4 from __future__ import unicode_literals
34 import xml.etree.ElementTree
38 compat_HTMLParseError,
42 compat_ctypes_WINFUNCTYPE,
43 compat_etree_fromstring,
46 compat_html_entities_html5,
52 compat_socket_create_connection,
58 compat_urllib_parse_urlencode,
59 compat_urllib_parse_urlparse,
60 compat_urllib_parse_unquote_plus,
61 compat_urllib_request,
72 def register_socks_protocols():
73 # "Register" SOCKS protocols
74 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
75 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
76 for scheme in ('socks', 'socks4', 'socks4a', 'socks5'):
77 if scheme not in compat_urlparse.uses_netloc:
78 compat_urlparse.uses_netloc.append(scheme)
81 # This is not clearly defined otherwise
82 compiled_regex_type = type(re.compile(''))
85 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:59.0) Gecko/20100101 Firefox/59.0 (Chrome)',
86 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
87 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
88 'Accept-Encoding': 'gzip, deflate',
89 'Accept-Language': 'en-us,en;q=0.5',
94 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
100 ENGLISH_MONTH_NAMES = [
101 'January', 'February', 'March', 'April', 'May', 'June',
102 'July', 'August', 'September', 'October', 'November', 'December']
105 'en': ENGLISH_MONTH_NAMES,
107 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
108 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
112 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
113 'flv', 'f4v', 'f4a', 'f4b',
114 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
115 'mkv', 'mka', 'mk3d',
124 'f4f', 'f4m', 'm3u8', 'smil')
126 # needed for sanitizing filenames in restricted mode
127 ACCENT_CHARS = dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
128 itertools.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUYP', ['ss'],
129 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuypy')))
152 '%Y-%m-%d %H:%M:%S.%f',
155 '%Y-%m-%dT%H:%M:%SZ',
156 '%Y-%m-%dT%H:%M:%S.%fZ',
157 '%Y-%m-%dT%H:%M:%S.%f0Z',
159 '%Y-%m-%dT%H:%M:%S.%f',
162 '%b %d %Y at %H:%M:%S',
164 '%B %d %Y at %H:%M:%S',
167 DATE_FORMATS_DAY_FIRST = list(DATE_FORMATS)
168 DATE_FORMATS_DAY_FIRST.extend([
177 DATE_FORMATS_MONTH_FIRST = list(DATE_FORMATS)
178 DATE_FORMATS_MONTH_FIRST.extend([
186 PACKED_CODES_RE = r"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
187 JSON_LD_RE = r'(?is)<script[^>]+type=(["\'])application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>'
190 def preferredencoding():
191 """Get preferred encoding.
193 Returns the best encoding scheme for the system, based on
194 locale.getpreferredencoding() and some further tweaks.
197 pref = locale.getpreferredencoding()
205 def write_json_file(obj, fn):
206 """ Encode obj as JSON and write it to fn, atomically if possible """
208 fn = encodeFilename(fn)
209 if sys.version_info < (3, 0) and sys.platform != 'win32':
210 encoding = get_filesystem_encoding()
211 # os.path.basename returns a bytes object, but NamedTemporaryFile
212 # will fail if the filename contains non ascii characters unless we
213 # use a unicode object
214 path_basename = lambda f: os.path.basename(fn).decode(encoding)
215 # the same for os.path.dirname
216 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
218 path_basename = os.path.basename
219 path_dirname = os.path.dirname
223 'prefix': path_basename(fn) + '.',
224 'dir': path_dirname(fn),
228 # In Python 2.x, json.dump expects a bytestream.
229 # In Python 3.x, it writes to a character stream
230 if sys.version_info < (3, 0):
238 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
243 if sys.platform == 'win32':
244 # Need to remove existing file on Windows, else os.rename raises
245 # WindowsError or FileExistsError.
250 os.rename(tf.name, fn)
259 if sys.version_info >= (2, 7):
260 def find_xpath_attr(node, xpath, key, val=None):
261 """ Find the xpath xpath[@key=val] """
262 assert re.match(r'^[a-zA-Z_-]+$', key)
263 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
264 return node.find(expr)
266 def find_xpath_attr(node, xpath, key, val=None):
267 for f in node.findall(compat_xpath(xpath)):
268 if key not in f.attrib:
270 if val is None or f.attrib.get(key) == val:
274 # On python2.6 the xml.etree.ElementTree.Element methods don't support
275 # the namespace parameter
278 def xpath_with_ns(path, ns_map):
279 components = [c.split(':') for c in path.split('/')]
283 replaced.append(c[0])
286 replaced.append('{%s}%s' % (ns_map[ns], tag))
287 return '/'.join(replaced)
290 def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
291 def _find_xpath(xpath):
292 return node.find(compat_xpath(xpath))
294 if isinstance(xpath, (str, compat_str)):
295 n = _find_xpath(xpath)
303 if default is not NO_DEFAULT:
306 name = xpath if name is None else name
307 raise ExtractorError('Could not find XML element %s' % name)
313 def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
314 n = xpath_element(node, xpath, name, fatal=fatal, default=default)
315 if n is None or n == default:
318 if default is not NO_DEFAULT:
321 name = xpath if name is None else name
322 raise ExtractorError('Could not find XML element\'s text %s' % name)
328 def xpath_attr(node, xpath, key, name=None, fatal=False, default=NO_DEFAULT):
329 n = find_xpath_attr(node, xpath, key)
331 if default is not NO_DEFAULT:
334 name = '%s[@%s]' % (xpath, key) if name is None else name
335 raise ExtractorError('Could not find XML attribute %s' % name)
341 def get_element_by_id(id, html):
342 """Return the content of the tag with the specified ID in the passed HTML document"""
343 return get_element_by_attribute('id', id, html)
346 def get_element_by_class(class_name, html):
347 """Return the content of the first tag with the specified class in the passed HTML document"""
348 retval = get_elements_by_class(class_name, html)
349 return retval[0] if retval else None
352 def get_element_by_attribute(attribute, value, html, escape_value=True):
353 retval = get_elements_by_attribute(attribute, value, html, escape_value)
354 return retval[0] if retval else None
357 def get_elements_by_class(class_name, html):
358 """Return the content of all tags with the specified class in the passed HTML document as a list"""
359 return get_elements_by_attribute(
360 'class', r'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
361 html, escape_value=False)
364 def get_elements_by_attribute(attribute, value, html, escape_value=True):
365 """Return the content of the tag with the specified attribute in the passed HTML document"""
367 value = re.escape(value) if escape_value else value
370 for m in re.finditer(r'''(?xs)
372 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?
374 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?
378 ''' % (re.escape(attribute), value), html):
379 res = m.group('content')
381 if res.startswith('"') or res.startswith("'"):
384 retlist.append(unescapeHTML(res))
389 class HTMLAttributeParser(compat_HTMLParser):
390 """Trivial HTML parser to gather the attributes for a single element"""
393 compat_HTMLParser.__init__(self)
395 def handle_starttag(self, tag, attrs):
396 self.attrs = dict(attrs)
399 def extract_attributes(html_element):
400 """Given a string for an HTML element such as
402 a="foo" B="bar" c="&98;az" d=boz
403 empty= noval entity="&"
406 Decode and return a dictionary of attributes.
408 'a': 'foo', 'b': 'bar', c: 'baz', d: 'boz',
409 'empty': '', 'noval': None, 'entity': '&',
410 'sq': '"', 'dq': '\''
412 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
413 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
415 parser = HTMLAttributeParser()
417 parser.feed(html_element)
419 # Older Python may throw HTMLParseError in case of malformed HTML
420 except compat_HTMLParseError:
425 def clean_html(html):
426 """Clean an HTML snippet into a readable string"""
428 if html is None: # Convenience for sanitizing descriptions etc.
432 html = html.replace('\n', ' ')
433 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
434 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
436 html = re.sub('<.*?>', '', html)
437 # Replace html entities
438 html = unescapeHTML(html)
442 def sanitize_open(filename, open_mode):
443 """Try to open the given filename, and slightly tweak it if this fails.
445 Attempts to open the given filename. If this fails, it tries to change
446 the filename slightly, step by step, until it's either able to open it
447 or it fails and raises a final exception, like the standard open()
450 It returns the tuple (stream, definitive_file_name).
454 if sys.platform == 'win32':
456 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
457 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
458 stream = open(encodeFilename(filename), open_mode)
459 return (stream, filename)
460 except (IOError, OSError) as err:
461 if err.errno in (errno.EACCES,):
464 # In case of error, try to remove win32 forbidden chars
465 alt_filename = sanitize_path(filename)
466 if alt_filename == filename:
469 # An exception here should be caught in the caller
470 stream = open(encodeFilename(alt_filename), open_mode)
471 return (stream, alt_filename)
474 def timeconvert(timestr):
475 """Convert RFC 2822 defined time string into system timestamp"""
477 timetuple = email.utils.parsedate_tz(timestr)
478 if timetuple is not None:
479 timestamp = email.utils.mktime_tz(timetuple)
483 def sanitize_filename(s, restricted=False, is_id=False):
484 """Sanitizes a string so it could be used as part of a filename.
