A function that gets called with the info_dict of
251 If it returns a message, the video is ignored.
252 If it returns None, the video is downloaded.
253 match_filter_func in utils.py is one example for this.
254 no_color: Do not emit color codes in output.
256 The following options determine which downloader is picked:
257 external_downloader: Executable of the external downloader to call.
258 None or unset for standard (built-in) downloader.
259 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv.
261 The following parameters are not used by YoutubeDL itself, they are used by
262 the downloader (see youtube_dl/downloader/common.py):
263 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
264 noresizebuffer, retries, continuedl, noprogress, consoletitle,
265 xattr_set_filesize, external_downloader_args.
267 The following options are used by the post processors:
268 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
269 otherwise prefer avconv.
270 postprocessor_args: A list of additional command-line arguments for the
277 _download_retcode = None
278 _num_downloads = None
281 def __init__(self, params=None, auto_init=True):
282 """Create a FileDownloader object with the given options."""
286 self._ies_instances = {}
288 self._progress_hooks = []
289 self._download_retcode = 0
290 self._num_downloads = 0
291 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
292 self._err_file = sys.stderr
295 'nocheckcertificate': False,
297 self.params.update(params)
298 self.cache = Cache(self)
300 if params.get('bidi_workaround', False):
303 master, slave = pty.openpty()
304 width = compat_get_terminal_size().columns
308 width_args = ['-w', str(width)]
310 stdin=subprocess.PIPE,
312 stderr=self._err_file)
314 self._output_process = subprocess.Popen(
315 ['bidiv'] + width_args, **sp_kwargs
318 self._output_process = subprocess.Popen(
319 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
320 self._output_channel = os.fdopen(master, 'rb')
321 except OSError as ose:
323 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 $PATH.')
327 if (sys.version_info >= (3,) and sys.platform != 'win32' and
328 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
329 not params.get('restrictfilenames', False)):
330 # On Python 3, the Unicode filesystem API will throw errors (#1474)
332 'Assuming --restrict-filenames since file system encoding '
333 'cannot encode all characters. '
334 'Set the LC_ALL environment variable to fix this.')
335 self.params['restrictfilenames'] = True
337 if isinstance(params.get('outtmpl'), bytes):
339 'Parameter outtmpl is bytes, but should be a unicode string. '
340 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
345 self.print_debug_header()
346 self.add_default_info_extractors()
348 for pp_def_raw in self.params.get('postprocessors', []):
349 pp_class = get_postprocessor(pp_def_raw['key'])
350 pp_def = dict(pp_def_raw)
352 pp = pp_class(self, **compat_kwargs(pp_def))
353 self.add_post_processor(pp)
355 for ph in self.params.get('progress_hooks', []):
356 self.add_progress_hook(ph)
358 def warn_if_short_id(self, argv):
359 # short YouTube ID starting with dash?
361 i for i, a in enumerate(argv)
362 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
366 [a for i, a in enumerate(argv) if i not in idxs] +
367 ['--'] + [argv[i] for i in idxs]
370 'Long argument string detected. '
371 'Use -- to separate parameters and URLs, like this:\n%s\n' %
372 args_to_str(correct_argv))
374 def add_info_extractor(self, ie):
375 """Add an InfoExtractor object to the end of the list."""
377 self._ies_instances[ie.ie_key()] = ie
378 ie.set_downloader(self)
380 def get_info_extractor(self, ie_key):
382 Get an instance of an IE with name ie_key, it will try to get one from
383 the _ies list, if there's no instance it will create a new one and add
384 it to the extractor list.
386 ie = self._ies_instances.get(ie_key)
388 ie = get_info_extractor(ie_key)()
389 self.add_info_extractor(ie)
392 def add_default_info_extractors(self):
394 Add the InfoExtractors returned by gen_extractors to the end of the list
396 for ie in gen_extractors():
397 self.add_info_extractor(ie)
399 def add_post_processor(self, pp):
400 """Add a PostProcessor object to the end of the chain."""
402 pp.set_downloader(self)
404 def add_progress_hook(self, ph):
405 """Add the progress hook (currently only for the file downloader)"""
406 self._progress_hooks.append(ph)
408 def _bidi_workaround(self, message):
409 if not hasattr(self, '_output_channel'):
412 assert hasattr(self, '_output_process')
413 assert isinstance(message, compat_str)
414 line_count = message.count('\n') + 1
415 self._output_process.stdin.write((message + '\n').encode('utf-8'))
416 self._output_process.stdin.flush()
417 res = ''.join(self._output_channel.readline().decode('utf-8')
418 for _ in range(line_count))
419 return res[:-len('\n')]
421 def to_screen(self, message, skip_eol=False):
422 """Print message to stdout if not in quiet mode."""
423 return self.to_stdout(message, skip_eol, check_quiet=True)
425 def _write_string(self, s, out=None):
426 write_string(s, out=out, encoding=self.params.get('encoding'))
428 def to_stdout(self, message, skip_eol=False, check_quiet=False):
429 """Print message to stdout if not in quiet mode."""
430 if self.params.get('logger'):
431 self.params['logger'].debug(message)
432 elif not check_quiet or not self.params.get('quiet', False):
433 message = self._bidi_workaround(message)
434 terminator = ['\n', ''][skip_eol]
435 output = message + terminator
437 self._write_string(output, self._screen_file)
439 def to_stderr(self, message):
440 """Print message to stderr."""
441 assert isinstance(message, compat_str)
442 if self.params.get('logger'):
443 self.params['logger'].error(message)
445 message = self._bidi_workaround(message)
446 output = message + '\n'
447 self._write_string(output, self._err_file)
449 def to_console_title(self, message):
450 if not self.params.get('consoletitle', False):
452 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
453 # c_wchar_p() might not be necessary if `message` is
454 # already of type unicode()
455 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
456 elif 'TERM' in os.environ:
457 self._write_string('\033]0;%s\007' % message, self._screen_file)
459 def save_console_title(self):
460 if not self.params.get('consoletitle', False):
462 if 'TERM' in os.environ:
463 # Save the title on stack
464 self._write_string('\033[22;0t', self._screen_file)
466 def restore_console_title(self):
467 if not self.params.get('consoletitle', False):
469 if 'TERM' in os.environ:
470 # Restore the title from stack
471 self._write_string('\033[23;0t', self._screen_file)
474 self.save_console_title()
477 def __exit__(self, *args):
478 self.restore_console_title()
480 if self.params.get('cookiefile') is not None:
481 self.cookiejar.save()
483 def trouble(self, message=None, tb=None):
484 """Determine action to take when a download problem appears.
