2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import, unicode_literals
32 compat_get_terminal_size,
37 compat_tokenize_tokenize,
39 compat_urllib_request,
40 compat_urllib_request_DataHandler,
63 PerRequestProxyHandler,
68 register_socks_protocols,
78 UnavailableVideoError,
83 YoutubeDLCookieProcessor,
86 from .cache import Cache
87 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
88 from .downloader import get_suitable_downloader
89 from .downloader.rtmp import rtmpdump_version
90 from .postprocessor import (
93 FFmpegFixupStretchedPP,
98 from .version import __version__
100 if compat_os_name == 'nt':
104 class YoutubeDL(object):
107 YoutubeDL objects are the ones responsible of downloading the
108 actual video file and writing it to disk if the user has requested
109 it, among some other tasks. In most cases there should be one per
110 program. As, given a video URL, the downloader doesn't know how to
111 extract all the needed information, task that InfoExtractors do, it
112 has to pass the URL to one of them.
114 For this, YoutubeDL objects have a method that allows
115 InfoExtractors to be registered in a given order. When it is passed
116 a URL, the YoutubeDL object handles it to the first InfoExtractor it
117 finds that reports being able to handle it. The InfoExtractor extracts
118 all the information about the video or videos the URL refers to, and
119 YoutubeDL process the extracted information, possibly using a File
120 Downloader to download the video.
122 YoutubeDL objects accept a lot of parameters. In order not to saturate
123 the object constructor with arguments, it receives a dictionary of
124 options instead. These options are available through the params
125 attribute for the InfoExtractors to use. The YoutubeDL also
126 registers itself as the downloader in charge for the InfoExtractors
127 that are added to it, so this is a "mutual registration".
131 username: Username for authentication purposes.
132 password: Password for authentication purposes.
133 videopassword: Password for accessing a video.
134 usenetrc: Use netrc for authentication instead.
135 verbose: Print additional info to stdout.
136 quiet: Do not print messages to stdout.
137 no_warnings: Do not print out anything for warnings.
138 forceurl: Force printing final URL.
139 forcetitle: Force printing title.
140 forceid: Force printing ID.
141 forcethumbnail: Force printing thumbnail URL.
142 forcedescription: Force printing description.
143 forcefilename: Force printing final filename.
144 forceduration: Force printing duration.
145 forcejson: Force printing info_dict as JSON.
146 dump_single_json: Force printing the info_dict of the whole playlist
147 (or video) as a single JSON line.
148 simulate: Do not download the video files.
149 format: Video format code. See options.py for more information.
150 outtmpl: Template for output names.
151 restrictfilenames: Do not allow "&" and spaces in file names
152 ignoreerrors: Do not stop on download errors.
153 force_generic_extractor: Force downloader to use the generic extractor
154 nooverwrites: Prevent overwriting files.
155 playliststart: Playlist item to start at.
156 playlistend: Playlist item to end at.
157 playlist_items: Specific indices of playlist to download.
158 playlistreverse: Download playlist items in reverse order.
159 matchtitle: Download only matching titles.
160 rejecttitle: Reject downloads for matching titles.
161 logger: Log messages to a logging.Logger instance.
162 logtostderr: Log messages to stderr instead of stdout.
163 writedescription: Write the video description to a .description file
164 writeinfojson: Write the video description to a .info.json file
165 writeannotations: Write the video annotations to a .annotations.xml file
166 writethumbnail: Write the thumbnail image to a file
167 write_all_thumbnails: Write all thumbnail formats to files
168 writesubtitles: Write the video subtitles to a file
169 writeautomaticsub: Write the automatically generated subtitles to a file
170 allsubtitles: Downloads all the subtitles of the video
171 (requires writesubtitles or writeautomaticsub)
172 listsubtitles: Lists all available subtitles for the video
173 subtitlesformat: The format code for subtitles
174 subtitleslangs: List of languages of the subtitles to download
175 keepvideo: Keep the video file after post-processing
176 daterange: A DateRange object, download only if the upload_date is in the range.
177 skip_download: Skip the actual download of the video file
178 cachedir: Location of the cache files in the filesystem.
179 False to disable filesystem cache.
180 noplaylist: Download single video instead of a playlist if in doubt.
181 age_limit: An integer representing the user's age in years.
182 Unsuitable videos for the given age are skipped.
183 min_views: An integer representing the minimum view count the video
184 must have in order to not be skipped.
185 Videos without view count information are always
186 downloaded. None for no limit.
187 max_views: An integer representing the maximum view count.
188 Videos that are more popular than that are not
190 Videos without view count information are always
191 downloaded. None for no limit.
192 download_archive: File name of a file where all downloads are recorded.
193 Videos already present in the file are not downloaded
195 cookiefile: File name where cookies should be read from and dumped to.
196 nocheckcertificate:Do not verify SSL certificates
197 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
198 At the moment, this is only supported by YouTube.
199 proxy: URL of the proxy server to use
200 geo_verification_proxy: URL of the proxy to use for IP address verification
201 on geo-restricted sites. (Experimental)
202 socket_timeout: Time to wait for unresponsive hosts, in seconds
203 bidi_workaround: Work around buggy terminals without bidirectional text
204 support, using fridibi
205 debug_printtraffic:Print out sent and received HTTP traffic
206 include_ads: Download ads as well
207 default_search: Prepend this string if an input url is not valid.
208 'auto' for elaborate guessing
209 encoding: Use this encoding instead of the system-specified.
210 extract_flat: Do not resolve URLs, return the immediate result.
211 Pass in 'in_playlist' to only show this behavior for
213 postprocessors: A list of dictionaries, each with an entry
214 * key: The name of the postprocessor. See
215 youtube_dl/postprocessor/__init__.py for a list.
216 as well as any further keyword arguments for the
218 progress_hooks: A list of functions that get called on download
219 progress, with a dictionary with the entries
220 * status: One of "downloading", "error", or "finished".
221 Check this first and ignore unknown values.
223 If status is one of "downloading", or "finished", the
224 following properties may also be present:
225 * filename: The final filename (always present)
226 * tmpfilename: The filename we're currently writing to
227 * downloaded_bytes: Bytes on disk
228 * total_bytes: Size of the whole file, None if unknown
229 * total_bytes_estimate: Guess of the eventual file size,
231 * elapsed: The number of seconds since download started.
232 * eta: The estimated time in seconds, None if unknown
233 * speed: The download speed in bytes/second, None if
235 * fragment_index: The counter of the currently
236 downloaded video fragment.
237 * fragment_count: The number of fragments (= individual
238 files that will be merged)
240 Progress hooks are guaranteed to be called at least once
241 (with status "finished") if the download is successful.
242 merge_output_format: Extension to use when merging formats.
243 fixup: Automatically correct known faults of the file.
245 - "never": do nothing
246 - "warn": only emit a warning
247 - "detect_or_warn": check whether we can do anything
248 about it, warn otherwise (default)
249 source_address: (Experimental) Client-side IP address to bind to.
250 call_home: Boolean, true iff we are allowed to contact the
251 youtube-dl servers for debugging.
252 sleep_interval: Number of seconds to sleep before each download.
253 listformats: Print an overview of available video formats and exit.
254 list_thumbnails: Print a table of all thumbnails and exit.
255 match_filter: A function that gets called with the info_dict of
257 If it returns a message, the video is ignored.
258 If it returns None, the video is downloaded.
259 match_filter_func in utils.py is one example for this.
260 no_color: Do not emit color codes in output.
262 The following options determine which downloader is picked:
263 external_downloader: Executable of the external downloader to call.
264 None or unset for standard (built-in) downloader.
265 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
266 if True, otherwise use ffmpeg/avconv if False, otherwise
267 use downloader suggested by extractor if None.