485 If restricted is set, use a stricter subset of allowed characters.
486 Set is_id if this is not an arbitrary string, but an ID that should be kept
489 def replace_insane(char):
490 if restricted and char in ACCENT_CHARS:
491 return ACCENT_CHARS[char]
492 if char == '?' or ord(char) < 32 or ord(char) == 127:
495 return '' if restricted else '\''
497 return '_-' if restricted else ' -'
498 elif char in '\\/|*<>':
500 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
502 if restricted and ord(char) > 127:
507 s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
508 result = ''.join(map(replace_insane, s))
510 while '__' in result:
511 result = result.replace('__', '_')
512 result = result.strip('_')
513 # Common case of "Foreign band name - English song title"
514 if restricted and result.startswith('-_'):
516 if result.startswith('-'):
517 result = '_' + result[len('-'):]
518 result = result.lstrip('.')
524 def sanitize_path(s):
525 """Sanitizes and normalizes path on Windows"""
526 if sys.platform != 'win32':
528 drive_or_unc, _ = os.path.splitdrive(s)
529 if sys.version_info < (2, 7) and not drive_or_unc:
530 drive_or_unc, _ = os.path.splitunc(s)
531 norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
535 path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part)
536 for path_part in norm_path]
538 sanitized_path.insert(0, drive_or_unc + os.path.sep)
539 return os.path.join(*sanitized_path)
542 def sanitize_url(url):
543 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
544 # the number of unwanted failures due to missing protocol
545 if url.startswith('//'):
546 return 'http:%s' % url
547 # Fix some common typos seen so far
549 # https://github.com/rg3/youtube-dl/issues/15649
550 (r'^httpss://', r'https://'),
551 # https://bx1.be/lives/direct-tv/
552 (r'^rmtp([es]?)://', r'rtmp\1://'),
554 for mistake, fixup in COMMON_TYPOS:
555 if re.match(mistake, url):
556 return re.sub(mistake, fixup, url)
560 def sanitized_Request(url, *args, **kwargs):
561 return compat_urllib_request.Request(sanitize_url(url), *args, **kwargs)
565 """Expand shell variables and ~"""
566 return os.path.expandvars(compat_expanduser(s))
569 def orderedSet(iterable):
570 """ Remove all duplicates from the input iterable """
578 def _htmlentity_transform(entity_with_semicolon):
579 """Transforms an HTML entity to a character."""
580 entity = entity_with_semicolon[:-1]
582 # Known non-numeric HTML entity
583 if entity in compat_html_entities.name2codepoint:
584 return compat_chr(compat_html_entities.name2codepoint[entity])
586 # TODO: HTML5 allows entities without a semicolon. For example,
587 # 'Éric' should be decoded as 'Éric'.
588 if entity_with_semicolon in compat_html_entities_html5:
589 return compat_html_entities_html5[entity_with_semicolon]
591 mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
593 numstr = mobj.group(1)
594 if numstr.startswith('x'):
596 numstr = '0%s' % numstr
599 # See https://github.com/rg3/youtube-dl/issues/7518
601 return compat_chr(int(numstr, base))
605 # Unknown entity in name, return its literal representation
606 return '&%s;' % entity
612 assert type(s) == compat_str
615 r'&([^&;]+;)', lambda m: _htmlentity_transform(m.group(1)), s)
618 def get_subprocess_encoding():
619 if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
620 # For subprocess calls, encode with locale encoding
621 # Refer to http://stackoverflow.com/a/9951851/35070
622 encoding = preferredencoding()
624 encoding = sys.getfilesystemencoding()
630 def encodeFilename(s, for_subprocess=False):
632 @param s The name of the file
635 assert type(s) == compat_str
637 # Python 3 has a Unicode API
638 if sys.version_info >= (3, 0):
641 # Pass '' directly to use Unicode APIs on Windows 2000 and up
642 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
643 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
644 if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
647 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
648 if sys.platform.startswith('java'):
651 return s.encode(get_subprocess_encoding(), 'ignore')
654 def decodeFilename(b, for_subprocess=False):
656 if sys.version_info >= (3, 0):
659 if not isinstance(b, bytes):
662 return b.decode(get_subprocess_encoding(), 'ignore')
665 def encodeArgument(s):
666 if not isinstance(s, compat_str):
667 # Legacy code that uses byte strings
668 # Uncomment the following line after fixing all post processors
669 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
670 s = s.decode('ascii')
671 return encodeFilename(s, True)
674 def decodeArgument(b):
675 return decodeFilename(b, True)
678 def decodeOption(optval):
681 if isinstance(optval, bytes):
682 optval = optval.decode(preferredencoding())
684 assert isinstance(optval, compat_str)
688 def formatSeconds(secs):
690 return '%d:%02d:%02d' % (secs // 3600, (secs % 3600) // 60, secs % 60)
692 return '%d:%02d' % (secs // 60, secs % 60)
697 def make_HTTPS_handler(params, **kwargs):
698 opts_no_check_certificate = params.get('nocheckcertificate', False)
699 if hasattr(ssl, 'create_default_context'): # Python >= 3.4 or 2.7.9
700 context = ssl.create_default_context(ssl.Purpose.SERVER_AUTH)
701 if opts_no_check_certificate:
702 context.check_hostname = False
703 context.verify_mode = ssl.CERT_NONE
705 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
708 # (create_default_context present but HTTPSHandler has no context=)
711 if sys.version_info < (3, 2):
712 return YoutubeDLHTTPSHandler(params, **kwargs)
714 context = ssl.SSLContext(ssl.PROTOCOL_TLSv1)
715 context.verify_mode = (ssl.CERT_NONE
716 if opts_no_check_certificate
717 else ssl.CERT_REQUIRED)
718 context.set_default_verify_paths()
719 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
722 def bug_reports_message():
723 if ytdl_is_updateable():
724 update_cmd = 'type youtube-dl -U to update'
726 update_cmd = 'see https://yt-dl.org/update on how to update'
727 msg = '; please report this issue on https://yt-dl.org/bug .'
728 msg += ' Make sure you are using the latest version; %s.' % update_cmd
729 msg += ' Be sure to call youtube-dl with the --verbose flag and include its complete output.'
733 class YoutubeDLError(Exception):
734 """Base exception for YoutubeDL errors."""
738 class ExtractorError(YoutubeDLError):
739 """Error during info extraction."""
741 def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):
742 """ tb, if given, is the original traceback (so that it can be printed out).
743 If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
746 if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError):
748 if video_id is not None:
749 msg = video_id + ': ' + msg
751 msg += ' (caused by %r)' % cause
753 msg += bug_reports_message()
754 super(ExtractorError, self).__init__(msg)
757 self.exc_info = sys.exc_info() # preserve original exception
759 self.video_id = video_id
761 def format_traceback(self):
762 if self.traceback is None:
764 return ''.join(traceback.format_tb(self.traceback))
767 class UnsupportedError(ExtractorError):
768 def __init__(self, url):
769 super(UnsupportedError, self).__init__(
770 'Unsupported URL: %s' % url, expected=True)
774 class RegexNotFoundError(ExtractorError):
775 """Error when a regex didn't match"""
779 class GeoRestrictedError(ExtractorError):
780 """Geographic restriction Error exception.
782 This exception may be thrown when a video is not available from your
783 geographic location due to geographic restrictions imposed by a website.
785 def __init__(self, msg, countries=None):
786 super(GeoRestrictedError, self).__init__(msg, expected=True)
788 self.countries = countries
791 class DownloadError(YoutubeDLError):
792 """Download Error exception.
794 This exception may be thrown by FileDownloader objects if they are not
795 configured to continue on errors. They will contain the appropriate
799 def __init__(self, msg, exc_info=None):
800 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
801 super(DownloadError, self).__init__(msg)
802 self.exc_info = exc_info
805 class SameFileError(YoutubeDLError):
806 """Same File exception.
808 This exception will be thrown by FileDownloader objects if they detect
809 multiple files would have to be downloaded to the same file on disk.
814 class PostProcessingError(YoutubeDLError):
815 """Post Processing exception.
817 This exception may be raised by PostProcessor's .run() method to
818 indicate an error in the postprocessing task.
821 def __init__(self, msg):
822 super(PostProcessingError, self).__init__(msg)
826 class MaxDownloadsReached(YoutubeDLError):
827 """ --max-downloads limit has been reached. """
831 class UnavailableVideoError(YoutubeDLError):
832 """Unavailable Format exception.
834 This exception will be thrown when a video is requested
835 in a format that is not available for that video.
840 class ContentTooShortError(YoutubeDLError):
841 """Content Too Short exception.