486 Depending on if the downloader has been configured to ignore
487 download errors or not, this method may throw an exception or
488 not when errors are found, after printing the message.
490 tb, if given, is additional traceback information.
492 if message is not None:
493 self.to_stderr(message)
494 if self.params.get('verbose'):
496 if sys.exc_info()[0]: # if .trouble has been called from an except block
498 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
499 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
500 tb += encode_compat_str(traceback.format_exc())
502 tb_data = traceback.format_list(traceback.extract_stack())
503 tb = ''.join(tb_data)
505 if not self.params.get('ignoreerrors', False):
506 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
507 exc_info = sys.exc_info()[1].exc_info
509 exc_info = sys.exc_info()
510 raise DownloadError(message, exc_info)
511 self._download_retcode = 1
513 def report_warning(self, message):
515 Print the message to stderr, it will be prefixed with 'WARNING:'
516 If stderr is a tty file the 'WARNING:' will be colored
518 if self.params.get('logger') is not None:
519 self.params['logger'].warning(message)
521 if self.params.get('no_warnings'):
523 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
524 _msg_header = '\033[0;33mWARNING:\033[0m'
526 _msg_header = 'WARNING:'
527 warning_message = '%s %s' % (_msg_header, message)
528 self.to_stderr(warning_message)
530 def report_error(self, message, tb=None):
532 Do the same as trouble, but prefixes the message with 'ERROR:', colored
533 in red if stderr is a tty file.
535 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
536 _msg_header = '\033[0;31mERROR:\033[0m'
538 _msg_header = 'ERROR:'
539 error_message = '%s %s' % (_msg_header, message)
540 self.trouble(error_message, tb)
542 def report_file_already_downloaded(self, file_name):
543 """Report file has already been fully downloaded."""
545 self.to_screen('[download] %s has already been downloaded' % file_name)
546 except UnicodeEncodeError:
547 self.to_screen('[download] The file has already been downloaded')
549 def prepare_filename(self, info_dict):
550 """Generate the output filename."""
552 template_dict = dict(info_dict)
554 template_dict['epoch'] = int(time.time())
555 autonumber_size = self.params.get('autonumber_size')
556 if autonumber_size is None:
558 autonumber_templ = '%0' + str(autonumber_size) + 'd'
559 template_dict['autonumber'] = autonumber_templ % self._num_downloads
560 if template_dict.get('playlist_index') is not None:
561 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
562 if template_dict.get('resolution') is None:
563 if template_dict.get('width') and template_dict.get('height'):
564 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
565 elif template_dict.get('height'):
566 template_dict['resolution'] = '%sp' % template_dict['height']
567 elif template_dict.get('width'):
568 template_dict['resolution'] = '?x%d' % template_dict['width']
570 sanitize = lambda k, v: sanitize_filename(
572 restricted=self.params.get('restrictfilenames'),
574 template_dict = dict((k, sanitize(k, v))
575 for k, v in template_dict.items()
577 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
579 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
580 tmpl = compat_expanduser(outtmpl)
581 filename = tmpl % template_dict
582 # Temporary fix for #4787
583 # 'Treat' all problem characters by passing filename through preferredencoding
584 # to workaround encoding issues with subprocess on python2 @ Windows
585 if sys.version_info < (3, 0) and sys.platform == 'win32':
586 filename = encodeFilename(filename, True).decode(preferredencoding())
587 return sanitize_path(filename)
588 except ValueError as err:
589 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
592 def _match_entry(self, info_dict, incomplete):
593 """ Returns None iff the file should be downloaded """
595 video_title = info_dict.get('title', info_dict.get('id', 'video'))
596 if 'title' in info_dict:
597 # This can happen when we're just evaluating the playlist
598 title = info_dict['title']
599 matchtitle = self.params.get('matchtitle', False)
601 if not re.search(matchtitle, title, re.IGNORECASE):
602 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
603 rejecttitle = self.params.get('rejecttitle', False)
605 if re.search(rejecttitle, title, re.IGNORECASE):
606 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
607 date = info_dict.get('upload_date', None)
609 dateRange = self.params.get('daterange', DateRange())
610 if date not in dateRange:
611 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
612 view_count = info_dict.get('view_count', None)
613 if view_count is not None:
614 min_views = self.params.get('min_views')
615 if min_views is not None and view_count < min_views:
616 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
617 max_views = self.params.get('max_views')
618 if max_views is not None and view_count > max_views:
619 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
620 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
621 return 'Skipping "%s" because it is age restricted' % video_title
622 if self.in_download_archive(info_dict):
623 return '%s has already been recorded in archive' % video_title
626 match_filter = self.params.get('match_filter')
627 if match_filter is not None:
628 ret = match_filter(info_dict)
635 def add_extra_info(info_dict, extra_info):
636 '''Set the keys from extra_info in info dict if they are missing'''
637 for key, value in extra_info.items():
638 info_dict.setdefault(key, value)
640 def extract_info(self, url, download=True, ie_key=None, extra_info={},
641 process=True, force_generic_extractor=False):
643 Returns a list with a dictionary for each video we find.
644 If 'download', also downloads the videos.
645 extra_info is a dict containing the extra values to add to each result
648 if not ie_key and force_generic_extractor:
652 ies = [self.get_info_extractor(ie_key)]
657 if not ie.suitable(url):
661 self.report_warning('The program functionality for this site has been marked as broken, '
662 'and will probably not work.')
665 ie_result = ie.extract(url)
666 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
668 if isinstance(ie_result, list):
669 # Backwards compatibility: old IE result format
671 '_type': 'compat_list',
672 'entries': ie_result,
674 self.add_default_extra_info(ie_result, ie, url)
676 return self.process_ie_result(ie_result, download, extra_info)
679 except ExtractorError as e: # An error we somewhat expected
680 self.report_error(compat_str(e), e.format_traceback())
682 except MaxDownloadsReached:
684 except Exception as e:
685 if self.params.get('ignoreerrors', False):
686 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
691 self.report_error('no suitable InfoExtractor for URL %s' % url)
693 def add_default_extra_info(self, ie_result, ie, url):
694 self.add_extra_info(ie_result, {
695 'extractor': ie.IE_NAME,
697 'webpage_url_basename': url_basename(url),
698 'extractor_key': ie.ie_key(),
701 def process_ie_result(self, ie_result, download=True, extra_info={}):
703 Take the result of the ie(may be modified) and resolve all unresolved
704 references (URLs, playlist items).