269 The following parameters are not used by YoutubeDL itself, they are used by
270 the downloader (see youtube_dl/downloader/common.py):
271 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
272 noresizebuffer, retries, continuedl, noprogress, consoletitle,
273 xattr_set_filesize, external_downloader_args, hls_use_mpegts.
275 The following options are used by the post processors:
276 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
277 otherwise prefer avconv.
278 postprocessor_args: A list of additional command-line arguments for the
285 _download_retcode = None
286 _num_downloads = None
289 def __init__(self, params=None, auto_init=True):
290 """Create a FileDownloader object with the given options."""
294 self._ies_instances = {}
296 self._progress_hooks = []
297 self._download_retcode = 0
298 self._num_downloads = 0
299 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
300 self._err_file = sys.stderr
303 'nocheckcertificate': False,
305 self.params.update(params)
306 self.cache = Cache(self)
308 if self.params.get('cn_verification_proxy') is not None:
309 self.report_warning('--cn-verification-proxy is deprecated. Use --geo-verification-proxy instead.')
310 if self.params.get('geo_verification_proxy') is None:
311 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
313 if params.get('bidi_workaround', False):
316 master, slave = pty.openpty()
317 width = compat_get_terminal_size().columns
321 width_args = ['-w', str(width)]
323 stdin=subprocess.PIPE,
325 stderr=self._err_file)
327 self._output_process = subprocess.Popen(
328 ['bidiv'] + width_args, **sp_kwargs
331 self._output_process = subprocess.Popen(
332 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
333 self._output_channel = os.fdopen(master, 'rb')
334 except OSError as ose:
335 if ose.errno == errno.ENOENT:
336 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 $PATH.')
340 if (sys.version_info >= (3,) and sys.platform != 'win32' and
341 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
342 not params.get('restrictfilenames', False)):
343 # On Python 3, the Unicode filesystem API will throw errors (#1474)
345 'Assuming --restrict-filenames since file system encoding '
346 'cannot encode all characters. '
347 'Set the LC_ALL environment variable to fix this.')
348 self.params['restrictfilenames'] = True
350 if isinstance(params.get('outtmpl'), bytes):
352 'Parameter outtmpl is bytes, but should be a unicode string. '
353 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
358 self.print_debug_header()
359 self.add_default_info_extractors()
361 for pp_def_raw in self.params.get('postprocessors', []):
362 pp_class = get_postprocessor(pp_def_raw['key'])
363 pp_def = dict(pp_def_raw)
365 pp = pp_class(self, **compat_kwargs(pp_def))
366 self.add_post_processor(pp)
368 for ph in self.params.get('progress_hooks', []):
369 self.add_progress_hook(ph)
371 register_socks_protocols()
373 def warn_if_short_id(self, argv):
374 # short YouTube ID starting with dash?
376 i for i, a in enumerate(argv)
377 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
381 [a for i, a in enumerate(argv) if i not in idxs] +
382 ['--'] + [argv[i] for i in idxs]
385 'Long argument string detected. '
386 'Use -- to separate parameters and URLs, like this:\n%s\n' %
387 args_to_str(correct_argv))
389 def add_info_extractor(self, ie):
390 """Add an InfoExtractor object to the end of the list."""
392 if not isinstance(ie, type):
393 self._ies_instances[ie.ie_key()] = ie
394 ie.set_downloader(self)
396 def get_info_extractor(self, ie_key):
398 Get an instance of an IE with name ie_key, it will try to get one from
399 the _ies list, if there's no instance it will create a new one and add
400 it to the extractor list.
402 ie = self._ies_instances.get(ie_key)
404 ie = get_info_extractor(ie_key)()
405 self.add_info_extractor(ie)
408 def add_default_info_extractors(self):
410 Add the InfoExtractors returned by gen_extractors to the end of the list
412 for ie in gen_extractor_classes():
413 self.add_info_extractor(ie)
415 def add_post_processor(self, pp):
416 """Add a PostProcessor object to the end of the chain."""
418 pp.set_downloader(self)
420 def add_progress_hook(self, ph):
421 """Add the progress hook (currently only for the file downloader)"""
422 self._progress_hooks.append(ph)
424 def _bidi_workaround(self, message):
425 if not hasattr(self, '_output_channel'):
428 assert hasattr(self, '_output_process')
429 assert isinstance(message, compat_str)
430 line_count = message.count('\n') + 1
431 self._output_process.stdin.write((message + '\n').encode('utf-8'))
432 self._output_process.stdin.flush()
433 res = ''.join(self._output_channel.readline().decode('utf-8')
434 for _ in range(line_count))
435 return res[:-len('\n')]
437 def to_screen(self, message, skip_eol=False):
438 """Print message to stdout if not in quiet mode."""
439 return self.to_stdout(message, skip_eol, check_quiet=True)
441 def _write_string(self, s, out=None):
442 write_string(s, out=out, encoding=self.params.get('encoding'))
444 def to_stdout(self, message, skip_eol=False, check_quiet=False):
445 """Print message to stdout if not in quiet mode."""
446 if self.params.get('logger'):
447 self.params['logger'].debug(message)
448 elif not check_quiet or not self.params.get('quiet', False):
449 message = self._bidi_workaround(message)
450 terminator = ['\n', ''][skip_eol]
451 output = message + terminator
453 self._write_string(output, self._screen_file)
455 def to_stderr(self, message):
456 """Print message to stderr."""
457 assert isinstance(message, compat_str)
458 if self.params.get('logger'):
459 self.params['logger'].error(message)
461 message = self._bidi_workaround(message)
462 output = message + '\n'
463 self._write_string(output, self._err_file)
465 def to_console_title(self, message):
466 if not self.params.get('consoletitle', False):
468 if compat_os_name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
469 # c_wchar_p() might not be necessary if `message` is
470 # already of type unicode()
471 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
472 elif 'TERM' in os.environ:
473 self._write_string('\033]0;%s\007' % message, self._screen_file)
475 def save_console_title(self):
476 if not self.params.get('consoletitle', False):
478 if 'TERM' in os.environ:
479 # Save the title on stack
480 self._write_string('\033[22;0t', self._screen_file)
482 def restore_console_title(self):
483 if not self.params.get('consoletitle', False):
485 if 'TERM' in os.environ:
486 # Restore the title from stack
487 self._write_string('\033[23;0t', self._screen_file)
490 self.save_console_title()
493 def __exit__(self, *args):
494 self.restore_console_title()
496 if self.params.get('cookiefile') is not None:
497 self.cookiejar.save()
499 def trouble(self, message=None, tb=None):
500 """Determine action to take when a download problem appears.
502 Depending on if the downloader has been configured to ignore
503 download errors or not, this method may throw an exception or
504 not when errors are found, after printing the message.
506 tb, if given, is additional traceback information.
508 if message is not None:
509 self.to_stderr(message)
510 if self.params.get('verbose'):
512 if sys.exc_info()[0]: # if .trouble has been called from an except block
514 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
515 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
516 tb += encode_compat_str(traceback.format_exc())
518 tb_data = traceback.format_list(traceback.extract_stack())
519 tb = ''.join(tb_data)
521 if not self.params.get('ignoreerrors', False):
522 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
523 exc_info = sys.exc_info()[1].exc_info
525 exc_info = sys.exc_info()
526 raise DownloadError(message, exc_info)
527 self._download_retcode = 1
529 def report_warning(self, message):
531 Print the message to stderr, it will be prefixed with 'WARNING:'
532 If stderr is a tty file the 'WARNING:' will be colored
534 if self.params.get('logger') is not None:
535 self.params['logger'].warning(message)
537 if self.params.get('no_warnings'):
539 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
540 _msg_header = '\033[0;33mWARNING:\033[0m'
542 _msg_header = 'WARNING:'
543 warning_message = '%s %s' % (_msg_header, message)
544 self.to_stderr(warning_message)
546 def report_error(self, message, tb=None):
548 Do the same as trouble, but prefixes the message with 'ERROR:', colored
549 in red if stderr is a tty file.