843 This exception may be raised by FileDownloader objects when a file they
844 download is too small for what the server announced first, indicating
845 the connection was probably interrupted.
848 def __init__(self, downloaded, expected):
849 super(ContentTooShortError, self).__init__(
850 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded, expected)
853 self.downloaded = downloaded
854 self.expected = expected
857 class XAttrMetadataError(YoutubeDLError):
858 def __init__(self, code=None, msg='Unknown error'):
859 super(XAttrMetadataError, self).__init__(msg)
863 # Parsing code and msg
864 if (self.code in (errno.ENOSPC, errno.EDQUOT) or
865 'No space left' in self.msg or 'Disk quota excedded' in self.msg):
866 self.reason = 'NO_SPACE'
867 elif self.code == errno.E2BIG or 'Argument list too long' in self.msg:
868 self.reason = 'VALUE_TOO_LONG'
870 self.reason = 'NOT_SUPPORTED'
873 class XAttrUnavailableError(YoutubeDLError):
877 def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
878 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
879 # expected HTTP responses to meet HTTP/1.0 or later (see also
880 # https://github.com/rg3/youtube-dl/issues/6727)
881 if sys.version_info < (3, 0):
882 kwargs['strict'] = True
883 hc = http_class(*args, **compat_kwargs(kwargs))
884 source_address = ydl_handler._params.get('source_address')
885 if source_address is not None:
886 sa = (source_address, 0)
887 if hasattr(hc, 'source_address'): # Python 2.7+
888 hc.source_address = sa
890 def _hc_connect(self, *args, **kwargs):
891 sock = compat_socket_create_connection(
892 (self.host, self.port), self.timeout, sa)
894 self.sock = ssl.wrap_socket(
895 sock, self.key_file, self.cert_file,
896 ssl_version=ssl.PROTOCOL_TLSv1)
899 hc.connect = functools.partial(_hc_connect, hc)
904 def handle_youtubedl_headers(headers):
905 filtered_headers = headers
907 if 'Youtubedl-no-compression' in filtered_headers:
908 filtered_headers = dict((k, v) for k, v in filtered_headers.items() if k.lower() != 'accept-encoding')
909 del filtered_headers['Youtubedl-no-compression']
911 return filtered_headers
914 class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
915 """Handler for HTTP requests and responses.
917 This class, when installed with an OpenerDirector, automatically adds
918 the standard headers to every HTTP request and handles gzipped and
919 deflated responses from web servers. If compression is to be avoided in
920 a particular request, the original request in the program code only has
921 to include the HTTP header "Youtubedl-no-compression", which will be
922 removed before making the real request.
924 Part of this code was copied from:
926 http://techknack.net/python-urllib2-handlers/
928 Andrew Rowls, the author of that code, agreed to release it to the
932 def __init__(self, params, *args, **kwargs):
933 compat_urllib_request.HTTPHandler.__init__(self, *args, **kwargs)
934 self._params = params
936 def http_open(self, req):
937 conn_class = compat_http_client.HTTPConnection
939 socks_proxy = req.headers.get('Ytdl-socks-proxy')
941 conn_class = make_socks_conn_class(conn_class, socks_proxy)
942 del req.headers['Ytdl-socks-proxy']
944 return self.do_open(functools.partial(
945 _create_http_connection, self, conn_class, False),
951 return zlib.decompress(data, -zlib.MAX_WBITS)
953 return zlib.decompress(data)
955 def http_request(self, req):
956 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
957 # always respected by websites, some tend to give out URLs with non percent-encoded
958 # non-ASCII characters (see telemb.py, ard.py [#3412])
959 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
960 # To work around aforementioned issue we will replace request's original URL with
961 # percent-encoded one
962 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
963 # the code of this workaround has been moved here from YoutubeDL.urlopen()
964 url = req.get_full_url()
965 url_escaped = escape_url(url)
967 # Substitute URL if any change after escaping
968 if url != url_escaped:
969 req = update_Request(req, url=url_escaped)
971 for h, v in std_headers.items():
972 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
973 # The dict keys are capitalized because of this bug by urllib
974 if h.capitalize() not in req.headers:
977 req.headers = handle_youtubedl_headers(req.headers)
979 if sys.version_info < (2, 7) and '#' in req.get_full_url():
980 # Python 2.6 is brain-dead when it comes to fragments
981 req._Request__original = req._Request__original.partition('#')[0]
982 req._Request__r_type = req._Request__r_type.partition('#')[0]
986 def http_response(self, req, resp):
989 if resp.headers.get('Content-encoding', '') == 'gzip':
990 content = resp.read()
991 gz = gzip.GzipFile(fileobj=io.BytesIO(content), mode='rb')
993 uncompressed = io.BytesIO(gz.read())
994 except IOError as original_ioerror:
995 # There may be junk add the end of the file
996 # See http://stackoverflow.com/q/4928560/35070 for details
997 for i in range(1, 1024):
999 gz = gzip.GzipFile(fileobj=io.BytesIO(content[:-i]), mode='rb')
1000 uncompressed = io.BytesIO(gz.read())
1005 raise original_ioerror
1006 resp = compat_urllib_request.addinfourl(uncompressed, old_resp.headers, old_resp.url, old_resp.code)
1007 resp.msg = old_resp.msg
1008 del resp.headers['Content-encoding']
1010 if resp.headers.get('Content-encoding', '') == 'deflate':
1011 gz = io.BytesIO(self.deflate(resp.read()))
1012 resp = compat_urllib_request.addinfourl(gz, old_resp.headers, old_resp.url, old_resp.code)
1013 resp.msg = old_resp.msg
1014 del resp.headers['Content-encoding']
1015 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1016 # https://github.com/rg3/youtube-dl/issues/6457).
1017 if 300 <= resp.code < 400:
1018 location = resp.headers.get('Location')
1020 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1021 if sys.version_info >= (3, 0):
1022 location = location.encode('iso-8859-1').decode('utf-8')
1024 location = location.decode('utf-8')
1025 location_escaped = escape_url(location)
1026 if location != location_escaped:
1027 del resp.headers['Location']
1028 if sys.version_info < (3, 0):
1029 location_escaped = location_escaped.encode('utf-8')
1030 resp.headers['Location'] = location_escaped
1033 https_request = http_request
1034 https_response = http_response
1037 def make_socks_conn_class(base_class, socks_proxy):
1038 assert issubclass(base_class, (
1039 compat_http_client.HTTPConnection, compat_http_client.HTTPSConnection))
1041 url_components = compat_urlparse.urlparse(socks_proxy)
1042 if url_components.scheme.lower() == 'socks5':
1043 socks_type = ProxyType.SOCKS5
1044 elif url_components.scheme.lower() in ('socks', 'socks4'):
1045 socks_type = ProxyType.SOCKS4
1046 elif url_components.scheme.lower() == 'socks4a':
1047 socks_type = ProxyType.SOCKS4A
1049 def unquote_if_non_empty(s):
1052 return compat_urllib_parse_unquote_plus(s)
1056 url_components.hostname, url_components.port or 1080,
1058 unquote_if_non_empty(url_components.username),
1059 unquote_if_non_empty(url_components.password),
1062 class SocksConnection(base_class):
1064 self.sock = sockssocket()
1065 self.sock.setproxy(*proxy_args)
1066 if type(self.timeout) in (int, float):
1067 self.sock.settimeout(self.timeout)
1068 self.sock.connect((self.host, self.port))
1070 if isinstance(self, compat_http_client.HTTPSConnection):
1071 if hasattr(self, '_context'): # Python > 2.6
1072 self.sock = self._context.wrap_socket(
1073 self.sock, server_hostname=self.host)
1075 self.sock = ssl.wrap_socket(self.sock)
1077 return SocksConnection
1080 class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
1081 def __init__(self, params, https_conn_class=None, *args, **kwargs):
1082 compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)
1083 self._https_conn_class = https_conn_class or compat_http_client.HTTPSConnection
1084 self._params = params
1086 def https_open(self, req):
1088 conn_class = self._https_conn_class
1090 if hasattr(self, '_context'): # python > 2.6
1091 kwargs['context'] = self._context
1092 if hasattr(self, '_check_hostname'): # python 3.x
1093 kwargs['check_hostname'] = self._check_hostname
1095 socks_proxy = req.headers.get('Ytdl-socks-proxy')
1097 conn_class = make_socks_conn_class(conn_class, socks_proxy)
1098 del req.headers['Ytdl-socks-proxy']
1100 return self.do_open(functools.partial(
1101 _create_http_connection, self, conn_class, True),
1105 class YoutubeDLCookieProcessor(compat_urllib_request.HTTPCookieProcessor):
1106 def __init__(self, cookiejar=None):
1107 compat_urllib_request.HTTPCookieProcessor.__init__(self, cookiejar)
1109 def http_response(self, request, response):
1110 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1111 # characters in Set-Cookie HTTP header of last response (see
1112 # https://github.com/rg3/youtube-dl/issues/6769).