706 It will also download the videos if 'download'.
707 Returns the resolved ie_result.
710 result_type = ie_result.get('_type', 'video')
712 if result_type in ('url', 'url_transparent'):
713 extract_flat = self.params.get('extract_flat', False)
714 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
715 extract_flat is True):
716 if self.params.get('forcejson', False):
717 self.to_stdout(json.dumps(ie_result))
720 if result_type == 'video':
721 self.add_extra_info(ie_result, extra_info)
722 return self.process_video_result(ie_result, download=download)
723 elif result_type == 'url':
724 # We have to add extra_info to the results because it may be
725 # contained in a playlist
726 return self.extract_info(ie_result['url'],
728 ie_key=ie_result.get('ie_key'),
729 extra_info=extra_info)
730 elif result_type == 'url_transparent':
731 # Use the information from the embedding page
732 info = self.extract_info(
733 ie_result['url'], ie_key=ie_result.get('ie_key'),
734 extra_info=extra_info, download=False, process=False)
736 force_properties = dict(
737 (k, v) for k, v in ie_result.items() if v is not None)
738 for f in ('_type', 'url'):
739 if f in force_properties:
740 del force_properties[f]
741 new_result = info.copy()
742 new_result.update(force_properties)
744 assert new_result.get('_type') != 'url_transparent'
746 return self.process_ie_result(
747 new_result, download=download, extra_info=extra_info)
748 elif result_type == 'playlist' or result_type == 'multi_video':
749 # We process each entry in the playlist
750 playlist = ie_result.get('title', None) or ie_result.get('id', None)
751 self.to_screen('[download] Downloading playlist: %s' % playlist)
753 playlist_results = []
755 playliststart = self.params.get('playliststart', 1) - 1
756 playlistend = self.params.get('playlistend', None)
757 # For backwards compatibility, interpret -1 as whole list
758 if playlistend == -1:
761 playlistitems_str = self.params.get('playlist_items', None)
763 if playlistitems_str is not None:
764 def iter_playlistitems(format):
765 for string_segment in format.split(','):
766 if '-' in string_segment:
767 start, end = string_segment.split('-')
768 for item in range(int(start), int(end) + 1):
771 yield int(string_segment)
772 playlistitems = iter_playlistitems(playlistitems_str)
774 ie_entries = ie_result['entries']
775 if isinstance(ie_entries, list):
776 n_all_entries = len(ie_entries)
779 ie_entries[i - 1] for i in playlistitems
780 if -n_all_entries <= i - 1 < n_all_entries]
782 entries = ie_entries[playliststart:playlistend]
783 n_entries = len(entries)
785 "[%s] playlist %s: Collected %d video ids (downloading %d of them)" %
786 (ie_result['extractor'], playlist, n_all_entries, n_entries))
787 elif isinstance(ie_entries, PagedList):
790 for item in playlistitems:
791 entries.extend(ie_entries.getslice(
795 entries = ie_entries.getslice(
796 playliststart, playlistend)
797 n_entries = len(entries)
799 "[%s] playlist %s: Downloading %d videos" %
800 (ie_result['extractor'], playlist, n_entries))
803 entry_list = list(ie_entries)
804 entries = [entry_list[i - 1] for i in playlistitems]
806 entries = list(itertools.islice(
807 ie_entries, playliststart, playlistend))
808 n_entries = len(entries)
810 "[%s] playlist %s: Downloading %d videos" %
811 (ie_result['extractor'], playlist, n_entries))
813 if self.params.get('playlistreverse', False):
814 entries = entries[::-1]
816 for i, entry in enumerate(entries, 1):
817 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
819 'n_entries': n_entries,
820 'playlist': playlist,
821 'playlist_id': ie_result.get('id'),
822 'playlist_title': ie_result.get('title'),
823 'playlist_index': i + playliststart,
824 'extractor': ie_result['extractor'],
825 'webpage_url': ie_result['webpage_url'],
826 'webpage_url_basename': url_basename(ie_result['webpage_url']),
827 'extractor_key': ie_result['extractor_key'],
830 reason = self._match_entry(entry, incomplete=True)
831 if reason is not None:
832 self.to_screen('[download] ' + reason)
835 entry_result = self.process_ie_result(entry,
838 playlist_results.append(entry_result)
839 ie_result['entries'] = playlist_results
840 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
842 elif result_type == 'compat_list':
844 'Extractor %s returned a compat_list result. '
845 'It needs to be updated.' % ie_result.get('extractor'))
851 'extractor': ie_result['extractor'],
852 'webpage_url': ie_result['webpage_url'],
853 'webpage_url_basename': url_basename(ie_result['webpage_url']),
854 'extractor_key': ie_result['extractor_key'],
858 ie_result['entries'] = [
859 self.process_ie_result(_fixup(r), download, extra_info)
860 for r in ie_result['entries']
864 raise Exception('Invalid result type: %s' % result_type)
866 def _build_format_filter(self, filter_spec):
867 " Returns a function to filter the formats according to the filter_spec "
877 operator_rex = re.compile(r'''(?x)\s*
878 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
879 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
880 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
882 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
883 m = operator_rex.search(filter_spec)
886 comparison_value = int(m.group('value'))
888 comparison_value = parse_filesize(m.group('value'))
889 if comparison_value is None:
890 comparison_value = parse_filesize(m.group('value') + 'B')
891 if comparison_value is None:
893 'Invalid value %r in format specification %r' % (
894 m.group('value'), filter_spec))
895 op = OPERATORS[m.group('op')]
902 str_operator_rex = re.compile(r'''(?x)
903 \s*(?P<key>ext|acodec|vcodec|container|protocol)
904 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
905 \s*(?P<value>[a-zA-Z0-9_-]+)
907 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
908 m = str_operator_rex.search(filter_spec)
910 comparison_value = m.group('value')
911 op = STR_OPERATORS[m.group('op')]
914 raise ValueError('Invalid filter specification %r' % filter_spec)
917 actual_value = f.get(m.group('key'))
918 if actual_value is None:
919 return m.group('none_inclusive')
920 return op(actual_value, comparison_value)
923 def build_format_selector(self, format_spec):
924 def syntax_error(note, start):
926 'Invalid format specification: '
927 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
928 return SyntaxError(message)
930 PICKFIRST = 'PICKFIRST'
934 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
936 def _parse_filter(tokens):
938 for type, string, start, _, _ in tokens:
939 if type == tokenize.OP and string == ']':
940 return ''.join(filter_parts)
942 filter_parts.append(string)
944 def _remove_unused_ops(tokens):
945 # Remove operators that we don't use and join them with the surrounding strings
946 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
947 ALLOWED_OPS = ('/', '+', ',', '(', ')')
948 last_string, last_start, last_end, last_line = None, None, None, None
949 for type, string, start, end, line in tokens:
950 if type == tokenize.OP and string == '[':
952 yield tokenize.NAME, last_string, last_start, last_end, last_line
954 yield type, string, start, end, line
955 # everything inside brackets will be handled by _parse_filter
956 for type, string, start, end, line in tokens:
957 yield type, string, start, end, line
958 if type == tokenize.OP and string == ']':
960 elif type == tokenize.OP and string in ALLOWED_OPS:
962 yield tokenize.NAME, last_string, last_start, last_end, last_line
964 yield type, string, start, end, line
965 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
971 last_string += string
973 yield tokenize.NAME, last_string, last_start, last_end, last_line
975 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
977 current_selector = None
978 for type, string, start, _, _ in tokens:
979 # ENCODING is only defined in python 3.x
980 if type == getattr(tokenize, 'ENCODING', None):