551 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
552 _msg_header = '\033[0;31mERROR:\033[0m'
554 _msg_header = 'ERROR:'
555 error_message = '%s %s' % (_msg_header, message)
556 self.trouble(error_message, tb)
558 def report_file_already_downloaded(self, file_name):
559 """Report file has already been fully downloaded."""
561 self.to_screen('[download] %s has already been downloaded' % file_name)
562 except UnicodeEncodeError:
563 self.to_screen('[download] The file has already been downloaded')
565 def prepare_filename(self, info_dict):
566 """Generate the output filename."""
568 template_dict = dict(info_dict)
570 template_dict['epoch'] = int(time.time())
571 autonumber_size = self.params.get('autonumber_size')
572 if autonumber_size is None:
574 autonumber_templ = '%0' + str(autonumber_size) + 'd'
575 template_dict['autonumber'] = autonumber_templ % self._num_downloads
576 if template_dict.get('playlist_index') is not None:
577 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
578 if template_dict.get('resolution') is None:
579 if template_dict.get('width') and template_dict.get('height'):
580 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
581 elif template_dict.get('height'):
582 template_dict['resolution'] = '%sp' % template_dict['height']
583 elif template_dict.get('width'):
584 template_dict['resolution'] = '%dx?' % template_dict['width']
586 sanitize = lambda k, v: sanitize_filename(
588 restricted=self.params.get('restrictfilenames'),
590 template_dict = dict((k, sanitize(k, v))
591 for k, v in template_dict.items()
592 if v is not None and not isinstance(v, (list, tuple, dict)))
593 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
595 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
596 tmpl = compat_expanduser(outtmpl)
597 filename = tmpl % template_dict
598 # Temporary fix for #4787
599 # 'Treat' all problem characters by passing filename through preferredencoding
600 # to workaround encoding issues with subprocess on python2 @ Windows
601 if sys.version_info < (3, 0) and sys.platform == 'win32':
602 filename = encodeFilename(filename, True).decode(preferredencoding())
603 return sanitize_path(filename)
604 except ValueError as err:
605 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
608 def _match_entry(self, info_dict, incomplete):
609 """ Returns None iff the file should be downloaded """
611 video_title = info_dict.get('title', info_dict.get('id', 'video'))
612 if 'title' in info_dict:
613 # This can happen when we're just evaluating the playlist
614 title = info_dict['title']
615 matchtitle = self.params.get('matchtitle', False)
617 if not re.search(matchtitle, title, re.IGNORECASE):
618 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
619 rejecttitle = self.params.get('rejecttitle', False)
621 if re.search(rejecttitle, title, re.IGNORECASE):
622 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
623 date = info_dict.get('upload_date')
625 dateRange = self.params.get('daterange', DateRange())
626 if date not in dateRange:
627 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
628 view_count = info_dict.get('view_count')
629 if view_count is not None:
630 min_views = self.params.get('min_views')
631 if min_views is not None and view_count < min_views:
632 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
633 max_views = self.params.get('max_views')
634 if max_views is not None and view_count > max_views:
635 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
636 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
637 return 'Skipping "%s" because it is age restricted' % video_title
638 if self.in_download_archive(info_dict):
639 return '%s has already been recorded in archive' % video_title
642 match_filter = self.params.get('match_filter')
643 if match_filter is not None:
644 ret = match_filter(info_dict)
651 def add_extra_info(info_dict, extra_info):
652 '''Set the keys from extra_info in info dict if they are missing'''
653 for key, value in extra_info.items():
654 info_dict.setdefault(key, value)
656 def extract_info(self, url, download=True, ie_key=None, extra_info={},
657 process=True, force_generic_extractor=False):
659 Returns a list with a dictionary for each video we find.
660 If 'download', also downloads the videos.
661 extra_info is a dict containing the extra values to add to each result
664 if not ie_key and force_generic_extractor:
668 ies = [self.get_info_extractor(ie_key)]
673 if not ie.suitable(url):
676 ie = self.get_info_extractor(ie.ie_key())
678 self.report_warning('The program functionality for this site has been marked as broken, '
679 'and will probably not work.')
682 ie_result = ie.extract(url)
683 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
685 if isinstance(ie_result, list):
686 # Backwards compatibility: old IE result format
688 '_type': 'compat_list',
689 'entries': ie_result,
691 self.add_default_extra_info(ie_result, ie, url)
693 return self.process_ie_result(ie_result, download, extra_info)
696 except ExtractorError as e: # An error we somewhat expected
697 self.report_error(compat_str(e), e.format_traceback())
699 except MaxDownloadsReached:
701 except Exception as e:
702 if self.params.get('ignoreerrors', False):
703 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
708 self.report_error('no suitable InfoExtractor for URL %s' % url)
710 def add_default_extra_info(self, ie_result, ie, url):
711 self.add_extra_info(ie_result, {
712 'extractor': ie.IE_NAME,
714 'webpage_url_basename': url_basename(url),
715 'extractor_key': ie.ie_key(),
718 def process_ie_result(self, ie_result, download=True, extra_info={}):
720 Take the result of the ie(may be modified) and resolve all unresolved
721 references (URLs, playlist items).
723 It will also download the videos if 'download'.
724 Returns the resolved ie_result.
726 result_type = ie_result.get('_type', 'video')
728 if result_type in ('url', 'url_transparent'):
729 ie_result['url'] = sanitize_url(ie_result['url'])
730 extract_flat = self.params.get('extract_flat', False)
731 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
732 extract_flat is True):
733 if self.params.get('forcejson', False):
734 self.to_stdout(json.dumps(ie_result))
737 if result_type == 'video':
738 self.add_extra_info(ie_result, extra_info)
739 return self.process_video_result(ie_result, download=download)
740 elif result_type == 'url':
741 # We have to add extra_info to the results because it may be
742 # contained in a playlist
743 return self.extract_info(ie_result['url'],
745 ie_key=ie_result.get('ie_key'),
746 extra_info=extra_info)
747 elif result_type == 'url_transparent':
748 # Use the information from the embedding page
749 info = self.extract_info(
750 ie_result['url'], ie_key=ie_result.get('ie_key'),
751 extra_info=extra_info, download=False, process=False)
753 force_properties = dict(
754 (k, v) for k, v in ie_result.items() if v is not None)
755 for f in ('_type', 'url', 'ie_key'):
756 if f in force_properties:
757 del force_properties[f]
758 new_result = info.copy()
759 new_result.update(force_properties)
761 assert new_result.get('_type') != 'url_transparent'
763 return self.process_ie_result(
764 new_result, download=download, extra_info=extra_info)
765 elif result_type == 'playlist' or result_type == 'multi_video':
766 # We process each entry in the playlist
767 playlist = ie_result.get('title') or ie_result.get('id')
768 self.to_screen('[download] Downloading playlist: %s' % playlist)
770 playlist_results = []
772 playliststart = self.params.get('playliststart', 1) - 1
773 playlistend = self.params.get('playlistend')
774 # For backwards compatibility, interpret -1 as whole list
775 if playlistend == -1:
778 playlistitems_str = self.params.get('playlist_items')
780 if playlistitems_str is not None:
781 def iter_playlistitems(format):
782 for string_segment in format.split(','):
783 if '-' in string_segment:
784 start, end = string_segment.split('-')
785 for item in range(int(start), int(end) + 1):
788 yield int(string_segment)
789 playlistitems = iter_playlistitems(playlistitems_str)
791 ie_entries = ie_result['entries']
792 if isinstance(ie_entries, list):
793 n_all_entries = len(ie_entries)
796 ie_entries[i - 1] for i in playlistitems
797 if -n_all_entries <= i - 1 < n_all_entries]
799 entries = ie_entries[playliststart:playlistend]
800 n_entries = len(entries)
802 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
803 (ie_result['extractor'], playlist, n_all_entries, n_entries))
804 elif isinstance(ie_entries, PagedList):
807 for item in playlistitems:
808 entries.extend(ie_entries.getslice(
812 entries = ie_entries.getslice(