1113 # In order to at least prevent crashing we will percent encode Set-Cookie
1114 # header before HTTPCookieProcessor starts processing it.
1115 # if sys.version_info < (3, 0) and response.headers:
1116 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1117 # set_cookie = response.headers.get(set_cookie_header)
1119 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1120 # if set_cookie != set_cookie_escaped:
1121 # del response.headers[set_cookie_header]
1122 # response.headers[set_cookie_header] = set_cookie_escaped
1123 return compat_urllib_request.HTTPCookieProcessor.http_response(self, request, response)
1125 https_request = compat_urllib_request.HTTPCookieProcessor.http_request
1126 https_response = http_response
1129 def extract_timezone(date_str):
1131 r'^.{8,}?(?P<tz>Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
1134 timezone = datetime.timedelta()
1136 date_str = date_str[:-len(m.group('tz'))]
1137 if not m.group('sign'):
1138 timezone = datetime.timedelta()
1140 sign = 1 if m.group('sign') == '+' else -1
1141 timezone = datetime.timedelta(
1142 hours=sign * int(m.group('hours')),
1143 minutes=sign * int(m.group('minutes')))
1144 return timezone, date_str
1147 def parse_iso8601(date_str, delimiter='T', timezone=None):
1148 """ Return a UNIX timestamp from the given date """
1150 if date_str is None:
1153 date_str = re.sub(r'\.[0-9]+', '', date_str)
1155 if timezone is None:
1156 timezone, date_str = extract_timezone(date_str)
1159 date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
1160 dt = datetime.datetime.strptime(date_str, date_format) - timezone
1161 return calendar.timegm(dt.timetuple())
1166 def date_formats(day_first=True):
1167 return DATE_FORMATS_DAY_FIRST if day_first else DATE_FORMATS_MONTH_FIRST
1170 def unified_strdate(date_str, day_first=True):
1171 """Return a string with the date in the format YYYYMMDD"""
1173 if date_str is None:
1177 date_str = date_str.replace(',', ' ')
1178 # Remove AM/PM + timezone
1179 date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
1180 _, date_str = extract_timezone(date_str)
1182 for expression in date_formats(day_first):
1184 upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
1187 if upload_date is None:
1188 timetuple = email.utils.parsedate_tz(date_str)
1191 upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
1194 if upload_date is not None:
1195 return compat_str(upload_date)
1198 def unified_timestamp(date_str, day_first=True):
1199 if date_str is None:
1202 date_str = re.sub(r'[,|]', '', date_str)
1204 pm_delta = 12 if re.search(r'(?i)PM', date_str) else 0
1205 timezone, date_str = extract_timezone(date_str)
1207 # Remove AM/PM + timezone
1208 date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
1210 # Remove unrecognized timezones from ISO 8601 alike timestamps
1211 m = re.search(r'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str)
1213 date_str = date_str[:-len(m.group('tz'))]
1215 # Python only supports microseconds, so remove nanoseconds
1216 m = re.search(r'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str)
1218 date_str = m.group(1)
1220 for expression in date_formats(day_first):
1222 dt = datetime.datetime.strptime(date_str, expression) - timezone + datetime.timedelta(hours=pm_delta)
1223 return calendar.timegm(dt.timetuple())
1226 timetuple = email.utils.parsedate_tz(date_str)
1228 return calendar.timegm(timetuple) + pm_delta * 3600
1231 def determine_ext(url, default_ext='unknown_video'):
1232 if url is None or '.' not in url:
1234 guess = url.partition('?')[0].rpartition('.')[2]
1235 if re.match(r'^[A-Za-z0-9]+$', guess):
1237 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1238 elif guess.rstrip('/') in KNOWN_EXTENSIONS:
1239 return guess.rstrip('/')
1244 def subtitles_filename(filename, sub_lang, sub_format):
1245 return filename.rsplit('.', 1)[0] + '.' + sub_lang + '.' + sub_format
1248 def date_from_str(date_str):
1250 Return a datetime object from a string in the format YYYYMMDD or
1251 (now|today)[+-][0-9](day|week|month|year)(s)?"""
1252 today = datetime.date.today()
1253 if date_str in ('now', 'today'):
1255 if date_str == 'yesterday':
1256 return today - datetime.timedelta(days=1)
1257 match = re.match(r'(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
1258 if match is not None:
1259 sign = match.group('sign')
1260 time = int(match.group('time'))
1263 unit = match.group('unit')
1264 # A bad approximation?
1268 elif unit == 'year':
1272 delta = datetime.timedelta(**{unit: time})
1273 return today + delta
1274 return datetime.datetime.strptime(date_str, '%Y%m%d').date()
1277 def hyphenate_date(date_str):
1279 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1280 match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str)
1281 if match is not None:
1282 return '-'.join(match.groups())
1287 class DateRange(object):
1288 """Represents a time interval between two dates"""
1290 def __init__(self, start=None, end=None):
1291 """start and end must be strings in the format accepted by date"""
1292 if start is not None:
1293 self.start = date_from_str(start)
1295 self.start = datetime.datetime.min.date()
1297 self.end = date_from_str(end)
1299 self.end = datetime.datetime.max.date()
1300 if self.start > self.end:
1301 raise ValueError('Date range: "%s" , the start date must be before the end date' % self)
1305 """Returns a range that only contains the given day"""
1306 return cls(day, day)
1308 def __contains__(self, date):
1309 """Check if the date is in the range"""
1310 if not isinstance(date, datetime.date):
1311 date = date_from_str(date)
1312 return self.start <= date <= self.end
1315 return '%s - %s' % (self.start.isoformat(), self.end.isoformat())
1318 def platform_name():
1319 """ Returns the platform name as a compat_str """
1320 res = platform.platform()
1321 if isinstance(res, bytes):
1322 res = res.decode(preferredencoding())
1324 assert isinstance(res, compat_str)
1328 def _windows_write_string(s, out):
1329 """ Returns True if the string was written using special methods,
1330 False if it has yet to be written out."""
1331 # Adapted from http://stackoverflow.com/a/3259271/35070
1334 import ctypes.wintypes
1342 fileno = out.fileno()
1343 except AttributeError:
1344 # If the output stream doesn't have a fileno, it's virtual
1346 except io.UnsupportedOperation:
1347 # Some strange Windows pseudo files?