982 elif type in [tokenize.NAME, tokenize.NUMBER]:
983 current_selector = FormatSelector(SINGLE, string, [])
984 elif type == tokenize.OP:
987 # ')' will be handled by the parentheses group
988 tokens.restore_last_token()
990 elif inside_merge and string in ['/', ',']:
991 tokens.restore_last_token()
993 elif inside_choice and string == ',':
994 tokens.restore_last_token()
997 if not current_selector:
998 raise syntax_error('"," must follow a format selector', start)
999 selectors.append(current_selector)
1000 current_selector = None
1002 if not current_selector:
1003 raise syntax_error('"/" must follow a format selector', start)
1004 first_choice = current_selector
1005 second_choice = _parse_format_selection(tokens, inside_choice=True)
1006 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1008 if not current_selector:
1009 current_selector = FormatSelector(SINGLE, 'best', [])
1010 format_filter = _parse_filter(tokens)
1011 current_selector.filters.append(format_filter)
1013 if current_selector:
1014 raise syntax_error('Unexpected "("', start)
1015 group = _parse_format_selection(tokens, inside_group=True)
1016 current_selector = FormatSelector(GROUP, group, [])
1018 video_selector = current_selector
1019 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1020 if not video_selector or not audio_selector:
1021 raise syntax_error('"+" must be between two format selectors', start)
1022 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1024 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1025 elif type == tokenize.ENDMARKER:
1027 if current_selector:
1028 selectors.append(current_selector)
1031 def _build_selector_function(selector):
1032 if isinstance(selector, list):
1033 fs = [_build_selector_function(s) for s in selector]
1035 def selector_function(formats):
1037 for format in f(formats):
1039 return selector_function
1040 elif selector.type == GROUP:
1041 selector_function = _build_selector_function(selector.selector)
1042 elif selector.type == PICKFIRST:
1043 fs = [_build_selector_function(s) for s in selector.selector]
1045 def selector_function(formats):
1047 picked_formats = list(f(formats))
1049 return picked_formats
1051 elif selector.type == SINGLE:
1052 format_spec = selector.selector
1054 def selector_function(formats):
1055 formats = list(formats)
1058 if format_spec == 'all':
1061 elif format_spec in ['best', 'worst', None]:
1062 format_idx = 0 if format_spec == 'worst' else -1
1063 audiovideo_formats = [
1065 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1066 if audiovideo_formats:
1067 yield audiovideo_formats[format_idx]
1068 # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format
1069 elif (all(f.get('acodec') != 'none' for f in formats) or
1070 all(f.get('vcodec') != 'none' for f in formats)):
1071 yield formats[format_idx]
1072 elif format_spec == 'bestaudio':
1075 if f.get('vcodec') == 'none']
1077 yield audio_formats[-1]
1078 elif format_spec == 'worstaudio':
1081 if f.get('vcodec') == 'none']
1083 yield audio_formats[0]
1084 elif format_spec == 'bestvideo':
1087 if f.get('acodec') == 'none']
1089 yield video_formats[-1]
1090 elif format_spec == 'worstvideo':
1093 if f.get('acodec') == 'none']
1095 yield video_formats[0]
1097 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1098 if format_spec in extensions:
1099 filter_f = lambda f: f['ext'] == format_spec
1101 filter_f = lambda f: f['format_id'] == format_spec
1102 matches = list(filter(filter_f, formats))
1105 elif selector.type == MERGE:
1106 def _merge(formats_info):
1107 format_1, format_2 = [f['format_id'] for f in formats_info]
1108 # The first format must contain the video and the
1110 if formats_info[0].get('vcodec') == 'none':
1111 self.report_error('The first format must '
1112 'contain the video, try using '
1113 '"-f %s+%s"' % (format_2, format_1))
1115 # Formats must be opposite (video+audio)
1116 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1118 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1119 % (format_1, format_2))
1122 formats_info[0]['ext']
1123 if self.params.get('merge_output_format') is None
1124 else self.params['merge_output_format'])
1126 'requested_formats': formats_info,
1127 'format': '%s+%s' % (formats_info[0].get('format'),
1128 formats_info[1].get('format')),
1129 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1130 formats_info[1].get('format_id')),
1131 'width': formats_info[0].get('width'),
1132 'height': formats_info[0].get('height'),
1133 'resolution': formats_info[0].get('resolution'),
1134 'fps': formats_info[0].get('fps'),
1135 'vcodec': formats_info[0].get('vcodec'),
1136 'vbr': formats_info[0].get('vbr'),
1137 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1138 'acodec': formats_info[1].get('acodec'),
1139 'abr': formats_info[1].get('abr'),
1142 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1144 def selector_function(formats):
1145 formats = list(formats)
1146 for pair in itertools.product(video_selector(formats), audio_selector(formats)):
1149 filters = [self._build_format_filter(f) for f in selector.filters]
1151 def final_selector(formats):
1152 for _filter in filters:
1153 formats = list(filter(_filter, formats))
1154 return selector_function(formats)
1155 return final_selector
1157 stream = io.BytesIO(format_spec.encode('utf-8'))
1159 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1160 except tokenize.TokenError:
1161 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1163 class TokenIterator(object):
1164 def __init__(self, tokens):
1165 self.tokens = tokens
1172 if self.counter >= len(self.tokens):
1173 raise StopIteration()
1174 value = self.tokens[self.counter]
1180 def restore_last_token(self):
1183 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1184 return _build_selector_function(parsed_selector)
1186 def _calc_headers(self, info_dict):
1187 res = std_headers.copy()
1189 add_headers = info_dict.get('http_headers')
1191 res.update(add_headers)
1193 cookies = self._calc_cookies(info_dict)
1195 res['Cookie'] = cookies
1199 def _calc_cookies(self, info_dict):
1200 pr = sanitized_Request(info_dict['url'])
1201 self.cookiejar.add_cookie_header(pr)
1202 return pr.get_header('Cookie')
1204 def process_video_result(self, info_dict, download=True):
1205 assert info_dict.get('_type', 'video') == 'video'
1207 if 'id' not in info_dict:
1208 raise ExtractorError('Missing "id" field in extractor result')
1209 if 'title' not in info_dict:
1210 raise ExtractorError('Missing "title" field in extractor result')
1212 if 'playlist' not in info_dict:
1213 # It isn't part of a playlist
1214 info_dict['playlist'] = None
1215 info_dict['playlist_index'] = None
1217 thumbnails = info_dict.get('thumbnails')
1218 if thumbnails is None:
1219 thumbnail = info_dict.get('thumbnail')
1221 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1223 thumbnails.sort(key=lambda t: (
1224 t.get('preference'), t.get('width'), t.get('height'),
1225 t.get('id'), t.get('url')))
1226 for i, t in enumerate(thumbnails):
1227 if t.get('width') and t.get('height'):
1228 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1229 if t.get('id') is None:
1232 if thumbnails and 'thumbnail' not in info_dict:
1233 info_dict['thumbnail'] = thumbnails[-1]['url']
1235 if 'display_id' not in info_dict and 'id' in info_dict:
1236 info_dict['display_id'] = info_dict['id']
1238 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1239 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1240 # see http://bugs.python.org/issue1646728)
1242 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1243 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1244 except (ValueError, OverflowError, OSError):
1247 # Auto generate title fields corresponding to the *_number fields when missing
1248 # in order to always have clean titles. This is very common for TV series.