813 playliststart, playlistend)
814 n_entries = len(entries)
816 '[%s] playlist %s: Downloading %d videos' %
817 (ie_result['extractor'], playlist, n_entries))
820 entry_list = list(ie_entries)
821 entries = [entry_list[i - 1] for i in playlistitems]
823 entries = list(itertools.islice(
824 ie_entries, playliststart, playlistend))
825 n_entries = len(entries)
827 '[%s] playlist %s: Downloading %d videos' %
828 (ie_result['extractor'], playlist, n_entries))
830 if self.params.get('playlistreverse', False):
831 entries = entries[::-1]
833 for i, entry in enumerate(entries, 1):
834 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
836 'n_entries': n_entries,
837 'playlist': playlist,
838 'playlist_id': ie_result.get('id'),
839 'playlist_title': ie_result.get('title'),
840 'playlist_index': i + playliststart,
841 'extractor': ie_result['extractor'],
842 'webpage_url': ie_result['webpage_url'],
843 'webpage_url_basename': url_basename(ie_result['webpage_url']),
844 'extractor_key': ie_result['extractor_key'],
847 reason = self._match_entry(entry, incomplete=True)
848 if reason is not None:
849 self.to_screen('[download] ' + reason)
852 entry_result = self.process_ie_result(entry,
855 playlist_results.append(entry_result)
856 ie_result['entries'] = playlist_results
857 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
859 elif result_type == 'compat_list':
861 'Extractor %s returned a compat_list result. '
862 'It needs to be updated.' % ie_result.get('extractor'))
868 'extractor': ie_result['extractor'],
869 'webpage_url': ie_result['webpage_url'],
870 'webpage_url_basename': url_basename(ie_result['webpage_url']),
871 'extractor_key': ie_result['extractor_key'],
875 ie_result['entries'] = [
876 self.process_ie_result(_fixup(r), download, extra_info)
877 for r in ie_result['entries']
881 raise Exception('Invalid result type: %s' % result_type)
883 def _build_format_filter(self, filter_spec):
884 " Returns a function to filter the formats according to the filter_spec "
894 operator_rex = re.compile(r'''(?x)\s*
895 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
896 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
897 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
899 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
900 m = operator_rex.search(filter_spec)
903 comparison_value = int(m.group('value'))
905 comparison_value = parse_filesize(m.group('value'))
906 if comparison_value is None:
907 comparison_value = parse_filesize(m.group('value') + 'B')
908 if comparison_value is None:
910 'Invalid value %r in format specification %r' % (
911 m.group('value'), filter_spec))
912 op = OPERATORS[m.group('op')]
918 '^=': lambda attr, value: attr.startswith(value),
919 '$=': lambda attr, value: attr.endswith(value),
920 '*=': lambda attr, value: value in attr,
922 str_operator_rex = re.compile(r'''(?x)
923 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
924 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
925 \s*(?P<value>[a-zA-Z0-9._-]+)
927 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
928 m = str_operator_rex.search(filter_spec)
930 comparison_value = m.group('value')
931 op = STR_OPERATORS[m.group('op')]
934 raise ValueError('Invalid filter specification %r' % filter_spec)
937 actual_value = f.get(m.group('key'))
938 if actual_value is None:
939 return m.group('none_inclusive')
940 return op(actual_value, comparison_value)
943 def build_format_selector(self, format_spec):
944 def syntax_error(note, start):
946 'Invalid format specification: '
947 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
948 return SyntaxError(message)
950 PICKFIRST = 'PICKFIRST'
954 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
956 def _parse_filter(tokens):
958 for type, string, start, _, _ in tokens:
959 if type == tokenize.OP and string == ']':
960 return ''.join(filter_parts)
962 filter_parts.append(string)
964 def _remove_unused_ops(tokens):
965 # Remove operators that we don't use and join them with the surrounding strings
966 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
967 ALLOWED_OPS = ('/', '+', ',', '(', ')')
968 last_string, last_start, last_end, last_line = None, None, None, None
969 for type, string, start, end, line in tokens:
970 if type == tokenize.OP and string == '[':
972 yield tokenize.NAME, last_string, last_start, last_end, last_line
974 yield type, string, start, end, line
975 # everything inside brackets will be handled by _parse_filter
976 for type, string, start, end, line in tokens:
977 yield type, string, start, end, line
978 if type == tokenize.OP and string == ']':
980 elif type == tokenize.OP and string in ALLOWED_OPS:
982 yield tokenize.NAME, last_string, last_start, last_end, last_line
984 yield type, string, start, end, line
985 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
991 last_string += string
993 yield tokenize.NAME, last_string, last_start, last_end, last_line
995 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
997 current_selector = None
998 for type, string, start, _, _ in tokens:
999 # ENCODING is only defined in python 3.x
1000 if type == getattr(tokenize, 'ENCODING', None):
1002 elif type in [tokenize.NAME, tokenize.NUMBER]:
1003 current_selector = FormatSelector(SINGLE, string, [])
1004 elif type == tokenize.OP:
1006 if not inside_group:
1007 # ')' will be handled by the parentheses group
1008 tokens.restore_last_token()
1010 elif inside_merge and string in ['/', ',']:
1011 tokens.restore_last_token()
1013 elif inside_choice and string == ',':
1014 tokens.restore_last_token()
1017 if not current_selector:
1018 raise syntax_error('"," must follow a format selector', start)
1019 selectors.append(current_selector)
1020 current_selector = None
1022 if not current_selector:
1023 raise syntax_error('"/" must follow a format selector', start)
1024 first_choice = current_selector
1025 second_choice = _parse_format_selection(tokens, inside_choice=True)
1026 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1028 if not current_selector:
1029 current_selector = FormatSelector(SINGLE, 'best', [])
1030 format_filter = _parse_filter(tokens)
1031 current_selector.filters.append(format_filter)
1033 if current_selector:
1034 raise syntax_error('Unexpected "("', start)
1035 group = _parse_format_selection(tokens, inside_group=True)
1036 current_selector = FormatSelector(GROUP, group, [])
1038 video_selector = current_selector
1039 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1040 if not video_selector or not audio_selector:
1041 raise syntax_error('"+" must be between two format selectors', start)
1042 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1044 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1045 elif type == tokenize.ENDMARKER:
1047 if current_selector:
1048 selectors.append(current_selector)
1051 def _build_selector_function(selector):
1052 if isinstance(selector, list):
1053 fs = [_build_selector_function(s) for s in selector]
1055 def selector_function(ctx):
1057 for format in f(ctx):
1059 return selector_function
1060 elif selector.type == GROUP:
1061 selector_function = _build_selector_function(selector.selector)
1062 elif selector.type == PICKFIRST:
1063 fs = [_build_selector_function(s) for s in selector.selector]
1065 def selector_function(ctx):
1067 picked_formats = list(f(ctx))
1069 return picked_formats
1071 elif selector.type == SINGLE:
1072 format_spec = selector.selector
1074 def selector_function(ctx):
1075 formats = list(ctx['formats'])
1078 if format_spec == 'all':
1081 elif format_spec in ['best', 'worst', None]:
1082 format_idx = 0 if format_spec == 'worst' else -1
1083 audiovideo_formats = [
1085 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1086 if audiovideo_formats:
1087 yield audiovideo_formats[format_idx]
1088 # for extractors with incomplete formats (audio only (soundcloud)
1089 # or video only (imgur)) we will fallback to best/worst
1090 # {video,audio}-only format
1091 elif ctx['incomplete_formats']:
1092 yield formats[format_idx]
1093 elif format_spec == 'bestaudio':
1096 if f.get('vcodec') == 'none']
1098 yield audio_formats[-1]
1099 elif format_spec == 'worstaudio':
1102 if f.get('vcodec') == 'none']
1104 yield audio_formats[0]
1105 elif format_spec == 'bestvideo':
1108 if f.get('acodec') == 'none']
1110 yield video_formats[-1]
1111 elif format_spec == 'worstvideo':
1114 if f.get('acodec') == 'none']
1116 yield video_formats[0]
1118 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1119 if format_spec in extensions:
1120 filter_f = lambda f: f['ext'] == format_spec
1122 filter_f = lambda f: f['format_id'] == format_spec
