2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import, unicode_literals
31 compat_get_terminal_size,
36 compat_tokenize_tokenize,
38 compat_urllib_request,
39 compat_urllib_request_DataHandler,
60 PerRequestProxyHandler,
71 UnavailableVideoError,
76 YoutubeDLCookieProcessor,
83 from .cache import Cache
84 from .extractor import get_info_extractor, gen_extractors
85 from .downloader import get_suitable_downloader
86 from .downloader.rtmp import rtmpdump_version
87 from .postprocessor import (
90 FFmpegFixupStretchedPP,
95 from .version import __version__
97 if compat_os_name == 'nt':
101 class YoutubeDL(object):
104 YoutubeDL objects are the ones responsible of downloading the
105 actual video file and writing it to disk if the user has requested
106 it, among some other tasks. In most cases there should be one per
107 program. As, given a video URL, the downloader doesn't know how to
108 extract all the needed information, task that InfoExtractors do, it
109 has to pass the URL to one of them.
111 For this, YoutubeDL objects have a method that allows
112 InfoExtractors to be registered in a given order. When it is passed
113 a URL, the YoutubeDL object handles it to the first InfoExtractor it
114 finds that reports being able to handle it. The InfoExtractor extracts
115 all the information about the video or videos the URL refers to, and
116 YoutubeDL process the extracted information, possibly using a File
117 Downloader to download the video.
119 YoutubeDL objects accept a lot of parameters. In order not to saturate
120 the object constructor with arguments, it receives a dictionary of
121 options instead. These options are available through the params
122 attribute for the InfoExtractors to use. The YoutubeDL also
123 registers itself as the downloader in charge for the InfoExtractors
124 that are added to it, so this is a "mutual registration".
128 username: Username for authentication purposes.
129 password: Password for authentication purposes.
130 videopassword: Password for accessing a video.
131 usenetrc: Use netrc for authentication instead.
132 verbose: Print additional info to stdout.
133 quiet: Do not print messages to stdout.
134 no_warnings: Do not print out anything for warnings.
135 forceurl: Force printing final URL.
136 forcetitle: Force printing title.
137 forceid: Force printing ID.
138 forcethumbnail: Force printing thumbnail URL.
139 forcedescription: Force printing description.
140 forcefilename: Force printing final filename.
141 forceduration: Force printing duration.
142 forcejson: Force printing info_dict as JSON.
143 dump_single_json: Force printing the info_dict of the whole playlist
144 (or video) as a single JSON line.
145 simulate: Do not download the video files.
146 format: Video format code. See options.py for more information.
147 outtmpl: Template for output names.
148 restrictfilenames: Do not allow "&" and spaces in file names
149 ignoreerrors: Do not stop on download errors.
150 force_generic_extractor: Force downloader to use the generic extractor
151 nooverwrites: Prevent overwriting files.
152 playliststart: Playlist item to start at.
153 playlistend: Playlist item to end at.
154 playlist_items: Specific indices of playlist to download.
155 playlistreverse: Download playlist items in reverse order.
156 matchtitle: Download only matching titles.
157 rejecttitle: Reject downloads for matching titles.
158 logger: Log messages to a logging.Logger instance.
159 logtostderr: Log messages to stderr instead of stdout.
160 writedescription: Write the video description to a .description file
161 writeinfojson: Write the video description to a .info.json file
162 writeannotations: Write the video annotations to a .annotations.xml file
163 writethumbnail: Write the thumbnail image to a file
164 write_all_thumbnails: Write all thumbnail formats to files
165 writesubtitles: Write the video subtitles to a file
166 writeautomaticsub: Write the automatically generated subtitles to a file
167 allsubtitles: Downloads all the subtitles of the video
168 (requires writesubtitles or writeautomaticsub)
169 listsubtitles: Lists all available subtitles for the video
170 subtitlesformat: The format code for subtitles
171 subtitleslangs: List of languages of the subtitles to download
172 keepvideo: Keep the video file after post-processing
173 daterange: A DateRange object, download only if the upload_date is in the range.
174 skip_download: Skip the actual download of the video file
175 cachedir: Location of the cache files in the filesystem.
176 False to disable filesystem cache.
177 noplaylist: Download single video instead of a playlist if in doubt.
178 age_limit: An integer representing the user's age in years.
179 Unsuitable videos for the given age are skipped.
180 min_views: An integer representing the minimum view count the video
181 must have in order to not be skipped.
182 Videos without view count information are always
183 downloaded. None for no limit.
184 max_views: An integer representing the maximum view count.
185 Videos that are more popular than that are not
187 Videos without view count information are always
188 downloaded. None for no limit.
189 download_archive: File name of a file where all downloads are recorded.
190 Videos already present in the file are not downloaded
192 cookiefile: File name where cookies should be read from and dumped to.
193 nocheckcertificate:Do not verify SSL certificates
194 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
195 At the moment, this is only supported by YouTube.
196 proxy: URL of the proxy server to use
197 cn_verification_proxy: URL of the proxy to use for IP address verification
198 on Chinese sites. (Experimental)
199 socket_timeout: Time to wait for unresponsive hosts, in seconds
200 bidi_workaround: Work around buggy terminals without bidirectional text
201 support, using fridibi
202 debug_printtraffic:Print out sent and received HTTP traffic
203 include_ads: Download ads as well
204 default_search: Prepend this string if an input url is not valid.
205 'auto' for elaborate guessing
206 encoding: Use this encoding instead of the system-specified.
207 extract_flat: Do not resolve URLs, return the immediate result.
208 Pass in 'in_playlist' to only show this behavior for
210 postprocessors: A list of dictionaries, each with an entry
211 * key: The name of the postprocessor. See
212 youtube_dl/postprocessor/__init__.py for a list.
213 as well as any further keyword arguments for the
215 progress_hooks: A list of functions that get called on download
216 progress, with a dictionary with the entries
217 * status: One of "downloading", "error", or "finished".
218 Check this first and ignore unknown values.
220 If status is one of "downloading", or "finished", the
221 following properties may also be present:
222 * filename: The final filename (always present)
223 * tmpfilename: The filename we're currently writing to
224 * downloaded_bytes: Bytes on disk
225 * total_bytes: Size of the whole file, None if unknown
226 * total_bytes_estimate: Guess of the eventual file size,
228 * elapsed: The number of seconds since download started.
229 * eta: The estimated time in seconds, None if unknown
230 * speed: The download speed in bytes/second, None if
232 * fragment_index: The counter of the currently
233 downloaded video fragment.
234 * fragment_count: The number of fragments (= individual
235 files that will be merged)
237 Progress hooks are guaranteed to be called at least once
238 (with status "finished") if the download is successful.
239 merge_output_format: Extension to use when merging formats.
240 fixup: Automatically correct known faults of the file.
242 - "never": do nothing
243 - "warn": only emit a warning
244 - "detect_or_warn": check whether we can do anything
245 about it, warn otherwise (default)
246 source_address: (Experimental) Client-side IP address to bind to.
247 call_home: Boolean, true iff we are allowed to contact the
248 youtube-dl servers for debugging.
249 sleep_interval: Number of seconds to sleep before each download.
250 listformats: Print an overview of available video formats and exit.
251 list_thumbnails: Print a table of all thumbnails and exit.
252 match_filter: A function that gets called with the info_dict of
254 If it returns a message, the video is ignored.
255 If it returns None, the video is downloaded.
256 match_filter_func in utils.py is one example for this.
257 no_color: Do not emit color codes in output.
259 The following options determine which downloader is picked:
260 external_downloader: Executable of the external downloader to call.