1349 if fileno not in WIN_OUTPUT_IDS:
1352 GetStdHandle = compat_ctypes_WINFUNCTYPE(
1353 ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
1354 ('GetStdHandle', ctypes.windll.kernel32))
1355 h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
1357 WriteConsoleW = compat_ctypes_WINFUNCTYPE(
1358 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
1359 ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
1360 ctypes.wintypes.LPVOID)(('WriteConsoleW', ctypes.windll.kernel32))
1361 written = ctypes.wintypes.DWORD(0)
1363 GetFileType = compat_ctypes_WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)(('GetFileType', ctypes.windll.kernel32))
1364 FILE_TYPE_CHAR = 0x0002
1365 FILE_TYPE_REMOTE = 0x8000
1366 GetConsoleMode = compat_ctypes_WINFUNCTYPE(
1367 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
1368 ctypes.POINTER(ctypes.wintypes.DWORD))(
1369 ('GetConsoleMode', ctypes.windll.kernel32))
1370 INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
1372 def not_a_console(handle):
1373 if handle == INVALID_HANDLE_VALUE or handle is None:
1375 return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR or
1376 GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0)
1378 if not_a_console(h):
1381 def next_nonbmp_pos(s):
1383 return next(i for i, c in enumerate(s) if ord(c) > 0xffff)
1384 except StopIteration:
1388 count = min(next_nonbmp_pos(s), 1024)
1390 ret = WriteConsoleW(
1391 h, s, count if count else 2, ctypes.byref(written), None)
1393 raise OSError('Failed to write string')
1394 if not count: # We just wrote a non-BMP character
1395 assert written.value == 2
1398 assert written.value > 0
1399 s = s[written.value:]
1403 def write_string(s, out=None, encoding=None):
1406 assert type(s) == compat_str
1408 if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'):
1409 if _windows_write_string(s, out):
1412 if ('b' in getattr(out, 'mode', '') or
1413 sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr
1414 byt = s.encode(encoding or preferredencoding(), 'ignore')
1416 elif hasattr(out, 'buffer'):
1417 enc = encoding or getattr(out, 'encoding', None) or preferredencoding()
1418 byt = s.encode(enc, 'ignore')
1419 out.buffer.write(byt)
1425 def bytes_to_intlist(bs):
1428 if isinstance(bs[0], int): # Python 3
1431 return [ord(c) for c in bs]
1434 def intlist_to_bytes(xs):
1437 return compat_struct_pack('%dB' % len(xs), *xs)
1440 # Cross-platform file locking
1441 if sys.platform == 'win32':
1442 import ctypes.wintypes
1445 class OVERLAPPED(ctypes.Structure):
1447 ('Internal', ctypes.wintypes.LPVOID),
1448 ('InternalHigh', ctypes.wintypes.LPVOID),
1449 ('Offset', ctypes.wintypes.DWORD),
1450 ('OffsetHigh', ctypes.wintypes.DWORD),
1451 ('hEvent', ctypes.wintypes.HANDLE),
1454 kernel32 = ctypes.windll.kernel32
1455 LockFileEx = kernel32.LockFileEx
1456 LockFileEx.argtypes = [
1457 ctypes.wintypes.HANDLE, # hFile
1458 ctypes.wintypes.DWORD, # dwFlags
1459 ctypes.wintypes.DWORD, # dwReserved
1460 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1461 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1462 ctypes.POINTER(OVERLAPPED) # Overlapped
1464 LockFileEx.restype = ctypes.wintypes.BOOL
1465 UnlockFileEx = kernel32.UnlockFileEx
1466 UnlockFileEx.argtypes = [
1467 ctypes.wintypes.HANDLE, # hFile
1468 ctypes.wintypes.DWORD, # dwReserved
1469 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1470 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1471 ctypes.POINTER(OVERLAPPED) # Overlapped
1473 UnlockFileEx.restype = ctypes.wintypes.BOOL
1474 whole_low = 0xffffffff
1475 whole_high = 0x7fffffff
1477 def _lock_file(f, exclusive):
1478 overlapped = OVERLAPPED()
1479 overlapped.Offset = 0
1480 overlapped.OffsetHigh = 0
1481 overlapped.hEvent = 0
1482 f._lock_file_overlapped_p = ctypes.pointer(overlapped)
1483 handle = msvcrt.get_osfhandle(f.fileno())
1484 if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0,
1485 whole_low, whole_high, f._lock_file_overlapped_p):
1486 raise OSError('Locking file failed: %r' % ctypes.FormatError())
1488 def _unlock_file(f):
1489 assert f._lock_file_overlapped_p
1490 handle = msvcrt.get_osfhandle(f.fileno())
1491 if not UnlockFileEx(handle, 0,
1492 whole_low, whole_high, f._lock_file_overlapped_p):
1493 raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
1496 # Some platforms, such as Jython, is missing fcntl
1500 def _lock_file(f, exclusive):
1501 fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
1503 def _unlock_file(f):
1504 fcntl.flock(f, fcntl.LOCK_UN)
1506 UNSUPPORTED_MSG = 'file locking is not supported on this platform'
1508 def _lock_file(f, exclusive):
1509 raise IOError(UNSUPPORTED_MSG)
1511 def _unlock_file(f):
1512 raise IOError(UNSUPPORTED_MSG)
1515 class locked_file(object):
1516 def __init__(self, filename, mode, encoding=None):
1517 assert mode in ['r', 'a', 'w']
1518 self.f = io.open(filename, mode, encoding=encoding)
1521 def __enter__(self):
1522 exclusive = self.mode != 'r'
1524 _lock_file(self.f, exclusive)
1530 def __exit__(self, etype, value, traceback):
1532 _unlock_file(self.f)
1539 def write(self, *args):
1540 return self.f.write(*args)
1542 def read(self, *args):
1543 return self.f.read(*args)
1546 def get_filesystem_encoding():
1547 encoding = sys.getfilesystemencoding()
1548 return encoding if encoding is not None else 'utf-8'
1551 def shell_quote(args):
1553 encoding = get_filesystem_encoding()
1555 if isinstance(a, bytes):
1556 # We may get a filename encoded with 'encodeFilename'
1557 a = a.decode(encoding)
1558 quoted_args.append(compat_shlex_quote(a))
1559 return ' '.join(quoted_args)
1562 def smuggle_url(url, data):
1563 """ Pass additional data in a URL for internal use. """
1565 url, idata = unsmuggle_url(url, {})
1567 sdata = compat_urllib_parse_urlencode(
1568 {'__youtubedl_smuggle': json.dumps(data)})
1569 return url + '#' + sdata
1572 def unsmuggle_url(smug_url, default=None):
1573 if '#__youtubedl_smuggle' not in smug_url:
1574 return smug_url, default
1575 url, _, sdata = smug_url.rpartition('#')
1576 jsond = compat_parse_qs(sdata)['__youtubedl_smuggle'][0]
1577 data = json.loads(jsond)
1581 def format_bytes(bytes):
1584 if type(bytes) is str:
1585 bytes = float(bytes)
1589 exponent = int(math.log(bytes, 1024.0))
1590 suffix = ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent]
1591 converted = float(bytes) / float(1024 ** exponent)
1592 return '%.2f%s' % (converted, suffix)
1595 def lookup_unit_table(unit_table, s):
1596 units_re = '|'.join(re.escape(u) for u in unit_table)
1598 r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re, s)
1601 num_str = m.group('num').replace(',', '.')
1602 mult = unit_table[m.group('unit')]
1603 return int(float(num_str) * mult)
1606 def parse_filesize(s):
1610 # The lower-case forms are of course incorrect and unofficial,
1611 # but we support those too
1628 'megabytes': 1000 ** 2,
1629 'mebibytes': 1024 ** 2,
1635 'gigabytes': 1000 ** 3,
1636 'gibibytes': 1024 ** 3,
1642 'terabytes': 1000 ** 4,
1643 'tebibytes': 1024 ** 4,
1649 'petabytes': 1000 ** 5,
1650 'pebibytes': 1024 ** 5,
1656 'exabytes': 1000 ** 6,
1657 'exbibytes': 1024 ** 6,
1663 'zettabytes': 1000 ** 7,
1664 'zebibytes': 1024 ** 7,
1670 'yottabytes': 1000 ** 8,
1671 'yobibytes': 1024 ** 8,
1674 return lookup_unit_table(_UNIT_TABLE, s)
1683 if re.match(r'^[\d,.]+$', s):
1684 return str_to_int(s)
1695 return lookup_unit_table(_UNIT_TABLE, s)
1698 def parse_resolution(s):
1702 mobj = re.search(r'\b(?P<w>\d+)\s*[xX×]\s*(?P<h>\d+)\b', s)
1705 'width': int(mobj.group('w')),
1706 'height': int(mobj.group('h')),
1709 mobj = re.search(r'\b(\d+)[pPiI]\b', s)
1711 return {'height': int(mobj.group(1))}
1713 mobj = re.search(r'\b([48])[kK]\b', s)
1715 return {'height': int(mobj.group(1)) * 540}
1720 def month_by_name(name, lang='en'):
1721 """ Return the number of a month by (locale-independently) English name """
1723 month_names = MONTH_NAMES.get(lang, MONTH_NAMES['en'])
1726 return month_names.index(name) + 1
1731 def month_by_abbreviation(abbrev):
1732 """ Return the number of a month by (locale-independently) English
1736 return [s[:3] for s in ENGLISH_MONTH_NAMES].index(abbrev) + 1
1741 def fix_xml_ampersands(xml_str):
1742 """Replace all the '&' by '&' in XML"""
1744 r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1749 def setproctitle(title):
1750 assert isinstance(title, compat_str)
1752 # ctypes in Jython is not complete
1753 # http://bugs.jython.org/issue2148
1754 if sys.platform.startswith('java'):
1758 libc = ctypes.cdll.LoadLibrary('libc.so.6')
1762 # LoadLibrary in Windows Python 2.7.13 only expects
1763 # a bytestring, but since unicode_literals turns
1764 # every string into a unicode string, it fails.