1249 for field in ('chapter', 'season', 'episode'):
1250 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1251 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1253 subtitles = info_dict.get('subtitles')
1255 for _, subtitle in subtitles.items():
1256 for subtitle_format in subtitle:
1257 if 'ext' not in subtitle_format:
1258 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1260 if self.params.get('listsubtitles', False):
1261 if 'automatic_captions' in info_dict:
1262 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1263 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1265 info_dict['requested_subtitles'] = self.process_subtitles(
1266 info_dict['id'], subtitles,
1267 info_dict.get('automatic_captions'))
1269 # We now pick which formats have to be downloaded
1270 if info_dict.get('formats') is None:
1271 # There's only one format available
1272 formats = [info_dict]
1274 formats = info_dict['formats']
1277 raise ExtractorError('No video formats found!')
1281 # We check that all the formats have the format and format_id fields
1282 for i, format in enumerate(formats):
1283 if 'url' not in format:
1284 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1286 if format.get('format_id') is None:
1287 format['format_id'] = compat_str(i)
1288 format_id = format['format_id']
1289 if format_id not in formats_dict:
1290 formats_dict[format_id] = []
1291 formats_dict[format_id].append(format)
1293 # Make sure all formats have unique format_id
1294 for format_id, ambiguous_formats in formats_dict.items():
1295 if len(ambiguous_formats) > 1:
1296 for i, format in enumerate(ambiguous_formats):
1297 format['format_id'] = '%s-%d' % (format_id, i)
1299 for i, format in enumerate(formats):
1300 if format.get('format') is None:
1301 format['format'] = '{id} - {res}{note}'.format(
1302 id=format['format_id'],
1303 res=self.format_resolution(format),
1304 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1306 # Automatically determine file extension if missing
1307 if 'ext' not in format:
1308 format['ext'] = determine_ext(format['url']).lower()
1309 # Add HTTP headers, so that external programs can use them from the
1311 full_format_info = info_dict.copy()
1312 full_format_info.update(format)
1313 format['http_headers'] = self._calc_headers(full_format_info)
1315 # TODO Central sorting goes here
1317 if formats[0] is not info_dict:
1318 # only set the 'formats' fields if the original info_dict list them
1319 # otherwise we end up with a circular reference, the first (and unique)
1320 # element in the 'formats' field in info_dict is info_dict itself,
1321 # which can't be exported to json
1322 info_dict['formats'] = formats
1323 if self.params.get('listformats'):
1324 self.list_formats(info_dict)
1326 if self.params.get('list_thumbnails'):
1327 self.list_thumbnails(info_dict)
1330 req_format = self.params.get('format')
1331 if req_format is None:
1332 req_format_list = []
1333 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1334 info_dict['extractor'] in ['youtube', 'ted'] and
1335 not info_dict.get('is_live')):
1336 merger = FFmpegMergerPP(self)
1337 if merger.available and merger.can_merge():
1338 req_format_list.append('bestvideo+bestaudio')
1339 req_format_list.append('best')
1340 req_format = '/'.join(req_format_list)
1341 format_selector = self.build_format_selector(req_format)
1342 formats_to_download = list(format_selector(formats))
1343 if not formats_to_download:
1344 raise ExtractorError('requested format not available',
1348 if len(formats_to_download) > 1:
1349 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1350 for format in formats_to_download:
1351 new_info = dict(info_dict)
1352 new_info.update(format)
1353 self.process_info(new_info)
1354 # We update the info dict with the best quality format (backwards compatibility)
1355 info_dict.update(formats_to_download[-1])
1358 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1359 """Select the requested subtitles and their format"""
1361 if normal_subtitles and self.params.get('writesubtitles'):
1362 available_subs.update(normal_subtitles)
1363 if automatic_captions and self.params.get('writeautomaticsub'):
1364 for lang, cap_info in automatic_captions.items():
1365 if lang not in available_subs:
1366 available_subs[lang] = cap_info
1368 if (not self.params.get('writesubtitles') and not
1369 self.params.get('writeautomaticsub') or not
1373 if self.params.get('allsubtitles', False):
1374 requested_langs = available_subs.keys()
1376 if self.params.get('subtitleslangs', False):
1377 requested_langs = self.params.get('subtitleslangs')
1378 elif 'en' in available_subs:
1379 requested_langs = ['en']
1381 requested_langs = [list(available_subs.keys())[0]]
1383 formats_query = self.params.get('subtitlesformat', 'best')
1384 formats_preference = formats_query.split('/') if formats_query else []
1386 for lang in requested_langs:
1387 formats = available_subs.get(lang)
1389 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1391 for ext in formats_preference:
1395 matches = list(filter(lambda f: f['ext'] == ext, formats))
1401 self.report_warning(
1402 'No subtitle format found matching "%s" for language %s, '
1403 'using %s' % (formats_query, lang, f['ext']))
1407 def process_info(self, info_dict):
1408 """Process a single resolved IE result."""