261 None or unset for standard (built-in) downloader.
262 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv.
264 The following parameters are not used by YoutubeDL itself, they are used by
265 the downloader (see youtube_dl/downloader/common.py):
266 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
267 noresizebuffer, retries, continuedl, noprogress, consoletitle,
268 xattr_set_filesize, external_downloader_args, hls_use_mpegts.
270 The following options are used by the post processors:
271 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
272 otherwise prefer avconv.
273 postprocessor_args: A list of additional command-line arguments for the
280 _download_retcode = None
281 _num_downloads = None
284 def __init__(self, params=None, auto_init=True):
285 """Create a FileDownloader object with the given options."""
289 self._ies_instances = {}
291 self._progress_hooks = []
292 self._download_retcode = 0
293 self._num_downloads = 0
294 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
295 self._err_file = sys.stderr
298 'nocheckcertificate': False,
300 self.params.update(params)
301 self.cache = Cache(self)
303 if params.get('bidi_workaround', False):
306 master, slave = pty.openpty()
307 width = compat_get_terminal_size().columns
311 width_args = ['-w', str(width)]
313 stdin=subprocess.PIPE,
315 stderr=self._err_file)
317 self._output_process = subprocess.Popen(
318 ['bidiv'] + width_args, **sp_kwargs
321 self._output_process = subprocess.Popen(
322 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
323 self._output_channel = os.fdopen(master, 'rb')
324 except OSError as ose:
326 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 $PATH.')
330 if (sys.version_info >= (3,) and sys.platform != 'win32' and
331 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
332 not params.get('restrictfilenames', False)):
333 # On Python 3, the Unicode filesystem API will throw errors (#1474)
335 'Assuming --restrict-filenames since file system encoding '
336 'cannot encode all characters. '
337 'Set the LC_ALL environment variable to fix this.')
338 self.params['restrictfilenames'] = True
340 if isinstance(params.get('outtmpl'), bytes):
342 'Parameter outtmpl is bytes, but should be a unicode string. '
343 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
348 self.print_debug_header()
349 self.add_default_info_extractors()
351 for pp_def_raw in self.params.get('postprocessors', []):
352 pp_class = get_postprocessor(pp_def_raw['key'])
353 pp_def = dict(pp_def_raw)
355 pp = pp_class(self, **compat_kwargs(pp_def))
356 self.add_post_processor(pp)
358 for ph in self.params.get('progress_hooks', []):
359 self.add_progress_hook(ph)
361 def warn_if_short_id(self, argv):
362 # short YouTube ID starting with dash?
364 i for i, a in enumerate(argv)
365 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
369 [a for i, a in enumerate(argv) if i not in idxs] +
370 ['--'] + [argv[i] for i in idxs]
373 'Long argument string detected. '
374 'Use -- to separate parameters and URLs, like this:\n%s\n' %
375 args_to_str(correct_argv))
377 def add_info_extractor(self, ie):
378 """Add an InfoExtractor object to the end of the list."""
380 self._ies_instances[ie.ie_key()] = ie
381 ie.set_downloader(self)
383 def get_info_extractor(self, ie_key):
385 Get an instance of an IE with name ie_key, it will try to get one from
386 the _ies list, if there's no instance it will create a new one and add
387 it to the extractor list.
389 ie = self._ies_instances.get(ie_key)
391 ie = get_info_extractor(ie_key)()
392 self.add_info_extractor(ie)
395 def add_default_info_extractors(self):
397 Add the InfoExtractors returned by gen_extractors to the end of the list
399 for ie in gen_extractors():
400 self.add_info_extractor(ie)
402 def add_post_processor(self, pp):
403 """Add a PostProcessor object to the end of the chain."""
405 pp.set_downloader(self)
407 def add_progress_hook(self, ph):
408 """Add the progress hook (currently only for the file downloader)"""
409 self._progress_hooks.append(ph)
411 def _bidi_workaround(self, message):
412 if not hasattr(self, '_output_channel'):
415 assert hasattr(self, '_output_process')
416 assert isinstance(message, compat_str)
417 line_count = message.count('\n') + 1
418 self._output_process.stdin.write((message + '\n').encode('utf-8'))
419 self._output_process.stdin.flush()
420 res = ''.join(self._output_channel.readline().decode('utf-8')
421 for _ in range(line_count))
422 return res[:-len('\n')]
424 def to_screen(self, message, skip_eol=False):
425 """Print message to stdout if not in quiet mode."""
426 return self.to_stdout(message, skip_eol, check_quiet=True)
428 def _write_string(self, s, out=None):
429 write_string(s, out=out, encoding=self.params.get('encoding'))
431 def to_stdout(self, message, skip_eol=False, check_quiet=False):
432 """Print message to stdout if not in quiet mode."""
433 if self.params.get('logger'):
434 self.params['logger'].debug(message)
435 elif not check_quiet or not self.params.get('quiet', False):
436 message = self._bidi_workaround(message)
437 terminator = ['\n', ''][skip_eol]
438 output = message + terminator
440 self._write_string(output, self._screen_file)
442 def to_stderr(self, message):
443 """Print message to stderr."""
444 assert isinstance(message, compat_str)
445 if self.params.get('logger'):
446 self.params['logger'].error(message)
448 message = self._bidi_workaround(message)
449 output = message + '\n'
450 self._write_string(output, self._err_file)
452 def to_console_title(self, message):
453 if not self.params.get('consoletitle', False):
455 if compat_os_name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
456 # c_wchar_p() might not be necessary if `message` is
457 # already of type unicode()
458 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
459 elif 'TERM' in os.environ:
460 self._write_string('\033]0;%s\007' % message, self._screen_file)
462 def save_console_title(self):
463 if not self.params.get('consoletitle', False):
465 if 'TERM' in os.environ:
466 # Save the title on stack
467 self._write_string('\033[22;0t', self._screen_file)
469 def restore_console_title(self):
470 if not self.params.get('consoletitle', False):
472 if 'TERM' in os.environ:
473 # Restore the title from stack
474 self._write_string('\033[23;0t', self._screen_file)
477 self.save_console_title()
480 def __exit__(self, *args):
481 self.restore_console_title()
483 if self.params.get('cookiefile') is not None:
484 self.cookiejar.save()
486 def trouble(self, message=None, tb=None):
487 """Determine action to take when a download problem appears.
489 Depending on if the downloader has been configured to ignore
490 download errors or not, this method may throw an exception or
491 not when errors are found, after printing the message.
493 tb, if given, is additional traceback information.
495 if message is not None:
496 self.to_stderr(message)
497 if self.params.get('verbose'):
499 if sys.exc_info()[0]: # if .trouble has been called from an except block
501 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
502 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
503 tb += encode_compat_str(traceback.format_exc())
505 tb_data = traceback.format_list(traceback.extract_stack())
506 tb = ''.join(tb_data)
508 if not self.params.get('ignoreerrors', False):
509 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
510 exc_info = sys.exc_info()[1].exc_info
512 exc_info = sys.exc_info()
513 raise DownloadError(message, exc_info)
514 self._download_retcode = 1
516 def report_warning(self, message):
518 Print the message to stderr, it will be prefixed with 'WARNING:'
519 If stderr is a tty file the 'WARNING:' will be colored
521 if self.params.get('logger') is not None:
522 self.params['logger'].warning(message)
524 if self.params.get('no_warnings'):
526 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
527 _msg_header = '\033[0;33mWARNING:\033[0m'
529 _msg_header = 'WARNING:'
530 warning_message = '%s %s' % (_msg_header, message)
531 self.to_stderr(warning_message)
533 def report_error(self, message, tb=None):
535 Do the same as trouble, but prefixes the message with 'ERROR:', colored
536 in red if stderr is a tty file.