1766 title_bytes = title.encode('utf-8')
1767 buf = ctypes.create_string_buffer(len(title_bytes))
1768 buf.value = title_bytes
1770 libc.prctl(15, buf, 0, 0, 0)
1771 except AttributeError:
1772 return # Strange libc, just skip this
1775 def remove_start(s, start):
1776 return s[len(start):] if s is not None and s.startswith(start) else s
1779 def remove_end(s, end):
1780 return s[:-len(end)] if s is not None and s.endswith(end) else s
1783 def remove_quotes(s):
1784 if s is None or len(s) < 2:
1786 for quote in ('"', "'", ):
1787 if s[0] == quote and s[-1] == quote:
1792 def url_basename(url):
1793 path = compat_urlparse.urlparse(url).path
1794 return path.strip('/').split('/')[-1]
1798 return re.match(r'https?://[^?#&]+/', url).group()
1801 def urljoin(base, path):
1802 if isinstance(path, bytes):
1803 path = path.decode('utf-8')
1804 if not isinstance(path, compat_str) or not path:
1806 if re.match(r'^(?:https?:)?//', path):
1808 if isinstance(base, bytes):
1809 base = base.decode('utf-8')
1810 if not isinstance(base, compat_str) or not re.match(
1811 r'^(?:https?:)?//', base):
1813 return compat_urlparse.urljoin(base, path)
1816 class HEADRequest(compat_urllib_request.Request):
1817 def get_method(self):
1821 class PUTRequest(compat_urllib_request.Request):
1822 def get_method(self):
1826 def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
1829 v = getattr(v, get_attr, None)
1835 return int(v) * invscale // scale
1840 def str_or_none(v, default=None):
1841 return default if v is None else compat_str(v)
1844 def str_to_int(int_str):
1845 """ A more relaxed version of int_or_none """
1848 int_str = re.sub(r'[,\.\+]', '', int_str)
1852 def float_or_none(v, scale=1, invscale=1, default=None):
1856 return float(v) * invscale / scale
1861 def bool_or_none(v, default=None):
1862 return v if isinstance(v, bool) else default
1865 def strip_or_none(v):
1866 return None if v is None else v.strip()
1869 def parse_duration(s):
1870 if not isinstance(s, compat_basestring):
1875 days, hours, mins, secs, ms = [None] * 5
1876 m = re.match(r'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?Z?$', s)
1878 days, hours, mins, secs, ms = m.groups()
1883 [0-9]+\s*y(?:ears?)?\s*
1886 [0-9]+\s*m(?:onths?)?\s*
1889 [0-9]+\s*w(?:eeks?)?\s*
1892 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
1896 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
1899 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
1902 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
1905 days, hours, mins, secs, ms = m.groups()
1907 m = re.match(r'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s)
1909 hours, mins = m.groups()
1915 duration += float(secs)
1917 duration += float(mins) * 60
1919 duration += float(hours) * 60 * 60
1921 duration += float(days) * 24 * 60 * 60
1923 duration += float(ms)
1927 def prepend_extension(filename, ext, expected_real_ext=None):
1928 name, real_ext = os.path.splitext(filename)
1930 '{0}.{1}{2}'.format(name, ext, real_ext)
1931 if not expected_real_ext or real_ext[1:] == expected_real_ext
1932 else '{0}.{1}'.format(filename, ext))
1935 def replace_extension(filename, ext, expected_real_ext=None):
1936 name, real_ext = os.path.splitext(filename)
1937 return '{0}.{1}'.format(
1938 name if not expected_real_ext or real_ext[1:] == expected_real_ext else filename,
1942 def check_executable(exe, args=[]):
1943 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
1944 args can be a list of arguments for a short output (like -version) """
1946 subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
1952 def get_exe_version(exe, args=['--version'],
1953 version_re=None, unrecognized='present'):
1954 """ Returns the version of the specified executable,
1955 or False if the executable is not present """
1957 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
1958 # SIGTTOU if youtube-dl is run in the background.
1959 # See https://github.com/rg3/youtube-dl/issues/955#issuecomment-209789656
1960 out, _ = subprocess.Popen(
1961 [encodeArgument(exe)] + args,
1962 stdin=subprocess.PIPE,
1963 stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
1966 if isinstance(out, bytes): # Python 2.x
1967 out = out.decode('ascii', 'ignore')
1968 return detect_exe_version(out, version_re, unrecognized)
1971 def detect_exe_version(output, version_re=None, unrecognized='present'):
1972 assert isinstance(output, compat_str)
1973 if version_re is None:
1974 version_re = r'version\s+([-0-9._a-zA-Z]+)'
1975 m = re.search(version_re, output)
1982 class PagedList(object):
1984 # This is only useful for tests
1985 return len(self.getslice())
1988 class OnDemandPagedList(PagedList):
1989 def __init__(self, pagefunc, pagesize, use_cache=True):
1990 self._pagefunc = pagefunc
1991 self._pagesize = pagesize
1992 self._use_cache = use_cache
1996 def getslice(self, start=0, end=None):
1998 for pagenum in itertools.count(start // self._pagesize):
1999 firstid = pagenum * self._pagesize
2000 nextfirstid = pagenum * self._pagesize + self._pagesize
2001 if start >= nextfirstid:
2006 page_results = self._cache.get(pagenum)
2007 if page_results is None:
2008 page_results = list(self._pagefunc(pagenum))
2010 self._cache[pagenum] = page_results
2013 start % self._pagesize
2014 if firstid <= start < nextfirstid
2018 ((end - 1) % self._pagesize) + 1
2019 if (end is not None and firstid <= end <= nextfirstid)
2022 if startv != 0 or endv is not None:
2023 page_results = page_results[startv:endv]
2024 res.extend(page_results)
2026 # A little optimization - if current page is not "full", ie. does
2027 # not contain page_size videos then we can assume that this page
2028 # is the last one - there are no more ids on further pages -
2029 # i.e. no need to query again.
2030 if len(page_results) + startv < self._pagesize:
2033 # If we got the whole page, but the next page is not interesting,
2034 # break out early as well
2035 if end == nextfirstid:
2040 class InAdvancePagedList(PagedList):
2041 def __init__(self, pagefunc, pagecount, pagesize):
2042 self._pagefunc = pagefunc
2043 self._pagecount = pagecount
2044 self._pagesize = pagesize
2046 def getslice(self, start=0, end=None):
2048 start_page = start // self._pagesize
2050 self._pagecount if end is None else (end // self._pagesize + 1))
2051 skip_elems = start - start_page * self._pagesize
2052 only_more = None if end is None else end - start
2053 for pagenum in range(start_page, end_page):
2054 page = list(self._pagefunc(pagenum))
2056 page = page[skip_elems:]
2058 if only_more is not None:
2059 if len(page) < only_more:
2060 only_more -= len(page)
2062 page = page[:only_more]
2069 def uppercase_escape(s):
2070 unicode_escape = codecs.getdecoder('unicode_escape')
2072 r'\\U[0-9a-fA-F]{8}',
2073 lambda m: unicode_escape(m.group(0))[0],
2077 def lowercase_escape(s):
2078 unicode_escape = codecs.getdecoder('unicode_escape')
2080 r'\\u[0-9a-fA-F]{4}',
2081 lambda m: unicode_escape(m.group(0))[0],
2085 def escape_rfc3986(s):
2086 """Escape non-ASCII characters as suggested by RFC 3986"""
2087 if sys.version_info < (3, 0) and isinstance(s, compat_str):
2088 s = s.encode('utf-8')
2089 return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]")
2092 def escape_url(url):
2093 """Escape URL as suggested by RFC 3986"""
2094 url_parsed = compat_urllib_parse_urlparse(url)
2095 return url_parsed._replace(
2096 netloc=url_parsed.netloc.encode('idna').decode('ascii'),
2097 path=escape_rfc3986(url_parsed.path),
2098 params=escape_rfc3986(url_parsed.params),
2099 query=escape_rfc3986(url_parsed.query),
2100 fragment=escape_rfc3986(url_parsed.fragment)
2104 def read_batch_urls(batch_fd):
2106 if not isinstance(url, compat_str):
2107 url = url.decode('utf-8', 'replace')
2108 BOM_UTF8 = '\xef\xbb\xbf'
2109 if url.startswith(BOM_UTF8):
2110 url = url[len(BOM_UTF8):]
2112 if url.startswith(('#', ';', ']')):
2116 with contextlib.closing(batch_fd) as fd:
2117 return [url for url in map(fixup, fd) if url]
2120 def urlencode_postdata(*args, **kargs):
2121 return compat_urllib_parse_urlencode(*args, **kargs).encode('ascii')
2124 def update_url_query(url, query):
2127 parsed_url = compat_urlparse.urlparse(url)
2128 qs = compat_parse_qs(parsed_url.query)
2130 return compat_urlparse.urlunparse(parsed_url._replace(
2131 query=compat_urllib_parse_urlencode(qs, True)))
2134 def update_Request(req, url=None, data=None, headers={}, query={}):
2135 req_headers = req.headers.copy()
2136 req_headers.update(headers)
2137 req_data = data or req.data
2138 req_url = update_url_query(url or req.get_full_url(), query)
2139 req_get_method = req.get_method()
2140 if req_get_method == 'HEAD':
2141 req_type = HEADRequest
2142 elif req_get_method == 'PUT':
2143 req_type = PUTRequest
2145 req_type = compat_urllib_request.Request
2147 req_url, data=req_data, headers=req_headers,
2148 origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
2149 if hasattr(req, 'timeout'):
2150 new_req.timeout = req.timeout
2154 def _multipart_encode_impl(data, boundary):
2155 content_type = 'multipart/form-data; boundary=%s' % boundary
2158 for k, v in data.items():
2159 out += b'--' + boundary.encode('ascii') + b'\r\n'
2160 if isinstance(k, compat_str):
2161 k = k.encode('utf-8')
2162 if isinstance(v, compat_str):
2163 v = v.encode('utf-8')
2164 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2165 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2166 content = b'Content-Disposition: form-data; name="' + k + b'"\r\n\r\n' + v + b'\r\n'
2167 if boundary.encode('ascii') in content:
2168 raise ValueError('Boundary overlaps with data')
2171 out += b'--' + boundary.encode('ascii') + b'--\r\n'
2173 return out, content_type
2176 def multipart_encode(data, boundary=None):
2178 Encode a dict to RFC 7578-compliant form-data
2181 A dict where keys and values can be either Unicode or bytes-like
2184 If specified a Unicode object, it's used as the boundary. Otherwise
2185 a random boundary is generated.