1123 matches = list(filter(filter_f, formats))
1126 elif selector.type == MERGE:
1127 def _merge(formats_info):
1128 format_1, format_2 = [f['format_id'] for f in formats_info]
1129 # The first format must contain the video and the
1131 if formats_info[0].get('vcodec') == 'none':
1132 self.report_error('The first format must '
1133 'contain the video, try using '
1134 '"-f %s+%s"' % (format_2, format_1))
1136 # Formats must be opposite (video+audio)
1137 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1139 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1140 % (format_1, format_2))
1143 formats_info[0]['ext']
1144 if self.params.get('merge_output_format') is None
1145 else self.params['merge_output_format'])
1147 'requested_formats': formats_info,
1148 'format': '%s+%s' % (formats_info[0].get('format'),
1149 formats_info[1].get('format')),
1150 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1151 formats_info[1].get('format_id')),
1152 'width': formats_info[0].get('width'),
1153 'height': formats_info[0].get('height'),
1154 'resolution': formats_info[0].get('resolution'),
1155 'fps': formats_info[0].get('fps'),
1156 'vcodec': formats_info[0].get('vcodec'),
1157 'vbr': formats_info[0].get('vbr'),
1158 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1159 'acodec': formats_info[1].get('acodec'),
1160 'abr': formats_info[1].get('abr'),
1163 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1165 def selector_function(ctx):
1166 for pair in itertools.product(
1167 video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
1170 filters = [self._build_format_filter(f) for f in selector.filters]
1172 def final_selector(ctx):
1173 ctx_copy = copy.deepcopy(ctx)
1174 for _filter in filters:
1175 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1176 return selector_function(ctx_copy)
1177 return final_selector
1179 stream = io.BytesIO(format_spec.encode('utf-8'))
1181 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1182 except tokenize.TokenError:
1183 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1185 class TokenIterator(object):
1186 def __init__(self, tokens):
1187 self.tokens = tokens
1194 if self.counter >= len(self.tokens):
1195 raise StopIteration()
1196 value = self.tokens[self.counter]
1202 def restore_last_token(self):
1205 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1206 return _build_selector_function(parsed_selector)
1208 def _calc_headers(self, info_dict):
1209 res = std_headers.copy()
1211 add_headers = info_dict.get('http_headers')
1213 res.update(add_headers)
1215 cookies = self._calc_cookies(info_dict)
1217 res['Cookie'] = cookies
1221 def _calc_cookies(self, info_dict):
1222 pr = sanitized_Request(info_dict['url'])
1223 self.cookiejar.add_cookie_header(pr)
1224 return pr.get_header('Cookie')
1226 def process_video_result(self, info_dict, download=True):
1227 assert info_dict.get('_type', 'video') == 'video'
1229 if 'id' not in info_dict:
1230 raise ExtractorError('Missing "id" field in extractor result')
1231 if 'title' not in info_dict:
1232 raise ExtractorError('Missing "title" field in extractor result')
1234 if not isinstance(info_dict['id'], compat_str):
1235 self.report_warning('"id" field is not a string - forcing string conversion')
1236 info_dict['id'] = compat_str(info_dict['id'])
1238 if 'playlist' not in info_dict:
1239 # It isn't part of a playlist
1240 info_dict['playlist'] = None
1241 info_dict['playlist_index'] = None
1243 thumbnails = info_dict.get('thumbnails')
1244 if thumbnails is None:
1245 thumbnail = info_dict.get('thumbnail')
1247 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1249 thumbnails.sort(key=lambda t: (
1250 t.get('preference'), t.get('width'), t.get('height'),
1251 t.get('id'), t.get('url')))
1252 for i, t in enumerate(thumbnails):
1253 t['url'] = sanitize_url(t['url'])
1254 if t.get('width') and t.get('height'):
1255 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1256 if t.get('id') is None:
1259 if self.params.get('list_thumbnails'):
1260 self.list_thumbnails(info_dict)
1263 thumbnail = info_dict.get('thumbnail')
1265 info_dict['thumbnail'] = sanitize_url(thumbnail)
1267 info_dict['thumbnail'] = thumbnails[-1]['url']
1269 if 'display_id' not in info_dict and 'id' in info_dict:
1270 info_dict['display_id'] = info_dict['id']
1272 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1273 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1274 # see http://bugs.python.org/issue1646728)
1276 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1277 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1278 except (ValueError, OverflowError, OSError):
1281 # Auto generate title fields corresponding to the *_number fields when missing
1282 # in order to always have clean titles. This is very common for TV series.
1283 for field in ('chapter', 'season', 'episode'):
1284 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1285 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1287 subtitles = info_dict.get('subtitles')
1289 for _, subtitle in subtitles.items():
1290 for subtitle_format in subtitle:
1291 if subtitle_format.get('url'):
1292 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1293 if 'ext' not in subtitle_format:
1294 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1296 if self.params.get('listsubtitles', False):
1297 if 'automatic_captions' in info_dict:
1298 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1299 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1301 info_dict['requested_subtitles'] = self.process_subtitles(
1302 info_dict['id'], subtitles,
1303 info_dict.get('automatic_captions'))
1305 # We now pick which formats have to be downloaded
1306 if info_dict.get('formats') is None:
1307 # There's only one format available
1308 formats = [info_dict]
1310 formats = info_dict['formats']
1313 raise ExtractorError('No video formats found!')
1317 # We check that all the formats have the format and format_id fields
1318 for i, format in enumerate(formats):
1319 if 'url' not in format:
1320 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1322 format['url'] = sanitize_url(format['url'])
1324 if format.get('format_id') is None:
1325 format['format_id'] = compat_str(i)
1327 # Sanitize format_id from characters used in format selector expression
1328 format['format_id'] = re.sub('[\s,/+\[\]()]', '_', format['format_id'])
1329 format_id = format['format_id']
1330 if format_id not in formats_dict:
1331 formats_dict[format_id] = []
1332 formats_dict[format_id].append(format)
1334 # Make sure all formats have unique format_id
1335 for format_id, ambiguous_formats in formats_dict.items():
1336 if len(ambiguous_formats) > 1:
1337 for i, format in enumerate(ambiguous_formats):
1338 format['format_id'] = '%s-%d' % (format_id, i)
1340 for i, format in enumerate(formats):
1341 if format.get('format') is None:
1342 format['format'] = '{id} - {res}{note}'.format(
1343 id=format['format_id'],
1344 res=self.format_resolution(format),
1345 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1347 # Automatically determine file extension if missing
1348 if 'ext' not in format:
1349 format['ext'] = determine_ext(format['url']).lower()
1350 # Automatically determine protocol if missing (useful for format
1351 # selection purposes)
1352 if 'protocol' not in format:
1353 format['protocol'] = determine_protocol(format)
1354 # Add HTTP headers, so that external programs can use them from the
1356 full_format_info = info_dict.copy()
1357 full_format_info.update(format)
1358 format['http_headers'] = self._calc_headers(full_format_info)
1360 # TODO Central sorting goes here
1362 if formats[0] is not info_dict:
1363 # only set the 'formats' fields if the original info_dict list them
1364 # otherwise we end up with a circular reference, the first (and unique)
1365 # element in the 'formats' field in info_dict is info_dict itself,
1366 # which can't be exported to json
1367 info_dict['formats'] = formats
1368 if self.params.get('listformats'):
1369 self.list_formats(info_dict)
1372 req_format = self.params.get('format')
1373 if req_format is None:
1374 req_format_list = []
1375 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1376 not info_dict.get('is_live')):
1377 merger = FFmpegMergerPP(self)
1378 if merger.available and merger.can_merge():
1379 req_format_list.append('bestvideo+bestaudio')
1380 req_format_list.append('best')
1381 req_format = '/'.join(req_format_list)
1382 format_selector = self.build_format_selector(req_format)
1384 # While in format selection we may need to have an access to the original
1385 # format set in order to calculate some metrics or do some processing.