538 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
539 _msg_header = '\033[0;31mERROR:\033[0m'
541 _msg_header = 'ERROR:'
542 error_message = '%s %s' % (_msg_header, message)
543 self.trouble(error_message, tb)
545 def report_file_already_downloaded(self, file_name):
546 """Report file has already been fully downloaded."""
548 self.to_screen('[download] %s has already been downloaded' % file_name)
549 except UnicodeEncodeError:
550 self.to_screen('[download] The file has already been downloaded')
552 def prepare_filename(self, info_dict):
553 """Generate the output filename."""
555 template_dict = dict(info_dict)
557 template_dict['epoch'] = int(time.time())
558 autonumber_size = self.params.get('autonumber_size')
559 if autonumber_size is None:
561 autonumber_templ = '%0' + str(autonumber_size) + 'd'
562 template_dict['autonumber'] = autonumber_templ % self._num_downloads
563 if template_dict.get('playlist_index') is not None:
564 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
565 if template_dict.get('resolution') is None:
566 if template_dict.get('width') and template_dict.get('height'):
567 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
568 elif template_dict.get('height'):
569 template_dict['resolution'] = '%sp' % template_dict['height']
570 elif template_dict.get('width'):
571 template_dict['resolution'] = '%dx?' % template_dict['width']
573 sanitize = lambda k, v: sanitize_filename(
575 restricted=self.params.get('restrictfilenames'),
577 template_dict = dict((k, sanitize(k, v))
578 for k, v in template_dict.items()
580 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
582 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
583 tmpl = compat_expanduser(outtmpl)
584 filename = tmpl % template_dict
585 # Temporary fix for #4787
586 # 'Treat' all problem characters by passing filename through preferredencoding
587 # to workaround encoding issues with subprocess on python2 @ Windows
588 if sys.version_info < (3, 0) and sys.platform == 'win32':
589 filename = encodeFilename(filename, True).decode(preferredencoding())
590 return sanitize_path(filename)
591 except ValueError as err:
592 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
595 def _match_entry(self, info_dict, incomplete):
596 """ Returns None iff the file should be downloaded """
598 video_title = info_dict.get('title', info_dict.get('id', 'video'))
599 if 'title' in info_dict:
600 # This can happen when we're just evaluating the playlist
601 title = info_dict['title']
602 matchtitle = self.params.get('matchtitle', False)
604 if not re.search(matchtitle, title, re.IGNORECASE):
605 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
606 rejecttitle = self.params.get('rejecttitle', False)
608 if re.search(rejecttitle, title, re.IGNORECASE):
609 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
610 date = info_dict.get('upload_date')
612 dateRange = self.params.get('daterange', DateRange())
613 if date not in dateRange:
614 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
615 view_count = info_dict.get('view_count')
616 if view_count is not None:
617 min_views = self.params.get('min_views')
618 if min_views is not None and view_count < min_views:
619 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
620 max_views = self.params.get('max_views')
621 if max_views is not None and view_count > max_views:
622 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
623 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
624 return 'Skipping "%s" because it is age restricted' % video_title
625 if self.in_download_archive(info_dict):
626 return '%s has already been recorded in archive' % video_title
629 match_filter = self.params.get('match_filter')
630 if match_filter is not None:
631 ret = match_filter(info_dict)
638 def add_extra_info(info_dict, extra_info):
639 '''Set the keys from extra_info in info dict if they are missing'''
640 for key, value in extra_info.items():
641 info_dict.setdefault(key, value)
643 def extract_info(self, url, download=True, ie_key=None, extra_info={},
644 process=True, force_generic_extractor=False):
646 Returns a list with a dictionary for each video we find.
647 If 'download', also downloads the videos.
648 extra_info is a dict containing the extra values to add to each result
651 if not ie_key and force_generic_extractor:
655 ies = [self.get_info_extractor(ie_key)]
660 if not ie.suitable(url):
664 self.report_warning('The program functionality for this site has been marked as broken, '
665 'and will probably not work.')
668 ie_result = ie.extract(url)
669 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
671 if isinstance(ie_result, list):
672 # Backwards compatibility: old IE result format
674 '_type': 'compat_list',
675 'entries': ie_result,
677 self.add_default_extra_info(ie_result, ie, url)
679 return self.process_ie_result(ie_result, download, extra_info)
682 except ExtractorError as e: # An error we somewhat expected
683 self.report_error(compat_str(e), e.format_traceback())
685 except MaxDownloadsReached:
687 except Exception as e:
688 if self.params.get('ignoreerrors', False):
689 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
694 self.report_error('no suitable InfoExtractor for URL %s' % url)
696 def add_default_extra_info(self, ie_result, ie, url):
697 self.add_extra_info(ie_result, {
698 'extractor': ie.IE_NAME,
700 'webpage_url_basename': url_basename(url),
701 'extractor_key': ie.ie_key(),
704 def process_ie_result(self, ie_result, download=True, extra_info={}):
706 Take the result of the ie(may be modified) and resolve all unresolved
707 references (URLs, playlist items).
709 It will also download the videos if 'download'.
710 Returns the resolved ie_result.
712 result_type = ie_result.get('_type', 'video')
714 if result_type in ('url', 'url_transparent'):
715 extract_flat = self.params.get('extract_flat', False)
716 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
717 extract_flat is True):
718 if self.params.get('forcejson', False):
719 self.to_stdout(json.dumps(ie_result))
722 if result_type == 'video':
723 self.add_extra_info(ie_result, extra_info)
724 return self.process_video_result(ie_result, download=download)
725 elif result_type == 'url':
726 # We have to add extra_info to the results because it may be
727 # contained in a playlist
728 return self.extract_info(ie_result['url'],
730 ie_key=ie_result.get('ie_key'),
731 extra_info=extra_info)
732 elif result_type == 'url_transparent':
733 # Use the information from the embedding page
734 info = self.extract_info(
735 ie_result['url'], ie_key=ie_result.get('ie_key'),
736 extra_info=extra_info, download=False, process=False)
738 force_properties = dict(
739 (k, v) for k, v in ie_result.items() if v is not None)
740 for f in ('_type', 'url', 'ie_key'):
741 if f in force_properties:
742 del force_properties[f]
743 new_result = info.copy()
744 new_result.update(force_properties)
746 assert new_result.get('_type') != 'url_transparent'
748 return self.process_ie_result(
749 new_result, download=download, extra_info=extra_info)
750 elif result_type == 'playlist' or result_type == 'multi_video':
751 # We process each entry in the playlist
752 playlist = ie_result.get('title') or ie_result.get('id')
753 self.to_screen('[download] Downloading playlist: %s' % playlist)
755 playlist_results = []
757 playliststart = self.params.get('playliststart', 1) - 1
758 playlistend = self.params.get('playlistend')
759 # For backwards compatibility, interpret -1 as whole list