1410 assert info_dict.get('_type', 'video') == 'video'
1412 max_downloads = self.params.get('max_downloads')
1413 if max_downloads is not None:
1414 if self._num_downloads >= int(max_downloads):
1415 raise MaxDownloadsReached()
1417 info_dict['fulltitle'] = info_dict['title']
1418 if len(info_dict['title']) > 200:
1419 info_dict['title'] = info_dict['title'][:197] + '...'
1421 if 'format' not in info_dict:
1422 info_dict['format'] = info_dict['ext']
1424 reason = self._match_entry(info_dict, incomplete=False)
1425 if reason is not None:
1426 self.to_screen('[download] ' + reason)
1429 self._num_downloads += 1
1431 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1434 if self.params.get('forcetitle', False):
1435 self.to_stdout(info_dict['fulltitle'])
1436 if self.params.get('forceid', False):
1437 self.to_stdout(info_dict['id'])
1438 if self.params.get('forceurl', False):
1439 if info_dict.get('requested_formats') is not None:
1440 for f in info_dict['requested_formats']:
1441 self.to_stdout(f['url'] + f.get('play_path', ''))
1443 # For RTMP URLs, also include the playpath
1444 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1445 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1446 self.to_stdout(info_dict['thumbnail'])
1447 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1448 self.to_stdout(info_dict['description'])
1449 if self.params.get('forcefilename', False) and filename is not None:
1450 self.to_stdout(filename)
1451 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1452 self.to_stdout(formatSeconds(info_dict['duration']))
1453 if self.params.get('forceformat', False):
1454 self.to_stdout(info_dict['format'])
1455 if self.params.get('forcejson', False):
1456 self.to_stdout(json.dumps(info_dict))
1458 # Do nothing else if in simulate mode
1459 if self.params.get('simulate', False):
1462 if filename is None:
1466 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1467 if dn and not os.path.exists(dn):
1469 except (OSError, IOError) as err:
1470 self.report_error('unable to create directory ' + error_to_compat_str(err))
1473 if self.params.get('writedescription', False):
1474 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1475 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1476 self.to_screen('[info] Video description is already present')
1477 elif info_dict.get('description') is None:
1478 self.report_warning('There\'s no description to write.')
1481 self.to_screen('[info] Writing video description to: ' + descfn)
1482 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1483 descfile.write(info_dict['description'])
1484 except (OSError, IOError):
1485 self.report_error('Cannot write description file ' + descfn)
1488 if self.params.get('writeannotations', False):
1489 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1490 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1491 self.to_screen('[info] Video annotations are already present')
1494 self.to_screen('[info] Writing video annotations to: ' + annofn)
1495 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1496 annofile.write(info_dict['annotations'])
1497 except (KeyError, TypeError):
1498 self.report_warning('There are no annotations to write.')
1499 except (OSError, IOError):
1500 self.report_error('Cannot write annotations file: ' + annofn)
1503 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1504 self.params.get('writeautomaticsub')])
1506 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1507 # subtitles download errors are already managed as troubles in relevant IE
1508 # that way it will silently go on when used with unsupporting IE
1509 subtitles = info_dict['requested_subtitles']
1510 ie = self.get_info_extractor(info_dict['extractor_key'])
1511 for sub_lang, sub_info in subtitles.items():
1512 sub_format = sub_info['ext']
1513 if sub_info.get('data') is not None:
1514 sub_data = sub_info['data']
1517 sub_data = ie._download_webpage(
1518 sub_info['url'], info_dict['id'], note=False)
1519 except ExtractorError as err:
1520 self.report_warning('Unable to download subtitle for "%s": %s' %
1521 (sub_lang, error_to_compat_str(err.cause)))
1524 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1525 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1526 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1528 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1529 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1530 subfile.write(sub_data)
1531 except (OSError, IOError):
1532 self.report_error('Cannot write subtitles file ' + sub_filename)
1535 if self.params.get('writeinfojson', False):
1536 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1537 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1538 self.to_screen('[info] Video description metadata is already present')
1540 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1542 write_json_file(self.filter_requested_info(info_dict), infofn)
1543 except (OSError, IOError):
1544 self.report_error('Cannot write metadata to JSON file ' + infofn)
1547 self._write_thumbnails(info_dict, filename)
1549 if not self.params.get('skip_download', False):
1552 fd = get_suitable_downloader(info, self.params)(self, self.params)
1553 for ph in self._progress_hooks:
1554 fd.add_progress_hook(ph)
1555 if self.params.get('verbose'):
1556 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1557 return fd.download(name, info)
1559 if info_dict.get('requested_formats') is not None:
1562 merger = FFmpegMergerPP(self)
1563 if not merger.available:
1565 self.report_warning('You have requested multiple '
1566 'formats but ffmpeg or avconv are not installed.'
1567 ' The formats won\'t be merged.')
1569 postprocessors = [merger]
1571 def compatible_formats(formats):
1572 video, audio = formats
1574 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1575 if video_ext and audio_ext:
1577 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1580 for exts in COMPATIBLE_EXTS:
1581 if video_ext in exts and audio_ext in exts:
1583 # TODO: Check acodec/vcodec
1586 filename_real_ext = os.path.splitext(filename)[1][1:]
1588 os.path.splitext(filename)[0]
1589 if filename_real_ext == info_dict['ext']
1591 requested_formats = info_dict['requested_formats']
1592 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1593 info_dict['ext'] = 'mkv'
1594 self.report_warning(
1595 'Requested formats are incompatible for merge and will be merged into mkv.')