1386 # For now we need to be able to guess whether original formats provided
1387 # by extractor are incomplete or not (i.e. whether extractor provides only
1388 # video-only or audio-only formats) for proper formats selection for
1389 # extractors with such incomplete formats (see
1390 # https://github.com/rg3/youtube-dl/pull/5556).
1391 # Since formats may be filtered during format selection and may not match
1392 # the original formats the results may be incorrect. Thus original formats
1393 # or pre-calculated metrics should be passed to format selection routines
1395 # We will pass a context object containing all necessary additional data
1396 # instead of just formats.
1397 # This fixes incorrect format selection issue (see
1398 # https://github.com/rg3/youtube-dl/issues/10083).
1399 incomplete_formats = all(
1400 # All formats are video-only or
1401 f.get('vcodec') != 'none' and f.get('acodec') == 'none' or
1402 # all formats are audio-only
1403 f.get('vcodec') == 'none' and f.get('acodec') != 'none'
1408 'incomplete_formats': incomplete_formats,
1411 formats_to_download = list(format_selector(ctx))
1412 if not formats_to_download:
1413 raise ExtractorError('requested format not available',
1417 if len(formats_to_download) > 1:
1418 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1419 for format in formats_to_download:
1420 new_info = dict(info_dict)
1421 new_info.update(format)
1422 self.process_info(new_info)
1423 # We update the info dict with the best quality format (backwards compatibility)
1424 info_dict.update(formats_to_download[-1])
1427 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1428 """Select the requested subtitles and their format"""
1430 if normal_subtitles and self.params.get('writesubtitles'):
1431 available_subs.update(normal_subtitles)
1432 if automatic_captions and self.params.get('writeautomaticsub'):
1433 for lang, cap_info in automatic_captions.items():
1434 if lang not in available_subs:
1435 available_subs[lang] = cap_info
1437 if (not self.params.get('writesubtitles') and not
1438 self.params.get('writeautomaticsub') or not
1442 if self.params.get('allsubtitles', False):
1443 requested_langs = available_subs.keys()
1445 if self.params.get('subtitleslangs', False):
1446 requested_langs = self.params.get('subtitleslangs')
1447 elif 'en' in available_subs:
1448 requested_langs = ['en']
1450 requested_langs = [list(available_subs.keys())[0]]
1452 formats_query = self.params.get('subtitlesformat', 'best')
1453 formats_preference = formats_query.split('/') if formats_query else []
1455 for lang in requested_langs:
1456 formats = available_subs.get(lang)
1458 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1460 for ext in formats_preference:
1464 matches = list(filter(lambda f: f['ext'] == ext, formats))
1470 self.report_warning(
1471 'No subtitle format found matching "%s" for language %s, '
1472 'using %s' % (formats_query, lang, f['ext']))
1476 def process_info(self, info_dict):
1477 """Process a single resolved IE result."""
1479 assert info_dict.get('_type', 'video') == 'video'
1481 max_downloads = self.params.get('max_downloads')
1482 if max_downloads is not None:
1483 if self._num_downloads >= int(max_downloads):
1484 raise MaxDownloadsReached()
1486 info_dict['fulltitle'] = info_dict['title']
1487 if len(info_dict['title']) > 200:
1488 info_dict['title'] = info_dict['title'][:197] + '...'
1490 if 'format' not in info_dict:
1491 info_dict['format'] = info_dict['ext']
1493 reason = self._match_entry(info_dict, incomplete=False)
1494 if reason is not None:
1495 self.to_screen('[download] ' + reason)
1498 self._num_downloads += 1
1500 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1503 if self.params.get('forcetitle', False):
1504 self.to_stdout(info_dict['fulltitle'])
1505 if self.params.get('forceid', False):
1506 self.to_stdout(info_dict['id'])
1507 if self.params.get('forceurl', False):
1508 if info_dict.get('requested_formats') is not None:
1509 for f in info_dict['requested_formats']:
1510 self.to_stdout(f['url'] + f.get('play_path', ''))
1512 # For RTMP URLs, also include the playpath
1513 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1514 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1515 self.to_stdout(info_dict['thumbnail'])
1516 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1517 self.to_stdout(info_dict['description'])
1518 if self.params.get('forcefilename', False) and filename is not None:
1519 self.to_stdout(filename)
1520 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1521 self.to_stdout(formatSeconds(info_dict['duration']))
1522 if self.params.get('forceformat', False):
1523 self.to_stdout(info_dict['format'])
1524 if self.params.get('forcejson', False):
1525 self.to_stdout(json.dumps(info_dict))
1527 # Do nothing else if in simulate mode
1528 if self.params.get('simulate', False):
1531 if filename is None:
1535 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1536 if dn and not os.path.exists(dn):
1538 except (OSError, IOError) as err:
1539 self.report_error('unable to create directory ' + error_to_compat_str(err))
1542 if self.params.get('writedescription', False):
1543 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1544 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1545 self.to_screen('[info] Video description is already present')
1546 elif info_dict.get('description') is None:
1547 self.report_warning('There\'s no description to write.')
1550 self.to_screen('[info] Writing video description to: ' + descfn)
1551 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1552 descfile.write(info_dict['description'])
1553 except (OSError, IOError):
1554 self.report_error('Cannot write description file ' + descfn)
1557 if self.params.get('writeannotations', False):
1558 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1559 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1560 self.to_screen('[info] Video annotations are already present')
1563 self.to_screen('[info] Writing video annotations to: ' + annofn)
1564 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1565 annofile.write(info_dict['annotations'])
1566 except (KeyError, TypeError):
1567 self.report_warning('There are no annotations to write.')
1568 except (OSError, IOError):
1569 self.report_error('Cannot write annotations file: ' + annofn)
1572 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1573 self.params.get('writeautomaticsub')])
1575 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1576 # subtitles download errors are already managed as troubles in relevant IE
1577 # that way it will silently go on when used with unsupporting IE
1578 subtitles = info_dict['requested_subtitles']
1579 ie = self.get_info_extractor(info_dict['extractor_key'])
1580 for sub_lang, sub_info in subtitles.items():
1581 sub_format = sub_info['ext']
1582 if sub_info.get('data') is not None:
1583 sub_data = sub_info['data']
1586 sub_data = ie._download_webpage(
1587 sub_info['url'], info_dict['id'], note=False)
1588 except ExtractorError as err:
1589 self.report_warning('Unable to download subtitle for "%s": %s' %
1590 (sub_lang, error_to_compat_str(err.cause)))
1593 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1594 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1595 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1597 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1598 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1599 subfile.write(sub_data)
1600 except (OSError, IOError):
1601 self.report_error('Cannot write subtitles file ' + sub_filename)
1604 if self.params.get('writeinfojson', False):
1605 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1606 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1607 self.to_screen('[info] Video description metadata is already present')
1609 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1611 write_json_file(self.filter_requested_info(info_dict), infofn)
1612 except (OSError, IOError):
1613 self.report_error('Cannot write metadata to JSON file ' + infofn)
1616 self._write_thumbnails(info_dict, filename)
1618 if not self.params.get('skip_download', False):
1621 fd = get_suitable_downloader(info, self.params)(self, self.params)
1622 for ph in self._progress_hooks:
1623 fd.add_progress_hook(ph)
1624 if self.params.get('verbose'):
1625 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1626 return fd.download(name, info)
1628 if info_dict.get('requested_formats') is not None:
1631 merger = FFmpegMergerPP(self)
1632 if not merger.available:
1634 self.report_warning('You have requested multiple '
1635 'formats but ffmpeg or avconv are not installed.'