760 if playlistend == -1:
763 playlistitems_str = self.params.get('playlist_items')
765 if playlistitems_str is not None:
766 def iter_playlistitems(format):
767 for string_segment in format.split(','):
768 if '-' in string_segment:
769 start, end = string_segment.split('-')
770 for item in range(int(start), int(end) + 1):
773 yield int(string_segment)
774 playlistitems = iter_playlistitems(playlistitems_str)
776 ie_entries = ie_result['entries']
777 if isinstance(ie_entries, list):
778 n_all_entries = len(ie_entries)
781 ie_entries[i - 1] for i in playlistitems
782 if -n_all_entries <= i - 1 < n_all_entries]
784 entries = ie_entries[playliststart:playlistend]
785 n_entries = len(entries)
787 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
788 (ie_result['extractor'], playlist, n_all_entries, n_entries))
789 elif isinstance(ie_entries, PagedList):
792 for item in playlistitems:
793 entries.extend(ie_entries.getslice(
797 entries = ie_entries.getslice(
798 playliststart, playlistend)
799 n_entries = len(entries)
801 '[%s] playlist %s: Downloading %d videos' %
802 (ie_result['extractor'], playlist, n_entries))
805 entry_list = list(ie_entries)
806 entries = [entry_list[i - 1] for i in playlistitems]
808 entries = list(itertools.islice(
809 ie_entries, playliststart, playlistend))
810 n_entries = len(entries)
812 '[%s] playlist %s: Downloading %d videos' %
813 (ie_result['extractor'], playlist, n_entries))
815 if self.params.get('playlistreverse', False):
816 entries = entries[::-1]
818 for i, entry in enumerate(entries, 1):
819 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
821 'n_entries': n_entries,
822 'playlist': playlist,
823 'playlist_id': ie_result.get('id'),
824 'playlist_title': ie_result.get('title'),
825 'playlist_index': i + playliststart,
826 'extractor': ie_result['extractor'],
827 'webpage_url': ie_result['webpage_url'],
828 'webpage_url_basename': url_basename(ie_result['webpage_url']),
829 'extractor_key': ie_result['extractor_key'],
832 reason = self._match_entry(entry, incomplete=True)
833 if reason is not None:
834 self.to_screen('[download] ' + reason)
837 entry_result = self.process_ie_result(entry,
840 playlist_results.append(entry_result)
841 ie_result['entries'] = playlist_results
842 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
844 elif result_type == 'compat_list':
846 'Extractor %s returned a compat_list result. '
847 'It needs to be updated.' % ie_result.get('extractor'))
853 'extractor': ie_result['extractor'],
854 'webpage_url': ie_result['webpage_url'],
855 'webpage_url_basename': url_basename(ie_result['webpage_url']),
856 'extractor_key': ie_result['extractor_key'],
860 ie_result['entries'] = [
861 self.process_ie_result(_fixup(r), download, extra_info)
862 for r in ie_result['entries']
866 raise Exception('Invalid result type: %s' % result_type)
868 def _build_format_filter(self, filter_spec):
869 " Returns a function to filter the formats according to the filter_spec "
879 operator_rex = re.compile(r'''(?x)\s*
880 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
881 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
882 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
884 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
885 m = operator_rex.search(filter_spec)
888 comparison_value = int(m.group('value'))
890 comparison_value = parse_filesize(m.group('value'))
891 if comparison_value is None:
892 comparison_value = parse_filesize(m.group('value') + 'B')
893 if comparison_value is None:
895 'Invalid value %r in format specification %r' % (
896 m.group('value'), filter_spec))
897 op = OPERATORS[m.group('op')]
903 '^=': lambda attr, value: attr.startswith(value),
904 '$=': lambda attr, value: attr.endswith(value),
905 '*=': lambda attr, value: value in attr,
907 str_operator_rex = re.compile(r'''(?x)
908 \s*(?P<key>ext|acodec|vcodec|container|protocol)
909 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
910 \s*(?P<value>[a-zA-Z0-9._-]+)
912 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
913 m = str_operator_rex.search(filter_spec)
915 comparison_value = m.group('value')
916 op = STR_OPERATORS[m.group('op')]
919 raise ValueError('Invalid filter specification %r' % filter_spec)
922 actual_value = f.get(m.group('key'))
923 if actual_value is None:
924 return m.group('none_inclusive')
925 return op(actual_value, comparison_value)
928 def build_format_selector(self, format_spec):
929 def syntax_error(note, start):
931 'Invalid format specification: '
932 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
933 return SyntaxError(message)
935 PICKFIRST = 'PICKFIRST'
939 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
941 def _parse_filter(tokens):
943 for type, string, start, _, _ in tokens:
944 if type == tokenize.OP and string == ']':
945 return ''.join(filter_parts)
947 filter_parts.append(string)
949 def _remove_unused_ops(tokens):
950 # Remove operators that we don't use and join them with the surrounding strings
951 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
952 ALLOWED_OPS = ('/', '+', ',', '(', ')')
953 last_string, last_start, last_end, last_line = None, None, None, None
954 for type, string, start, end, line in tokens:
955 if type == tokenize.OP and string == '[':
957 yield tokenize.NAME, last_string, last_start, last_end, last_line
959 yield type, string, start, end, line
960 # everything inside brackets will be handled by _parse_filter
961 for type, string, start, end, line in tokens:
962 yield type, string, start, end, line
963 if type == tokenize.OP and string == ']':
965 elif type == tokenize.OP and string in ALLOWED_OPS:
967 yield tokenize.NAME, last_string, last_start, last_end, last_line
969 yield type, string, start, end, line
970 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
976 last_string += string
978 yield tokenize.NAME, last_string, last_start, last_end, last_line
980 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
982 current_selector = None
983 for type, string, start, _, _ in tokens:
984 # ENCODING is only defined in python 3.x
985 if type == getattr(tokenize, 'ENCODING', None):
987 elif type in [tokenize.NAME, tokenize.NUMBER]:
988 current_selector = FormatSelector(SINGLE, string, [])
989 elif type == tokenize.OP:
992 # ')' will be handled by the parentheses group
993 tokens.restore_last_token()
995 elif inside_merge and string in ['/', ',']:
996 tokens.restore_last_token()
998 elif inside_choice and string == ',':
999 tokens.restore_last_token()
1002 if not current_selector:
1003 raise syntax_error('"," must follow a format selector', start)
1004 selectors.append(current_selector)
1005 current_selector = None
1007 if not current_selector:
1008 raise syntax_error('"/" must follow a format selector', start)
1009 first_choice = current_selector
1010 second_choice = _parse_format_selection(tokens, inside_choice=True)
1011 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1013 if not current_selector:
1014 current_selector = FormatSelector(SINGLE, 'best', [])
1015 format_filter = _parse_filter(tokens)
1016 current_selector.filters.append(format_filter)
1018 if current_selector:
1019 raise syntax_error('Unexpected "("', start)
1020 group = _parse_format_selection(tokens, inside_group=True)
1021 current_selector = FormatSelector(GROUP, group, [])
1023 video_selector = current_selector
1024 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1025 if not video_selector or not audio_selector:
1026 raise syntax_error('"+" must be between two format selectors', start)
1027 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1029 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1030 elif type == tokenize.ENDMARKER:
1032 if current_selector:
1033 selectors.append(current_selector)
1036 def _build_selector_function(selector):
1037 if isinstance(selector, list):
1038 fs = [_build_selector_function(s) for s in selector]
1040 def selector_function(formats):
1042 for format in f(formats):
1044 return selector_function
1045 elif selector.type == GROUP:
1046 selector_function = _build_selector_function(selector.selector)
1047 elif selector.type == PICKFIRST:
1048 fs = [_build_selector_function(s) for s in selector.selector]
1050 def selector_function(formats):
1052 picked_formats = list(f(formats))
1054 return picked_formats
1056 elif selector.type == SINGLE:
1057 format_spec = selector.selector
1059 def selector_function(formats):
1060 formats = list(formats)
1063 if format_spec == 'all':
1066 elif format_spec in ['best', 'worst', None]:
1067 format_idx = 0 if format_spec == 'worst' else -1
1068 audiovideo_formats = [
1070 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1071 if audiovideo_formats:
1072 yield audiovideo_formats[format_idx]
1073 # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format
1074 elif (all(f.get('acodec') != 'none' for f in formats) or
1075 all(f.get('vcodec') != 'none' for f in formats)):
1076 yield formats[format_idx]
1077 elif format_spec == 'bestaudio':
1080 if f.get('vcodec') == 'none']
1082 yield audio_formats[-1]
1083 elif format_spec == 'worstaudio':
1086 if f.get('vcodec') == 'none']
1088 yield audio_formats[0]
1089 elif format_spec == 'bestvideo':
1092 if f.get('acodec') == 'none']
1094 yield video_formats[-1]
1095 elif format_spec == 'worstvideo':
1098 if f.get('acodec') == 'none']
1100 yield video_formats[0]
1102 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1103 if format_spec in extensions:
1104 filter_f = lambda f: f['ext'] == format_spec
1106 filter_f = lambda f: f['format_id'] == format_spec
1107 matches = list(filter(filter_f, formats))
1110 elif selector.type == MERGE:
1111 def _merge(formats_info):
1112 format_1, format_2 = [f['format_id'] for f in formats_info]
1113 # The first format must contain the video and the
1115 if formats_info[0].get('vcodec') == 'none':
1116 self.report_error('The first format must '
1117 'contain the video, try using '
1118 '"-f %s+%s"' % (format_2, format_1))
1120 # Formats must be opposite (video+audio)
1121 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1123 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1124 % (format_1, format_2))
1127 formats_info[0]['ext']
1128 if self.params.get('merge_output_format') is None
1129 else self.params['merge_output_format'])
1131 'requested_formats': formats_info,
1132 'format': '%s+%s' % (formats_info[0].get('format'),
1133 formats_info[1].get('format')),
1134 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1135 formats_info[1].get('format_id')),
1136 'width': formats_info[0].get('width'),
1137 'height': formats_info[0].get('height'),
1138 'resolution': formats_info[0].get('resolution'),
1139 'fps': formats_info[0].get('fps'),
1140 'vcodec': formats_info[0].get('vcodec'),
1141 'vbr': formats_info[0].get('vbr'),
1142 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1143 'acodec': formats_info[1].get('acodec'),
1144 'abr': formats_info[1].get('abr'),
1147 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1149 def selector_function(formats):
1150 formats = list(formats)
1151 for pair in itertools.product(video_selector(formats), audio_selector(formats)):
1154 filters = [self._build_format_filter(f) for f in selector.filters]
1156 def final_selector(formats):
1157 for _filter in filters:
1158 formats = list(filter(_filter, formats))
1159 return selector_function(formats)
1160 return final_selector
1162 stream = io.BytesIO(format_spec.encode('utf-8'))
1164 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1165 except tokenize.TokenError:
1166 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1168 class TokenIterator(object):
1169 def __init__(self, tokens):
1170 self.tokens = tokens
1177 if self.counter >= len(self.tokens):
1178 raise StopIteration()
1179 value = self.tokens[self.counter]
1185 def restore_last_token(self):
1188 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1189 return _build_selector_function(parsed_selector)
1191 def _calc_headers(self, info_dict):
1192 res = std_headers.copy()
1194 add_headers = info_dict.get('http_headers')
1196 res.update(add_headers)
1198 cookies = self._calc_cookies(info_dict)
1200 res['Cookie'] = cookies
1204 def _calc_cookies(self, info_dict):
1205 pr = sanitized_Request(info_dict['url'])
1206 self.cookiejar.add_cookie_header(pr)
1207 return pr.get_header('Cookie')
1209 def process_video_result(self, info_dict, download=True):
1210 assert info_dict.get('_type', 'video') == 'video'
1212 if 'id' not in info_dict:
1213 raise ExtractorError('Missing "id" field in extractor result')
1214 if 'title' not in info_dict:
1215 raise ExtractorError('Missing "title" field in extractor result')
1217 if 'playlist' not in info_dict:
1218 # It isn't part of a playlist
1219 info_dict['playlist'] = None
1220 info_dict['playlist_index'] = None
1222 thumbnails = info_dict.get('thumbnails')
1223 if thumbnails is None:
1224 thumbnail = info_dict.get('thumbnail')
1226 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1228 thumbnails.sort(key=lambda t: (
1229 t.get('preference'), t.get('width'), t.get('height'),
1230 t.get('id'), t.get('url')))
1231 for i, t in enumerate(thumbnails):
1232 if t.get('width') and t.get('height'):
1233 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1234 if t.get('id') is None:
1237 if thumbnails and 'thumbnail' not in info_dict:
1238 info_dict['thumbnail'] = thumbnails[-1]['url']
1240 if 'display_id' not in info_dict and 'id' in info_dict:
1241 info_dict['display_id'] = info_dict['id']
1243 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1244 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1245 # see http://bugs.python.org/issue1646728)
1247 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1248 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1249 except (ValueError, OverflowError, OSError):
1252 # Auto generate title fields corresponding to the *_number fields when missing
1253 # in order to always have clean titles. This is very common for TV series.
1254 for field in ('chapter', 'season', 'episode'):
1255 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1256 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1258 subtitles = info_dict.get('subtitles')
1260 for _, subtitle in subtitles.items():
1261 for subtitle_format in subtitle:
1262 if 'ext' not in subtitle_format:
1263 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1265 if self.params.get('listsubtitles', False):
1266 if 'automatic_captions' in info_dict:
1267 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1268 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1270 info_dict['requested_subtitles'] = self.process_subtitles(
1271 info_dict['id'], subtitles,
1272 info_dict.get('automatic_captions'))
1274 # We now pick which formats have to be downloaded
1275 if info_dict.get('formats') is None:
1276 # There's only one format available
1277 formats = [info_dict]
1279 formats = info_dict['formats']
1282 raise ExtractorError('No video formats found!')