2187 Reference: https://tools.ietf.org/html/rfc7578
2189 has_specified_boundary = boundary is not None
2192 if boundary is None:
2193 boundary = '---------------' + str(random.randrange(0x0fffffff, 0xffffffff))
2196 out, content_type = _multipart_encode_impl(data, boundary)
2199 if has_specified_boundary:
2203 return out, content_type
2206 def dict_get(d, key_or_keys, default=None, skip_false_values=True):
2207 if isinstance(key_or_keys, (list, tuple)):
2208 for key in key_or_keys:
2209 if key not in d or d[key] is None or skip_false_values and not d[key]:
2213 return d.get(key_or_keys, default)
2216 def try_get(src, getter, expected_type=None):
2217 if not isinstance(getter, (list, tuple)):
2222 except (AttributeError, KeyError, TypeError, IndexError):
2225 if expected_type is None or isinstance(v, expected_type):
2229 def merge_dicts(*dicts):
2231 for a_dict in dicts:
2232 for k, v in a_dict.items():
2235 if (k not in merged or
2236 (isinstance(v, compat_str) and v and
2237 isinstance(merged[k], compat_str) and
2243 def encode_compat_str(string, encoding=preferredencoding(), errors='strict'):
2244 return string if isinstance(string, compat_str) else compat_str(string, encoding, errors)
2256 TV_PARENTAL_GUIDELINES = {
2266 def parse_age_limit(s):
2268 return s if 0 <= s <= 21 else None
2269 if not isinstance(s, compat_basestring):
2271 m = re.match(r'^(?P<age>\d{1,2})\+?$', s)
2273 return int(m.group('age'))
2275 return US_RATINGS[s]
2276 m = re.match(r'^TV[_-]?(%s)$' % '|'.join(k[3:] for k in TV_PARENTAL_GUIDELINES), s)
2278 return TV_PARENTAL_GUIDELINES['TV-' + m.group(1)]
2282 def strip_jsonp(code):
2285 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]+)
2286 (?:\s*&&\s*(?P=func_name))?
2287 \s*\(\s*(?P<callback_data>.*)\);?
2288 \s*?(?://[^\n]*)*$''',
2289 r'\g<callback_data>', code)
2292 def js_to_json(code):
2293 COMMENT_RE = r'/\*(?:(?!\*/).)*?\*/|//[^\n]*'
2294 SKIP_RE = r'\s*(?:{comment})?\s*'.format(comment=COMMENT_RE)
2296 (r'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip=SKIP_RE), 16),
2297 (r'(?s)^(0+[0-7]+){skip}:?$'.format(skip=SKIP_RE), 8),
2302 if v in ('true', 'false', 'null'):
2304 elif v.startswith('/*') or v.startswith('//') or v == ',':
2307 if v[0] in ("'", '"'):
2308 v = re.sub(r'(?s)\\.|"', lambda m: {
2313 }.get(m.group(0), m.group(0)), v[1:-1])
2315 for regex, base in INTEGER_TABLE:
2316 im = re.match(regex, v)
2318 i = int(im.group(1), base)
2319 return '"%d":' % i if v.endswith(':') else '%d' % i
2323 return re.sub(r'''(?sx)
2324 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
2325 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
2326 {comment}|,(?={skip}[\]}}])|
2327 (?:(?<![0-9])[eE]|[a-df-zA-DF-Z_])[.a-zA-Z_0-9]*|
2328 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
2330 '''.format(comment=COMMENT_RE, skip=SKIP_RE), fix_kv, code)
2333 def qualities(quality_ids):
2334 """ Get a numeric quality value out of a list of possible values """
2337 return quality_ids.index(qid)
2343 DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s'
2346 def limit_length(s, length):
2347 """ Add ellipses to overly long strings """
2352 return s[:length - len(ELLIPSES)] + ELLIPSES
2356 def version_tuple(v):
2357 return tuple(int(e) for e in re.split(r'[-.]', v))
2360 def is_outdated_version(version, limit, assume_new=True):
2362 return not assume_new
2364 return version_tuple(version) < version_tuple(limit)
2366 return not assume_new
2369 def ytdl_is_updateable():
2370 """ Returns if youtube-dl can be updated with -U """
2371 from zipimport import zipimporter
2373 return isinstance(globals().get('__loader__'), zipimporter) or hasattr(sys, 'frozen')
2376 def args_to_str(args):
2377 # Get a short string representation for a subprocess command
2378 return ' '.join(compat_shlex_quote(a) for a in args)
2381 def error_to_compat_str(err):
2383 # On python 2 error byte string must be decoded with proper
2384 # encoding rather than ascii
2385 if sys.version_info[0] < 3:
2386 err_str = err_str.decode(preferredencoding())
2390 def mimetype2ext(mt):
2396 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
2397 # it's the most popular one
2398 'audio/mpeg': 'mp3',
2403 _, _, res = mt.rpartition('/')
2404 res = res.split(';')[0].strip().lower()
2408 'smptett+xml': 'tt',
2412 'x-mp4-fragmented': 'mp4',
2413 'x-ms-sami': 'sami',
2416 'x-mpegurl': 'm3u8',
2417 'vnd.apple.mpegurl': 'm3u8',
2421 'vnd.ms-sstr+xml': 'ism',
2427 def parse_codecs(codecs_str):
2428 # http://tools.ietf.org/html/rfc6381
2431 splited_codecs = list(filter(None, map(
2432 lambda str: str.strip(), codecs_str.strip().strip(',').split(','))))
2433 vcodec, acodec = None, None
2434 for full_codec in splited_codecs:
2435 codec = full_codec.split('.')[0]
2436 if codec in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v', 'hvc1'):
2439 elif codec in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
2443 write_string('WARNING: Unknown codec %s\n' % full_codec, sys.stderr)
2444 if not vcodec and not acodec:
2445 if len(splited_codecs) == 2:
2450 elif len(splited_codecs) == 1:
2457 'vcodec': vcodec or 'none',
2458 'acodec': acodec or 'none',
2463 def urlhandle_detect_ext(url_handle):
2464 getheader = url_handle.headers.get
2466 cd = getheader('Content-Disposition')
2468 m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd)
2470 e = determine_ext(m.group('filename'), default_ext=None)
2474 return mimetype2ext(getheader('Content-Type'))
2477 def encode_data_uri(data, mime_type):
2478 return 'data:%s;base64,%s' % (mime_type, base64.b64encode(data).decode('ascii'))
2481 def age_restricted(content_limit, age_limit):
2482 """ Returns True iff the content should be blocked """
2484 if age_limit is None: # No limit set
2486 if content_limit is None:
2487 return False # Content available for everyone
2488 return age_limit < content_limit
2491 def is_html(first_bytes):
2492 """ Detect whether a file contains HTML by examining its first bytes. """
2495 (b'\xef\xbb\xbf', 'utf-8'),
2496 (b'\x00\x00\xfe\xff', 'utf-32-be'),
2497 (b'\xff\xfe\x00\x00', 'utf-32-le'),
2498 (b'\xff\xfe', 'utf-16-le'),
2499 (b'\xfe\xff', 'utf-16-be'),
2501 for bom, enc in BOMS:
2502 if first_bytes.startswith(bom):
2503 s = first_bytes[len(bom):].decode(enc, 'replace')
2506 s = first_bytes.decode('utf-8', 'replace')
2508 return re.match(r'^\s*<', s)
2511 def determine_protocol(info_dict):
2512 protocol = info_dict.get('protocol')
2513 if protocol is not None:
2516 url = info_dict['url']
2517 if url.startswith('rtmp'):
2519 elif url.startswith('mms'):
2521 elif url.startswith('rtsp'):
2524 ext = determine_ext(url)
2530 return compat_urllib_parse_urlparse(url).scheme
2533 def render_table(header_row, data):
2534 """ Render a list of rows, each as a list of values """
2535 table = [header_row] + data
2536 max_lens = [max(len(compat_str(v)) for v in col) for col in zip(*table)]
2537 format_str = ' '.join('%-' + compat_str(ml + 1) + 's' for ml in max_lens[:-1]) + '%s'
2538 return '\n'.join(format_str % tuple(row) for row in table)
2541 def _match_one(filter_part, dct):
2542 COMPARISON_OPERATORS = {
2550 operator_rex = re.compile(r'''(?x)\s*
2552 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
2554 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
2555 (?P<quote>["\'])(?P<quotedstrval>(?:\\.|(?!(?P=quote)|\\).)+?)(?P=quote)|
2556 (?P<strval>(?![0-9.])[a-z0-9A-Z]*)
2559 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
2560 m = operator_rex.search(filter_part)
2562 op = COMPARISON_OPERATORS[m.group('op')]
2563 actual_value = dct.get(m.group('key'))
2564 if (m.group('quotedstrval') is not None or
2565 m.group('strval') is not None or
2566 # If the original field is a string and matching comparisonvalue is
2567 # a number we should respect the origin of the original field
2568 # and process comparison value as a string (see
2569 # https://github.com/rg3/youtube-dl/issues/11082).