1636 ' The formats won\'t be merged.')
1638 postprocessors = [merger]
1640 def compatible_formats(formats):
1641 video, audio = formats
1643 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1644 if video_ext and audio_ext:
1646 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1649 for exts in COMPATIBLE_EXTS:
1650 if video_ext in exts and audio_ext in exts:
1652 # TODO: Check acodec/vcodec
1655 filename_real_ext = os.path.splitext(filename)[1][1:]
1657 os.path.splitext(filename)[0]
1658 if filename_real_ext == info_dict['ext']
1660 requested_formats = info_dict['requested_formats']
1661 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1662 info_dict['ext'] = 'mkv'
1663 self.report_warning(
1664 'Requested formats are incompatible for merge and will be merged into mkv.')
1665 # Ensure filename always has a correct extension for successful merge
1666 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1667 if os.path.exists(encodeFilename(filename)):
1669 '[download] %s has already been downloaded and '
1670 'merged' % filename)
1672 for f in requested_formats:
1673 new_info = dict(info_dict)
1675 fname = self.prepare_filename(new_info)
1676 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1677 downloaded.append(fname)
1678 partial_success = dl(fname, new_info)
1679 success = success and partial_success
1680 info_dict['__postprocessors'] = postprocessors
1681 info_dict['__files_to_merge'] = downloaded
1683 # Just a single file
1684 success = dl(filename, info_dict)
1685 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1686 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1688 except (OSError, IOError) as err:
1689 raise UnavailableVideoError(err)
1690 except (ContentTooShortError, ) as err:
1691 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1694 if success and filename != '-':
1696 fixup_policy = self.params.get('fixup')
1697 if fixup_policy is None:
1698 fixup_policy = 'detect_or_warn'
1700 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1702 stretched_ratio = info_dict.get('stretched_ratio')
1703 if stretched_ratio is not None and stretched_ratio != 1:
1704 if fixup_policy == 'warn':
1705 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1706 info_dict['id'], stretched_ratio))
1707 elif fixup_policy == 'detect_or_warn':
1708 stretched_pp = FFmpegFixupStretchedPP(self)
1709 if stretched_pp.available:
1710 info_dict.setdefault('__postprocessors', [])
1711 info_dict['__postprocessors'].append(stretched_pp)
1713 self.report_warning(
1714 '%s: Non-uniform pixel ratio (%s). %s'
1715 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1717 assert fixup_policy in ('ignore', 'never')
1719 if (info_dict.get('requested_formats') is None and
1720 info_dict.get('container') == 'm4a_dash'):
1721 if fixup_policy == 'warn':
1722 self.report_warning(
1723 '%s: writing DASH m4a. '
1724 'Only some players support this container.'
1726 elif fixup_policy == 'detect_or_warn':
1727 fixup_pp = FFmpegFixupM4aPP(self)
1728 if fixup_pp.available:
1729 info_dict.setdefault('__postprocessors', [])
1730 info_dict['__postprocessors'].append(fixup_pp)
1732 self.report_warning(
1733 '%s: writing DASH m4a. '
1734 'Only some players support this container. %s'
1735 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1737 assert fixup_policy in ('ignore', 'never')
1739 if (info_dict.get('protocol') == 'm3u8_native' or
1740 info_dict.get('protocol') == 'm3u8' and
1741 self.params.get('hls_prefer_native')):
1742 if fixup_policy == 'warn':
1743 self.report_warning('%s: malformated aac bitstream.' % (
1745 elif fixup_policy == 'detect_or_warn':
1746 fixup_pp = FFmpegFixupM3u8PP(self)
1747 if fixup_pp.available:
1748 info_dict.setdefault('__postprocessors', [])
1749 info_dict['__postprocessors'].append(fixup_pp)
1751 self.report_warning(
1752 '%s: malformated aac bitstream. %s'
1753 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1755 assert fixup_policy in ('ignore', 'never')
1758 self.post_process(filename, info_dict)
1759 except (PostProcessingError) as err:
1760 self.report_error('postprocessing: %s' % str(err))
1762 self.record_download_archive(info_dict)
1764 def download(self, url_list):
1765 """Download a given list of URLs."""
1766 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1767 if (len(url_list) > 1 and
1768 '%' not in outtmpl and
1769 self.params.get('max_downloads') != 1):
1770 raise SameFileError(outtmpl)
1772 for url in url_list:
1774 # It also downloads the videos
1775 res = self.extract_info(
1776 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1777 except UnavailableVideoError:
1778 self.report_error('unable to download video')
1779 except MaxDownloadsReached:
1780 self.to_screen('[info] Maximum number of downloaded files reached.')
1783 if self.params.get('dump_single_json', False):
1784 self.to_stdout(json.dumps(res))
1786 return self._download_retcode
1788 def download_with_info_file(self, info_filename):
1789 with contextlib.closing(fileinput.FileInput(
1790 [info_filename], mode='r',
1791 openhook=fileinput.hook_encoded('utf-8'))) as f:
1792 # FileInput doesn't have a read method, we can't call json.load
1793 info = self.filter_requested_info(json.loads('\n'.join(f)))
1795 self.process_ie_result(info, download=True)
1796 except DownloadError:
1797 webpage_url = info.get('webpage_url')
1798 if webpage_url is not None:
1799 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1800 return self.download([webpage_url])
1803 return self._download_retcode
1806 def filter_requested_info(info_dict):
1808 (k, v) for k, v in info_dict.items()
1809 if k not in ['requested_formats', 'requested_subtitles'])
1811 def post_process(self, filename, ie_info):
1812 """Run all the postprocessors on the given file."""