1286 # We check that all the formats have the format and format_id fields
1287 for i, format in enumerate(formats):
1288 if 'url' not in format:
1289 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1291 if format.get('format_id') is None:
1292 format['format_id'] = compat_str(i)
1294 # Sanitize format_id from characters used in format selector expression
1295 format['format_id'] = re.sub('[\s,/+\[\]()]', '_', format['format_id'])
1296 format_id = format['format_id']
1297 if format_id not in formats_dict:
1298 formats_dict[format_id] = []
1299 formats_dict[format_id].append(format)
1301 # Make sure all formats have unique format_id
1302 for format_id, ambiguous_formats in formats_dict.items():
1303 if len(ambiguous_formats) > 1:
1304 for i, format in enumerate(ambiguous_formats):
1305 format['format_id'] = '%s-%d' % (format_id, i)
1307 for i, format in enumerate(formats):
1308 if format.get('format') is None:
1309 format['format'] = '{id} - {res}{note}'.format(
1310 id=format['format_id'],
1311 res=self.format_resolution(format),
1312 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1314 # Automatically determine file extension if missing
1315 if 'ext' not in format:
1316 format['ext'] = determine_ext(format['url']).lower()
1317 # Automatically determine protocol if missing (useful for format
1318 # selection purposes)
1319 if 'protocol' not in format:
1320 format['protocol'] = determine_protocol(format)
1321 # Add HTTP headers, so that external programs can use them from the
1323 full_format_info = info_dict.copy()
1324 full_format_info.update(format)
1325 format['http_headers'] = self._calc_headers(full_format_info)
1327 # TODO Central sorting goes here
1329 if formats[0] is not info_dict:
1330 # only set the 'formats' fields if the original info_dict list them
1331 # otherwise we end up with a circular reference, the first (and unique)
1332 # element in the 'formats' field in info_dict is info_dict itself,
1333 # which can't be exported to json
1334 info_dict['formats'] = formats
1335 if self.params.get('listformats'):
1336 self.list_formats(info_dict)
1338 if self.params.get('list_thumbnails'):
1339 self.list_thumbnails(info_dict)
1342 req_format = self.params.get('format')
1343 if req_format is None:
1344 req_format_list = []
1345 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1346 not info_dict.get('is_live')):
1347 merger = FFmpegMergerPP(self)
1348 if merger.available and merger.can_merge():
1349 req_format_list.append('bestvideo+bestaudio')
1350 req_format_list.append('best')
1351 req_format = '/'.join(req_format_list)
1352 format_selector = self.build_format_selector(req_format)
1353 formats_to_download = list(format_selector(formats))
1354 if not formats_to_download:
1355 raise ExtractorError('requested format not available',
1359 if len(formats_to_download) > 1:
1360 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1361 for format in formats_to_download:
1362 new_info = dict(info_dict)
1363 new_info.update(format)
1364 self.process_info(new_info)
1365 # We update the info dict with the best quality format (backwards compatibility)
1366 info_dict.update(formats_to_download[-1])
1369 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1370 """Select the requested subtitles and their format"""
1372 if normal_subtitles and self.params.get('writesubtitles'):
1373 available_subs.update(normal_subtitles)
1374 if automatic_captions and self.params.get('writeautomaticsub'):
1375 for lang, cap_info in automatic_captions.items():
1376 if lang not in available_subs:
1377 available_subs[lang] = cap_info
1379 if (not self.params.get('writesubtitles') and not
1380 self.params.get('writeautomaticsub') or not
1384 if self.params.get('allsubtitles', False):
1385 requested_langs = available_subs.keys()
1387 if self.params.get('subtitleslangs', False):
1388 requested_langs = self.params.get('subtitleslangs')
1389 elif 'en' in available_subs:
1390 requested_langs = ['en']
1392 requested_langs = [list(available_subs.keys())[0]]
1394 formats_query = self.params.get('subtitlesformat', 'best')
1395 formats_preference = formats_query.split('/') if formats_query else []
1397 for lang in requested_langs:
1398 formats = available_subs.get(lang)
1400 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1402 for ext in formats_preference:
1406 matches = list(filter(lambda f: f['ext'] == ext, formats))
1412 self.report_warning(
1413 'No subtitle format found matching "%s" for language %s, '
1414 'using %s' % (formats_query, lang, f['ext']))
1418 def process_info(self, info_dict):
1419 """Process a single resolved IE result."""
1421 assert info_dict.get('_type', 'video') == 'video'
1423 max_downloads = self.params.get('max_downloads')
1424 if max_downloads is not None:
1425 if self._num_downloads >= int(max_downloads):
1426 raise MaxDownloadsReached()
1428 info_dict['fulltitle'] = info_dict['title']
1429 if len(info_dict['title']) > 200:
1430 info_dict['title'] = info_dict['title'][:197] + '...'
1432 if 'format' not in info_dict:
1433 info_dict['format'] = info_dict['ext']
1435 reason = self._match_entry(info_dict, incomplete=False)
1436 if reason is not None:
1437 self.to_screen('[download] ' + reason)
1440 self._num_downloads += 1
1442 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1445 if self.params.get('forcetitle', False):
1446 self.to_stdout(info_dict['fulltitle'])
1447 if self.params.get('forceid', False):
1448 self.to_stdout(info_dict['id'])
1449 if self.params.get('forceurl', False):
1450 if info_dict.get('requested_formats') is not None:
1451 for f in info_dict['requested_formats']:
1452 self.to_stdout(f['url'] + f.get('play_path', ''))
1454 # For RTMP URLs, also include the playpath
1455 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1456 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1457 self.to_stdout(info_dict['thumbnail'])
1458 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1459 self.to_stdout(info_dict['description'])
1460 if self.params.get('forcefilename', False) and filename is not None:
1461 self.to_stdout(filename)
1462 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1463 self.to_stdout(formatSeconds(info_dict['duration']))
1464 if self.params.get('forceformat', False):
1465 self.to_stdout(info_dict['format'])
1466 if self.params.get('forcejson', False):
1467 self.to_stdout(json.dumps(info_dict))
1469 # Do nothing else if in simulate mode
1470 if self.params.get('simulate', False):
1473 if filename is None:
1477 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1478 if dn and not os.path.exists(dn):
1480 except (OSError, IOError) as err:
1481 self.report_error('unable to create directory ' + error_to_compat_str(err))
1484 if self.params.get('writedescription', False):
1485 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1486 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1487 self.to_screen('[info] Video description is already present')
1488 elif info_dict.get('description') is None:
1489 self.report_warning('There\'s no description to write.')
1492 self.to_screen('[info] Writing video description to: ' + descfn)
1493 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1494 descfile.write(info_dict['description'])
1495 except (OSError, IOError):
1496 self.report_error('Cannot write description file ' + descfn)
1499 if self.params.get('writeannotations', False):
1500 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1501 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1502 self.to_screen('[info] Video annotations are already present')
1505 self.to_screen('[info] Writing video annotations to: ' + annofn)
1506 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1507 annofile.write(info_dict['annotations'])
1508 except (KeyError, TypeError):
1509 self.report_warning('There are no annotations to write.')
1510 except (OSError, IOError):
1511 self.report_error('Cannot write annotations file: ' + annofn)
1514 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1515 self.params.get('writeautomaticsub')])
1517 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1518 # subtitles download errors are already managed as troubles in relevant IE
1519 # that way it will silently go on when used with unsupporting IE
1520 subtitles = info_dict['requested_subtitles']
1521 ie = self.get_info_extractor(info_dict['extractor_key'])
1522 for sub_lang, sub_info in subtitles.items():
1523 sub_format = sub_info['ext']
1524 if sub_info.get('data') is not None:
1525 sub_data = sub_info['data']
1528 sub_data = ie._download_webpage(
1529 sub_info['url'], info_dict['id'], note=False)
1530 except ExtractorError as err:
1531 self.report_warning('Unable to download subtitle for "%s": %s' %
1532 (sub_lang, error_to_compat_str(err.cause)))
1535 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1536 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1537 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1539 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1540 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1541 subfile.write(sub_data)
1542 except (OSError, IOError):
1543 self.report_error('Cannot write subtitles file ' + sub_filename)
1546 if self.params.get('writeinfojson', False):
1547 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1548 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1549 self.to_screen('[info] Video description metadata is already present')
1551 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1553 write_json_file(self.filter_requested_info(info_dict), infofn)
1554 except (OSError, IOError):
1555 self.report_error('Cannot write metadata to JSON file ' + infofn)
1558 self._write_thumbnails(info_dict, filename)
1560 if not self.params.get('skip_download', False):
1563 fd = get_suitable_downloader(info, self.params)(self, self.params)
1564 for ph in self._progress_hooks:
1565 fd.add_progress_hook(ph)
1566 if self.params.get('verbose'):
1567 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1568 return fd.download(name, info)
1570 if info_dict.get('requested_formats') is not None:
1573 merger = FFmpegMergerPP(self)
1574 if not merger.available:
1576 self.report_warning('You have requested multiple '
1577 'formats but ffmpeg or avconv are not installed.'
1578 ' The formats won\'t be merged.')