1596 # Ensure filename always has a correct extension for successful merge
1597 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1598 if os.path.exists(encodeFilename(filename)):
1600 '[download] %s has already been downloaded and '
1601 'merged' % filename)
1603 for f in requested_formats:
1604 new_info = dict(info_dict)
1606 fname = self.prepare_filename(new_info)
1607 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1608 downloaded.append(fname)
1609 partial_success = dl(fname, new_info)
1610 success = success and partial_success
1611 info_dict['__postprocessors'] = postprocessors
1612 info_dict['__files_to_merge'] = downloaded
1614 # Just a single file
1615 success = dl(filename, info_dict)
1616 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1617 self.report_error('unable to download video data: %s' % str(err))
1619 except (OSError, IOError) as err:
1620 raise UnavailableVideoError(err)
1621 except (ContentTooShortError, ) as err:
1622 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1627 fixup_policy = self.params.get('fixup')
1628 if fixup_policy is None:
1629 fixup_policy = 'detect_or_warn'
1631 stretched_ratio = info_dict.get('stretched_ratio')
1632 if stretched_ratio is not None and stretched_ratio != 1:
1633 if fixup_policy == 'warn':
1634 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1635 info_dict['id'], stretched_ratio))
1636 elif fixup_policy == 'detect_or_warn':
1637 stretched_pp = FFmpegFixupStretchedPP(self)
1638 if stretched_pp.available:
1639 info_dict.setdefault('__postprocessors', [])
1640 info_dict['__postprocessors'].append(stretched_pp)
1642 self.report_warning(
1643 '%s: Non-uniform pixel ratio (%s). Install ffmpeg or avconv to fix this automatically.' % (
1644 info_dict['id'], stretched_ratio))
1646 assert fixup_policy in ('ignore', 'never')
1648 if info_dict.get('requested_formats') is None and info_dict.get('container') == 'm4a_dash':
1649 if fixup_policy == 'warn':
1650 self.report_warning('%s: writing DASH m4a. Only some players support this container.' % (
1652 elif fixup_policy == 'detect_or_warn':
1653 fixup_pp = FFmpegFixupM4aPP(self)
1654 if fixup_pp.available:
1655 info_dict.setdefault('__postprocessors', [])
1656 info_dict['__postprocessors'].append(fixup_pp)
1658 self.report_warning(
1659 '%s: writing DASH m4a. Only some players support this container. Install ffmpeg or avconv to fix this automatically.' % (
1662 assert fixup_policy in ('ignore', 'never')
1665 self.post_process(filename, info_dict)
1666 except (PostProcessingError) as err:
1667 self.report_error('postprocessing: %s' % str(err))
1669 self.record_download_archive(info_dict)
1671 def download(self, url_list):
1672 """Download a given list of URLs."""
1673 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1674 if (len(url_list) > 1 and
1675 '%' not in outtmpl and
1676 self.params.get('max_downloads') != 1):
1677 raise SameFileError(outtmpl)
1679 for url in url_list:
1681 # It also downloads the videos
1682 res = self.extract_info(
1683 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1684 except UnavailableVideoError:
1685 self.report_error('unable to download video')
1686 except MaxDownloadsReached:
1687 self.to_screen('[info] Maximum number of downloaded files reached.')
1690 if self.params.get('dump_single_json', False):
1691 self.to_stdout(json.dumps(res))
1693 return self._download_retcode
1695 def download_with_info_file(self, info_filename):
1696 with contextlib.closing(fileinput.FileInput(
1697 [info_filename], mode='r',
1698 openhook=fileinput.hook_encoded('utf-8'))) as f:
1699 # FileInput doesn't have a read method, we can't call json.load
1700 info = self.filter_requested_info(json.loads('\n'.join(f)))
1702 self.process_ie_result(info, download=True)
1703 except DownloadError:
1704 webpage_url = info.get('webpage_url')
1705 if webpage_url is not None:
1706 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1707 return self.download([webpage_url])
1710 return self._download_retcode
1713 def filter_requested_info(info_dict):
1715 (k, v) for k, v in info_dict.items()
1716 if k not in ['requested_formats', 'requested_subtitles'])
1718 def post_process(self, filename, ie_info):
1719 """Run all the postprocessors on the given file."""
1720 info = dict(ie_info)
1721 info['filepath'] = filename
1723 if ie_info.get('__postprocessors') is not None:
1724 pps_chain.extend(ie_info['__postprocessors'])
1725 pps_chain.extend(self._pps)
1726 for pp in pps_chain:
1727 files_to_delete = []
1729 files_to_delete, info = pp.run(info)
1730 except PostProcessingError as e:
1731 self.report_error(e.msg)
1732 if files_to_delete and not self.params.get('keepvideo', False):
1733 for old_filename in files_to_delete:
1734 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1736 os.remove(encodeFilename(old_filename))
1737 except (IOError, OSError):
1738 self.report_warning('Unable to remove downloaded original file')
1740 def _make_archive_id(self, info_dict):
1741 # Future-proof against any change in case
1742 # and backwards compatibility with prior versions
1743 extractor = info_dict.get('extractor_key')
1744 if extractor is None:
1745 if 'id' in info_dict:
1746 extractor = info_dict.get('ie_key') # key in a playlist
1747 if extractor is None:
1748 return None # Incomplete video information
1749 return extractor.lower() + ' ' + info_dict['id']
1751 def in_download_archive(self, info_dict):
1752 fn = self.params.get('download_archive')
1756 vid_id = self._make_archive_id(info_dict)
1758 return False # Incomplete video information
1761 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1762 for line in archive_file:
1763 if line.strip() == vid_id:
1765 except IOError as ioe:
1766 if ioe.errno != errno.ENOENT:
1770 def record_download_archive(self, info_dict):
1771 fn = self.params.get('download_archive')
1774 vid_id = self._make_archive_id(info_dict)
1776 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1777 archive_file.write(vid_id + '\n')
1780 def format_resolution(format, default='unknown'):
1781 if format.get('vcodec') == 'none':
1783 if format.get('resolution') is not None:
1784 return format['resolution']
1785 if format.get('height') is not None:
1786 if format.get('width') is not None:
1787 res = '%sx%s' % (format['width'], format['height'])
1789 res = '%sp' % format['height']
1790 elif format.get('width') is not None:
1791 res = '?x%d' % format['width']
1796 def _format_note(self, fdict):
1798 if fdict.get('ext') in ['f4f', 'f4m']:
1799 res += '(unsupported) '
1800 if fdict.get('language'):
1803 res += '[%s]' % fdict['language']
1804 if fdict.get('format_note') is not None:
1805 res += fdict['format_note'] + ' '
1806 if fdict.get('tbr') is not None:
1807 res += '%4dk ' % fdict['tbr']
1808 if fdict.get('container') is not None:
1811 res += '%s container' % fdict['container']
1812 if (fdict.get('vcodec') is not None and
1813 fdict.get('vcodec') != 'none'):
1816 res += fdict['vcodec']
1817 if fdict.get('vbr') is not None:
1819 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1821 if fdict.get('vbr') is not None:
1822 res += '%4dk' % fdict['vbr']
1823 if fdict.get('fps') is not None:
1824 res += ', %sfps' % fdict['fps']
1825 if fdict.get('acodec') is not None:
1828 if fdict['acodec'] == 'none':
1831 res += '%-5s' % fdict['acodec']
1832 elif fdict.get('abr') is not None:
1836 if fdict.get('abr') is not None:
1837 res += '@%3dk' % fdict['abr']
1838 if fdict.get('asr') is not None:
1839 res += ' (%5dHz)' % fdict['asr']
1840 if fdict.get('filesize') is not None:
1843 res += format_bytes(fdict['filesize'])
1844 elif fdict.get('filesize_approx') is not None:
1847 res += '~' + format_bytes(fdict['filesize_approx'])
1850 def list_formats(self, info_dict):
1851 formats = info_dict.get('formats', [info_dict])
1853 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1855 if f.get('preference') is None or f['preference'] >= -1000]
1856 if len(formats) > 1:
1857 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1859 header_line = ['format code', 'extension', 'resolution', 'note']
1861 '[info] Available formats for %s:\n%s' %
1862 (info_dict['id'], render_table(header_line, table)))
1864 def list_thumbnails(self, info_dict):
1865 thumbnails = info_dict.get('thumbnails')
1867 tn_url = info_dict.get('thumbnail')
1869 thumbnails = [{'id': '0', 'url': tn_url}]
1872 '[info] No thumbnails present for %s' % info_dict['id'])
1876 '[info] Thumbnails for %s:' % info_dict['id'])
1877 self.to_screen(render_table(
1878 ['ID', 'width', 'height', 'URL'],
1879 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1881 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1883 self.to_screen('%s has no %s' % (video_id, name))
1886 'Available %s for %s:' % (name, video_id))
1887 self.to_screen(render_table(
1888 ['Language', 'formats'],
1889 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1890 for lang, formats in subtitles.items()]))
1892 def urlopen(self, req):
1893 """ Start an HTTP download """
1894 if isinstance(req, compat_basestring):
1895 req = sanitized_Request(req)
1896 return self._opener.open(req, timeout=self._socket_timeout)
1898 def print_debug_header(self):
1899 if not self.params.get('verbose'):
1902 if type('') is not compat_str:
1903 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1904 self.report_warning(
1905 'Your Python is broken! Update to a newer and supported version')
1907 stdout_encoding = getattr(
1908 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1910 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1911 locale.getpreferredencoding(),
1912 sys.getfilesystemencoding(),
1914 self.get_encoding()))
1915 write_string(encoding_str, encoding=None)
1917 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1919 sp = subprocess.Popen(
1920 ['git', 'rev-parse', '--short', 'HEAD'],
1921 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1922 cwd=os.path.dirname(os.path.abspath(__file__)))
1923 out, err = sp.communicate()
1924 out = out.decode().strip()
1925 if re.match('[0-9a-f]+', out):
1926 self._write_string('[debug] Git HEAD: ' + out + '\n')
1932 self._write_string('[debug] Python version %s - %s\n' % (
1933 platform.python_version(), platform_name()))
1935 exe_versions = FFmpegPostProcessor.get_versions(self)
1936 exe_versions['rtmpdump'] = rtmpdump_version()
1937 exe_str = ', '.join(
1939 for exe, v in sorted(exe_versions.items())
1944 self._write_string('[debug] exe versions: %s\n' % exe_str)
1947 for handler in self._opener.handlers:
1948 if hasattr(handler, 'proxies'):
1949 proxy_map.update(handler.proxies)
1950 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1952 if self.params.get('call_home', False):
1953 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
1954 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
1955 latest_version = self.urlopen(
1956 'https://yt-dl.org/latest/version').read().decode('utf-8')
1957 if version_tuple(latest_version) > version_tuple(__version__):
1958 self.report_warning(
1959 'You are using an outdated version (newest version: %s)! '
1960 'See https://yt-dl.org/update if you need help updating.' %
1963 def _setup_opener(self):
1964 timeout_val = self.params.get('socket_timeout')
1965 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
1967 opts_cookiefile = self.params.get('cookiefile')
1968 opts_proxy = self.params.get('proxy')
1970 if opts_cookiefile is None:
1971 self.cookiejar = compat_cookiejar.CookieJar()
1973 self.cookiejar = compat_cookiejar.MozillaCookieJar(
1975 if os.access(opts_cookiefile, os.R_OK):
1976 self.cookiejar.load()
1978 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
1979 if opts_proxy is not None:
1980 if opts_proxy == '':
1983 proxies = {'http': opts_proxy, 'https': opts_proxy}
1985 proxies = compat_urllib_request.getproxies()
1986 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
1987 if 'http' in proxies and 'https' not in proxies:
1988 proxies['https'] = proxies['http']
1989 proxy_handler = PerRequestProxyHandler(proxies)
1991 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
1992 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
1993 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
1994 data_handler = compat_urllib_request_DataHandler()
1996 # When passing our own FileHandler instance, build_opener won't add the
1997 # default FileHandler and allows us to disable the file protocol, which
1998 # can be used for malicious purposes (see
1999 # https://github.com/rg3/youtube-dl/issues/8227)
2000 file_handler = compat_urllib_request.FileHandler()
2002 def file_open(*args, **kwargs):
2003 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2004 file_handler.file_open = file_open
2006 opener = compat_urllib_request.build_opener(
2007 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2009 # Delete the default user-agent header, which would otherwise apply in
2010 # cases where our custom HTTP handler doesn't come into play
2011 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2012 opener.addheaders = []
2013 self._opener = opener
2015 def encode(self, s):
2016 if isinstance(s, bytes):
2017 return s # Already encoded
2020 return s.encode(self.get_encoding())
2021 except UnicodeEncodeError as err:
2022 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2025 def get_encoding(self):
2026 encoding = self.params.get('encoding')
2027 if encoding is None:
2028 encoding = preferredencoding()
2031 def _write_thumbnails(self, info_dict, filename):
2032 if self.params.get('writethumbnail', False):
2033 thumbnails = info_dict.get('thumbnails')
2035 thumbnails = [thumbnails[-1]]
2036 elif self.params.get('write_all_thumbnails', False):
2037 thumbnails = info_dict.get('thumbnails')
2042 # No thumbnails present, so return immediately
2045 for t in thumbnails:
2046 thumb_ext = determine_ext(t['url'], 'jpg')
2047 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2048 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2049 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2051 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2052 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2053 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2055 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2056 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2058 uf = self.urlopen(t['url'])
2059 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2060 shutil.copyfileobj(uf, thumbf)
2061 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2062 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2063 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2064 self.report_warning('Unable to download thumbnail "%s": %s' %
2065 (t['url'], error_to_compat_str(err)))