1813 info = dict(ie_info)
1814 info['filepath'] = filename
1816 if ie_info.get('__postprocessors') is not None:
1817 pps_chain.extend(ie_info['__postprocessors'])
1818 pps_chain.extend(self._pps)
1819 for pp in pps_chain:
1820 files_to_delete = []
1822 files_to_delete, info = pp.run(info)
1823 except PostProcessingError as e:
1824 self.report_error(e.msg)
1825 if files_to_delete and not self.params.get('keepvideo', False):
1826 for old_filename in files_to_delete:
1827 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1829 os.remove(encodeFilename(old_filename))
1830 except (IOError, OSError):
1831 self.report_warning('Unable to remove downloaded original file')
1833 def _make_archive_id(self, info_dict):
1834 # Future-proof against any change in case
1835 # and backwards compatibility with prior versions
1836 extractor = info_dict.get('extractor_key')
1837 if extractor is None:
1838 if 'id' in info_dict:
1839 extractor = info_dict.get('ie_key') # key in a playlist
1840 if extractor is None:
1841 return None # Incomplete video information
1842 return extractor.lower() + ' ' + info_dict['id']
1844 def in_download_archive(self, info_dict):
1845 fn = self.params.get('download_archive')
1849 vid_id = self._make_archive_id(info_dict)
1851 return False # Incomplete video information
1854 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1855 for line in archive_file:
1856 if line.strip() == vid_id:
1858 except IOError as ioe:
1859 if ioe.errno != errno.ENOENT:
1863 def record_download_archive(self, info_dict):
1864 fn = self.params.get('download_archive')
1867 vid_id = self._make_archive_id(info_dict)
1869 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1870 archive_file.write(vid_id + '\n')
1873 def format_resolution(format, default='unknown'):
1874 if format.get('vcodec') == 'none':
1876 if format.get('resolution') is not None:
1877 return format['resolution']
1878 if format.get('height') is not None:
1879 if format.get('width') is not None:
1880 res = '%sx%s' % (format['width'], format['height'])
1882 res = '%sp' % format['height']
1883 elif format.get('width') is not None:
1884 res = '%dx?' % format['width']
1889 def _format_note(self, fdict):
1891 if fdict.get('ext') in ['f4f', 'f4m']:
1892 res += '(unsupported) '
1893 if fdict.get('language'):
1896 res += '[%s] ' % fdict['language']
1897 if fdict.get('format_note') is not None:
1898 res += fdict['format_note'] + ' '
1899 if fdict.get('tbr') is not None:
1900 res += '%4dk ' % fdict['tbr']
1901 if fdict.get('container') is not None:
1904 res += '%s container' % fdict['container']
1905 if (fdict.get('vcodec') is not None and
1906 fdict.get('vcodec') != 'none'):
1909 res += fdict['vcodec']
1910 if fdict.get('vbr') is not None:
1912 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1914 if fdict.get('vbr') is not None:
1915 res += '%4dk' % fdict['vbr']
1916 if fdict.get('fps') is not None:
1919 res += '%sfps' % fdict['fps']
1920 if fdict.get('acodec') is not None:
1923 if fdict['acodec'] == 'none':
1926 res += '%-5s' % fdict['acodec']
1927 elif fdict.get('abr') is not None:
1931 if fdict.get('abr') is not None:
1932 res += '@%3dk' % fdict['abr']
1933 if fdict.get('asr') is not None:
1934 res += ' (%5dHz)' % fdict['asr']
1935 if fdict.get('filesize') is not None:
1938 res += format_bytes(fdict['filesize'])
1939 elif fdict.get('filesize_approx') is not None:
1942 res += '~' + format_bytes(fdict['filesize_approx'])
1945 def list_formats(self, info_dict):
1946 formats = info_dict.get('formats', [info_dict])
1948 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1950 if f.get('preference') is None or f['preference'] >= -1000]
1951 if len(formats) > 1:
1952 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1954 header_line = ['format code', 'extension', 'resolution', 'note']
1956 '[info] Available formats for %s:\n%s' %
1957 (info_dict['id'], render_table(header_line, table)))
1959 def list_thumbnails(self, info_dict):
1960 thumbnails = info_dict.get('thumbnails')
1962 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
1966 '[info] Thumbnails for %s:' % info_dict['id'])
1967 self.to_screen(render_table(
1968 ['ID', 'width', 'height', 'URL'],
1969 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1971 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1973 self.to_screen('%s has no %s' % (video_id, name))
1976 'Available %s for %s:' % (name, video_id))
1977 self.to_screen(render_table(
1978 ['Language', 'formats'],
1979 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1980 for lang, formats in subtitles.items()]))
1982 def urlopen(self, req):
1983 """ Start an HTTP download """
1984 if isinstance(req, compat_basestring):
1985 req = sanitized_Request(req)
1986 return self._opener.open(req, timeout=self._socket_timeout)
1988 def print_debug_header(self):
1989 if not self.params.get('verbose'):
1992 if type('') is not compat_str:
1993 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1994 self.report_warning(
1995 'Your Python is broken! Update to a newer and supported version')
1997 stdout_encoding = getattr(
1998 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2000 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2001 locale.getpreferredencoding(),
2002 sys.getfilesystemencoding(),
2004 self.get_encoding()))
2005 write_string(encoding_str, encoding=None)
2007 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
2009 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2011 sp = subprocess.Popen(
2012 ['git', 'rev-parse', '--short', 'HEAD'],
2013 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2014 cwd=os.path.dirname(os.path.abspath(__file__)))
2015 out, err = sp.communicate()
2016 out = out.decode().strip()
2017 if re.match('[0-9a-f]+', out):
2018 self._write_string('[debug] Git HEAD: ' + out + '\n')
2024 self._write_string('[debug] Python version %s - %s\n' % (
2025 platform.python_version(), platform_name()))
2027 exe_versions = FFmpegPostProcessor.get_versions(self)
2028 exe_versions['rtmpdump'] = rtmpdump_version()
2029 exe_str = ', '.join(
2031 for exe, v in sorted(exe_versions.items())
2036 self._write_string('[debug] exe versions: %s\n' % exe_str)
2039 for handler in self._opener.handlers:
2040 if hasattr(handler, 'proxies'):
2041 proxy_map.update(handler.proxies)
2042 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2044 if self.params.get('call_home', False):
2045 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2046 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2047 latest_version = self.urlopen(
2048 'https://yt-dl.org/latest/version').read().decode('utf-8')
2049 if version_tuple(latest_version) > version_tuple(__version__):
2050 self.report_warning(
2051 'You are using an outdated version (newest version: %s)! '
2052 'See https://yt-dl.org/update if you need help updating.' %
2055 def _setup_opener(self):
2056 timeout_val = self.params.get('socket_timeout')
2057 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2059 opts_cookiefile = self.params.get('cookiefile')
2060 opts_proxy = self.params.get('proxy')
2062 if opts_cookiefile is None:
2063 self.cookiejar = compat_cookiejar.CookieJar()
2065 opts_cookiefile = compat_expanduser(opts_cookiefile)
2066 self.cookiejar = compat_cookiejar.MozillaCookieJar(
2068 if os.access(opts_cookiefile, os.R_OK):
2069 self.cookiejar.load()
2071 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2072 if opts_proxy is not None:
2073 if opts_proxy == '':
2076 proxies = {'http': opts_proxy, 'https': opts_proxy}
2078 proxies = compat_urllib_request.getproxies()
2079 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2080 if 'http' in proxies and 'https' not in proxies:
2081 proxies['https'] = proxies['http']
2082 proxy_handler = PerRequestProxyHandler(proxies)
2084 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2085 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2086 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2087 data_handler = compat_urllib_request_DataHandler()
2089 # When passing our own FileHandler instance, build_opener won't add the
2090 # default FileHandler and allows us to disable the file protocol, which
2091 # can be used for malicious purposes (see
2092 # https://github.com/rg3/youtube-dl/issues/8227)
2093 file_handler = compat_urllib_request.FileHandler()
2095 def file_open(*args, **kwargs):
2096 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2097 file_handler.file_open = file_open
2099 opener = compat_urllib_request.build_opener(
2100 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2102 # Delete the default user-agent header, which would otherwise apply in
2103 # cases where our custom HTTP handler doesn't come into play
2104 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2105 opener.addheaders = []
2106 self._opener = opener
2108 def encode(self, s):
2109 if isinstance(s, bytes):
2110 return s # Already encoded
2113 return s.encode(self.get_encoding())
2114 except UnicodeEncodeError as err:
2115 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2118 def get_encoding(self):
2119 encoding = self.params.get('encoding')
2120 if encoding is None:
2121 encoding = preferredencoding()
2124 def _write_thumbnails(self, info_dict, filename):
2125 if self.params.get('writethumbnail', False):
2126 thumbnails = info_dict.get('thumbnails')
2128 thumbnails = [thumbnails[-1]]
2129 elif self.params.get('write_all_thumbnails', False):
2130 thumbnails = info_dict.get('thumbnails')
2135 # No thumbnails present, so return immediately
2138 for t in thumbnails:
2139 thumb_ext = determine_ext(t['url'], 'jpg')
2140 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2141 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2142 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2144 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2145 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2146 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2148 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2149 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2151 uf = self.urlopen(t['url'])
2152 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2153 shutil.copyfileobj(uf, thumbf)
2154 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2155 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2156 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2157 self.report_warning('Unable to download thumbnail "%s": %s' %
2158 (t['url'], error_to_compat_str(err)))