1580 postprocessors = [merger]
1582 def compatible_formats(formats):
1583 video, audio = formats
1585 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1586 if video_ext and audio_ext:
1588 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1591 for exts in COMPATIBLE_EXTS:
1592 if video_ext in exts and audio_ext in exts:
1594 # TODO: Check acodec/vcodec
1597 filename_real_ext = os.path.splitext(filename)[1][1:]
1599 os.path.splitext(filename)[0]
1600 if filename_real_ext == info_dict['ext']
1602 requested_formats = info_dict['requested_formats']
1603 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1604 info_dict['ext'] = 'mkv'
1605 self.report_warning(
1606 'Requested formats are incompatible for merge and will be merged into mkv.')
1607 # Ensure filename always has a correct extension for successful merge
1608 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1609 if os.path.exists(encodeFilename(filename)):
1611 '[download] %s has already been downloaded and '
1612 'merged' % filename)
1614 for f in requested_formats:
1615 new_info = dict(info_dict)
1617 fname = self.prepare_filename(new_info)
1618 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1619 downloaded.append(fname)
1620 partial_success = dl(fname, new_info)
1621 success = success and partial_success
1622 info_dict['__postprocessors'] = postprocessors
1623 info_dict['__files_to_merge'] = downloaded
1625 # Just a single file
1626 success = dl(filename, info_dict)
1627 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1628 self.report_error('unable to download video data: %s' % str(err))
1630 except (OSError, IOError) as err:
1631 raise UnavailableVideoError(err)
1632 except (ContentTooShortError, ) as err:
1633 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1636 if success and filename != '-':
1638 fixup_policy = self.params.get('fixup')
1639 if fixup_policy is None:
1640 fixup_policy = 'detect_or_warn'
1642 stretched_ratio = info_dict.get('stretched_ratio')
1643 if stretched_ratio is not None and stretched_ratio != 1:
1644 if fixup_policy == 'warn':
1645 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1646 info_dict['id'], stretched_ratio))
1647 elif fixup_policy == 'detect_or_warn':
1648 stretched_pp = FFmpegFixupStretchedPP(self)
1649 if stretched_pp.available:
1650 info_dict.setdefault('__postprocessors', [])
1651 info_dict['__postprocessors'].append(stretched_pp)
1653 self.report_warning(
1654 '%s: Non-uniform pixel ratio (%s). Install ffmpeg or avconv to fix this automatically.' % (
1655 info_dict['id'], stretched_ratio))
1657 assert fixup_policy in ('ignore', 'never')
1659 if info_dict.get('requested_formats') is None and info_dict.get('container') == 'm4a_dash':
1660 if fixup_policy == 'warn':
1661 self.report_warning('%s: writing DASH m4a. Only some players support this container.' % (
1663 elif fixup_policy == 'detect_or_warn':
1664 fixup_pp = FFmpegFixupM4aPP(self)
1665 if fixup_pp.available:
1666 info_dict.setdefault('__postprocessors', [])
1667 info_dict['__postprocessors'].append(fixup_pp)
1669 self.report_warning(
1670 '%s: writing DASH m4a. Only some players support this container. Install ffmpeg or avconv to fix this automatically.' % (
1673 assert fixup_policy in ('ignore', 'never')
1675 if info_dict.get('protocol') == 'm3u8_native' or info_dict.get('protocol') == 'm3u8' and self._downloader.params.get('hls_prefer_native', False):
1676 if fixup_policy == 'warn':
1677 self.report_warning('%s: malformated aac bitstream.' % (
1679 elif fixup_policy == 'detect_or_warn':
1680 fixup_pp = FFmpegFixupM3u8PP(self)
1681 if fixup_pp.available:
1682 info_dict.setdefault('__postprocessors', [])
1683 info_dict['__postprocessors'].append(fixup_pp)
1685 self.report_warning(
1686 '%s: malformated aac bitstream. Install ffmpeg or avconv to fix this automatically.' % (
1689 assert fixup_policy in ('ignore', 'never')
1692 self.post_process(filename, info_dict)
1693 except (PostProcessingError) as err:
1694 self.report_error('postprocessing: %s' % str(err))
1696 self.record_download_archive(info_dict)
1698 def download(self, url_list):
1699 """Download a given list of URLs."""
1700 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1701 if (len(url_list) > 1 and
1702 '%' not in outtmpl and
1703 self.params.get('max_downloads') != 1):
1704 raise SameFileError(outtmpl)
1706 for url in url_list:
1708 # It also downloads the videos
1709 res = self.extract_info(
1710 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1711 except UnavailableVideoError:
1712 self.report_error('unable to download video')
1713 except MaxDownloadsReached:
1714 self.to_screen('[info] Maximum number of downloaded files reached.')
1717 if self.params.get('dump_single_json', False):
1718 self.to_stdout(json.dumps(res))
1720 return self._download_retcode
1722 def download_with_info_file(self, info_filename):
1723 with contextlib.closing(fileinput.FileInput(
1724 [info_filename], mode='r',
1725 openhook=fileinput.hook_encoded('utf-8'))) as f:
1726 # FileInput doesn't have a read method, we can't call json.load
1727 info = self.filter_requested_info(json.loads('\n'.join(f)))
1729 self.process_ie_result(info, download=True)
1730 except DownloadError:
1731 webpage_url = info.get('webpage_url')
1732 if webpage_url is not None:
1733 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1734 return self.download([webpage_url])
1737 return self._download_retcode
1740 def filter_requested_info(info_dict):
1742 (k, v) for k, v in info_dict.items()
1743 if k not in ['requested_formats', 'requested_subtitles'])
1745 def post_process(self, filename, ie_info):
1746 """Run all the postprocessors on the given file."""
1747 info = dict(ie_info)
1748 info['filepath'] = filename
1750 if ie_info.get('__postprocessors') is not None:
1751 pps_chain.extend(ie_info['__postprocessors'])
1752 pps_chain.extend(self._pps)
1753 for pp in pps_chain:
1754 files_to_delete = []
1756 files_to_delete, info = pp.run(info)
1757 except PostProcessingError as e:
1758 self.report_error(e.msg)
1759 if files_to_delete and not self.params.get('keepvideo', False):
1760 for old_filename in files_to_delete:
1761 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1763 os.remove(encodeFilename(old_filename))
1764 except (IOError, OSError):
1765 self.report_warning('Unable to remove downloaded original file')
1767 def _make_archive_id(self, info_dict):
1768 # Future-proof against any change in case
1769 # and backwards compatibility with prior versions
1770 extractor = info_dict.get('extractor_key')
1771 if extractor is None:
1772 if 'id' in info_dict:
1773 extractor = info_dict.get('ie_key') # key in a playlist
1774 if extractor is None:
1775 return None # Incomplete video information
1776 return extractor.lower() + ' ' + info_dict['id']
1778 def in_download_archive(self, info_dict):
1779 fn = self.params.get('download_archive')
1783 vid_id = self._make_archive_id(info_dict)
1785 return False # Incomplete video information
1788 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1789 for line in archive_file:
1790 if line.strip() == vid_id:
1792 except IOError as ioe:
1793 if ioe.errno != errno.ENOENT:
1797 def record_download_archive(self, info_dict):
1798 fn = self.params.get('download_archive')
1801 vid_id = self._make_archive_id(info_dict)
1803 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1804 archive_file.write(vid_id + '\n')
1807 def format_resolution(format, default='unknown'):
1808 if format.get('vcodec') == 'none':
1810 if format.get('resolution') is not None:
1811 return format['resolution']
1812 if format.get('height') is not None:
1813 if format.get('width') is not None:
1814 res = '%sx%s' % (format['width'], format['height'])
1816 res = '%sp' % format['height']
1817 elif format.get('width') is not None:
1818 res = '%dx?' % format['width']
1823 def _format_note(self, fdict):
1825 if fdict.get('ext') in ['f4f', 'f4m']:
1826 res += '(unsupported) '
1827 if fdict.get('language'):
1830 res += '[%s]' % fdict['language']
1831 if fdict.get('format_note') is not None:
1832 res += fdict['format_note'] + ' '
1833 if fdict.get('tbr') is not None:
1834 res += '%4dk ' % fdict['tbr']
1835 if fdict.get('container') is not None:
1838 res += '%s container' % fdict['container']
1839 if (fdict.get('vcodec') is not None and
1840 fdict.get('vcodec') != 'none'):
1843 res += fdict['vcodec']
1844 if fdict.get('vbr') is not None:
1846 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1848 if fdict.get('vbr') is not None:
1849 res += '%4dk' % fdict['vbr']
1850 if fdict.get('fps') is not None:
1851 res += ', %sfps' % fdict['fps']
1852 if fdict.get('acodec') is not None:
1855 if fdict['acodec'] == 'none':
1858 res += '%-5s' % fdict['acodec']
1859 elif fdict.get('abr') is not None:
1863 if fdict.get('abr') is not None:
1864 res += '@%3dk' % fdict['abr']
1865 if fdict.get('asr') is not None:
1866 res += ' (%5dHz)' % fdict['asr']
1867 if fdict.get('filesize') is not None:
1870 res += format_bytes(fdict['filesize'])
1871 elif fdict.get('filesize_approx') is not None:
1874 res += '~' + format_bytes(fdict['filesize_approx'])
1877 def list_formats(self, info_dict):
1878 formats = info_dict.get('formats', [info_dict])
1880 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1882 if f.get('preference') is None or f['preference'] >= -1000]
1883 if len(formats) > 1:
1884 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1886 header_line = ['format code', 'extension', 'resolution', 'note']
1888 '[info] Available formats for %s:\n%s' %
1889 (info_dict['id'], render_table(header_line, table)))
1891 def list_thumbnails(self, info_dict):
1892 thumbnails = info_dict.get('thumbnails')
1894 tn_url = info_dict.get('thumbnail')
1896 thumbnails = [{'id': '0', 'url': tn_url}]
1899 '[info] No thumbnails present for %s' % info_dict['id'])
1903 '[info] Thumbnails for %s:' % info_dict['id'])
1904 self.to_screen(render_table(
1905 ['ID', 'width', 'height', 'URL'],
1906 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1908 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1910 self.to_screen('%s has no %s' % (video_id, name))
1913 'Available %s for %s:' % (name, video_id))
1914 self.to_screen(render_table(
1915 ['Language', 'formats'],
1916 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1917 for lang, formats in subtitles.items()]))
1919 def urlopen(self, req):
1920 """ Start an HTTP download """
1921 if isinstance(req, compat_basestring):
1922 req = sanitized_Request(req)
1923 return self._opener.open(req, timeout=self._socket_timeout)
1925 def print_debug_header(self):
1926 if not self.params.get('verbose'):
1929 if type('') is not compat_str:
1930 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1931 self.report_warning(
1932 'Your Python is broken! Update to a newer and supported version')
1934 stdout_encoding = getattr(
1935 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1937 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1938 locale.getpreferredencoding(),
1939 sys.getfilesystemencoding(),
1941 self.get_encoding()))
1942 write_string(encoding_str, encoding=None)
1944 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1946 sp = subprocess.Popen(
1947 ['git', 'rev-parse', '--short', 'HEAD'],
1948 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1949 cwd=os.path.dirname(os.path.abspath(__file__)))
1950 out, err = sp.communicate()
1951 out = out.decode().strip()
1952 if re.match('[0-9a-f]+', out):
1953 self._write_string('[debug] Git HEAD: ' + out + '\n')
1959 self._write_string('[debug] Python version %s - %s\n' % (
1960 platform.python_version(), platform_name()))
1962 exe_versions = FFmpegPostProcessor.get_versions(self)
1963 exe_versions['rtmpdump'] = rtmpdump_version()
1964 exe_str = ', '.join(
1966 for exe, v in sorted(exe_versions.items())
1971 self._write_string('[debug] exe versions: %s\n' % exe_str)
1974 for handler in self._opener.handlers:
1975 if hasattr(handler, 'proxies'):
1976 proxy_map.update(handler.proxies)
1977 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1979 if self.params.get('call_home', False):
1980 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
1981 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
1982 latest_version = self.urlopen(
1983 'https://yt-dl.org/latest/version').read().decode('utf-8')
1984 if version_tuple(latest_version) > version_tuple(__version__):
1985 self.report_warning(
1986 'You are using an outdated version (newest version: %s)! '
1987 'See https://yt-dl.org/update if you need help updating.' %
1990 def _setup_opener(self):
1991 timeout_val = self.params.get('socket_timeout')
1992 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
1994 opts_cookiefile = self.params.get('cookiefile')
1995 opts_proxy = self.params.get('proxy')
1997 if opts_cookiefile is None:
1998 self.cookiejar = compat_cookiejar.CookieJar()
2000 self.cookiejar = compat_cookiejar.MozillaCookieJar(
2002 if os.access(opts_cookiefile, os.R_OK):
2003 self.cookiejar.load()
2005 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2006 if opts_proxy is not None:
2007 if opts_proxy == '':
2010 proxies = {'http': opts_proxy, 'https': opts_proxy}
2012 proxies = compat_urllib_request.getproxies()
2013 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2014 if 'http' in proxies and 'https' not in proxies:
2015 proxies['https'] = proxies['http']
2016 proxy_handler = PerRequestProxyHandler(proxies)
2018 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2019 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2020 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2021 data_handler = compat_urllib_request_DataHandler()
2023 # When passing our own FileHandler instance, build_opener won't add the
2024 # default FileHandler and allows us to disable the file protocol, which
2025 # can be used for malicious purposes (see
2026 # https://github.com/rg3/youtube-dl/issues/8227)
2027 file_handler = compat_urllib_request.FileHandler()
2029 def file_open(*args, **kwargs):
2030 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2031 file_handler.file_open = file_open
2033 opener = compat_urllib_request.build_opener(
2034 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2036 # Delete the default user-agent header, which would otherwise apply in
2037 # cases where our custom HTTP handler doesn't come into play
2038 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2039 opener.addheaders = []
2040 self._opener = opener
2042 def encode(self, s):
2043 if isinstance(s, bytes):
2044 return s # Already encoded
2047 return s.encode(self.get_encoding())
2048 except UnicodeEncodeError as err:
2049 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2052 def get_encoding(self):
2053 encoding = self.params.get('encoding')
2054 if encoding is None:
2055 encoding = preferredencoding()
2058 def _write_thumbnails(self, info_dict, filename):
2059 if self.params.get('writethumbnail', False):
2060 thumbnails = info_dict.get('thumbnails')
2062 thumbnails = [thumbnails[-1]]
2063 elif self.params.get('write_all_thumbnails', False):
2064 thumbnails = info_dict.get('thumbnails')
2069 # No thumbnails present, so return immediately
2072 for t in thumbnails:
2073 thumb_ext = determine_ext(t['url'], 'jpg')
2074 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2075 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2076 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2078 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2079 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2080 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2082 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2083 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2085 uf = self.urlopen(t['url'])
2086 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2087 shutil.copyfileobj(uf, thumbf)
2088 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2089 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2090 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2091 self.report_warning('Unable to download thumbnail "%s": %s' %
2092 (t['url'], error_to_compat_str(err)))