2570 actual_value is not None and m.group('intval') is not None and
2571 isinstance(actual_value, compat_str)):
2572 if m.group('op') not in ('=', '!='):
2574 'Operator %s does not support string values!' % m.group('op'))
2575 comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval')
2576 quote = m.group('quote')
2577 if quote is not None:
2578 comparison_value = comparison_value.replace(r'\%s' % quote, quote)
2581 comparison_value = int(m.group('intval'))
2583 comparison_value = parse_filesize(m.group('intval'))
2584 if comparison_value is None:
2585 comparison_value = parse_filesize(m.group('intval') + 'B')
2586 if comparison_value is None:
2588 'Invalid integer value %r in filter part %r' % (
2589 m.group('intval'), filter_part))
2590 if actual_value is None:
2591 return m.group('none_inclusive')
2592 return op(actual_value, comparison_value)
2595 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
2596 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
2598 operator_rex = re.compile(r'''(?x)\s*
2599 (?P<op>%s)\s*(?P<key>[a-z_]+)
2601 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
2602 m = operator_rex.search(filter_part)
2604 op = UNARY_OPERATORS[m.group('op')]
2605 actual_value = dct.get(m.group('key'))
2606 return op(actual_value)
2608 raise ValueError('Invalid filter part %r' % filter_part)
2611 def match_str(filter_str, dct):
2612 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """
2615 _match_one(filter_part, dct) for filter_part in filter_str.split('&'))
2618 def match_filter_func(filter_str):
2619 def _match_func(info_dict):
2620 if match_str(filter_str, info_dict):
2623 video_title = info_dict.get('title', info_dict.get('id', 'video'))
2624 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
2628 def parse_dfxp_time_expr(time_expr):
2632 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
2634 return float(mobj.group('time_offset'))
2636 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
2638 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
2641 def srt_subtitles_timecode(seconds):
2642 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
2645 def dfxp2srt(dfxp_data):
2647 @param dfxp_data A bytes-like object containing DFXP data
2648 @returns A unicode object containing converted SRT data
2650 LEGACY_NAMESPACES = (
2651 (b'http://www.w3.org/ns/ttml', [
2652 b'http://www.w3.org/2004/11/ttaf1',
2653 b'http://www.w3.org/2006/04/ttaf1',
2654 b'http://www.w3.org/2006/10/ttaf1',
2656 (b'http://www.w3.org/ns/ttml#styling', [
2657 b'http://www.w3.org/ns/ttml#style',
2661 SUPPORTED_STYLING = [
2670 _x = functools.partial(xpath_with_ns, ns_map={
2671 'xml': 'http://www.w3.org/XML/1998/namespace',
2672 'ttml': 'http://www.w3.org/ns/ttml',
2673 'tts': 'http://www.w3.org/ns/ttml#styling',
2679 class TTMLPElementParser(object):
2681 _unclosed_elements = []
2682 _applied_styles = []
2684 def start(self, tag, attrib):
2685 if tag in (_x('ttml:br'), 'br'):
2688 unclosed_elements = []
2690 element_style_id = attrib.get('style')
2692 style.update(default_style)
2693 if element_style_id:
2694 style.update(styles.get(element_style_id, {}))
2695 for prop in SUPPORTED_STYLING:
2696 prop_val = attrib.get(_x('tts:' + prop))
2698 style[prop] = prop_val
2701 for k, v in sorted(style.items()):
2702 if self._applied_styles and self._applied_styles[-1].get(k) == v:
2705 font += ' color="%s"' % v
2706 elif k == 'fontSize':
2707 font += ' size="%s"' % v
2708 elif k == 'fontFamily':
2709 font += ' face="%s"' % v
2710 elif k == 'fontWeight' and v == 'bold':
2712 unclosed_elements.append('b')
2713 elif k == 'fontStyle' and v == 'italic':
2715 unclosed_elements.append('i')
2716 elif k == 'textDecoration' and v == 'underline':
2718 unclosed_elements.append('u')
2720 self._out += '<font' + font + '>'
2721 unclosed_elements.append('font')
2723 if self._applied_styles:
2724 applied_style.update(self._applied_styles[-1])
2725 applied_style.update(style)
2726 self._applied_styles.append(applied_style)
2727 self._unclosed_elements.append(unclosed_elements)
2730 if tag not in (_x('ttml:br'), 'br'):
2731 unclosed_elements = self._unclosed_elements.pop()
2732 for element in reversed(unclosed_elements):
2733 self._out += '</%s>' % element
2734 if unclosed_elements and self._applied_styles:
2735 self._applied_styles.pop()
2737 def data(self, data):
2741 return self._out.strip()
2743 def parse_node(node):
2744 target = TTMLPElementParser()
2745 parser = xml.etree.ElementTree.XMLParser(target=target)
2746 parser.feed(xml.etree.ElementTree.tostring(node))
2747 return parser.close()
2749 for k, v in LEGACY_NAMESPACES:
2751 dfxp_data = dfxp_data.replace(ns, k)
2753 dfxp = compat_etree_fromstring(dfxp_data)
2755 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
2758 raise ValueError('Invalid dfxp/TTML subtitle')
2762 for style in dfxp.findall(_x('.//ttml:style')):
2763 style_id = style.get('id') or style.get(_x('xml:id'))
2766 parent_style_id = style.get('style')
2768 if parent_style_id not in styles:
2771 styles[style_id] = styles[parent_style_id].copy()
2772 for prop in SUPPORTED_STYLING:
2773 prop_val = style.get(_x('tts:' + prop))
2775 styles.setdefault(style_id, {})[prop] = prop_val
2781 for p in ('body', 'div'):
2782 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
2785 style = styles.get(ele.get('style'))
2788 default_style.update(style)
2790 for para, index in zip(paras, itertools.count(1)):
2791 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
2792 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
2793 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
2794 if begin_time is None:
2799 end_time = begin_time + dur
2800 out.append('%d\n%s --> %s\n%s\n\n' % (
2802 srt_subtitles_timecode(begin_time),
2803 srt_subtitles_timecode(end_time),
2809 def cli_option(params, command_option, param):
2810 param = params.get(param)
2812 param = compat_str(param)
2813 return [command_option, param] if param is not None else []
2816 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
2817 param = params.get(param)
2820 assert isinstance(param, bool)
2822 return [command_option + separator + (true_value if param else false_value)]
2823 return [command_option, true_value if param else false_value]
2826 def cli_valueless_option(params, command_option, param, expected_value=True):
2827 param = params.get(param)
2828 return [command_option] if param == expected_value else []
2831 def cli_configuration_args(params, param, default=[]):
2832 ex_args = params.get(param)
2835 assert isinstance(ex_args, list)
2839 class ISO639Utils(object):
2840 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3029 def short2long(cls, code):
3030 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3031 return cls._lang_map.get(code[:2])
3034 def long2short(cls, code):
3035 """Convert language code from ISO 639-2/T to ISO 639-1"""
3036 for short_name, long_name in cls._lang_map.items():
3037 if long_name == code:
3041 class ISO3166Utils(object):
3042 # From http://data.okfn.org/data/core/country-list
3044 'AF': 'Afghanistan',
3045 'AX': 'Åland Islands',
3048 'AS': 'American Samoa',
3053 'AG': 'Antigua and Barbuda',
3070 'BO': 'Bolivia, Plurinational State of',
3071 'BQ': 'Bonaire, Sint Eustatius and Saba',
3072 'BA': 'Bosnia and Herzegovina',
3074 'BV': 'Bouvet Island',
3076 'IO': 'British Indian Ocean Territory',
3077 'BN': 'Brunei Darussalam',
3079 'BF': 'Burkina Faso',
3085 'KY': 'Cayman Islands',
3086 'CF': 'Central African Republic',
3090 'CX': 'Christmas Island',
3091 'CC': 'Cocos (Keeling) Islands',
3095 'CD': 'Congo, the Democratic Republic of the',
3096 'CK': 'Cook Islands',
3098 'CI': 'Côte d\'Ivoire',
3103 'CZ': 'Czech Republic',
3107 'DO': 'Dominican Republic',
3110 'SV': 'El Salvador',
3111 'GQ': 'Equatorial Guinea',
3115 'FK': 'Falkland Islands (Malvinas)',
3116 'FO': 'Faroe Islands',
3120 'GF': 'French Guiana',
3121 'PF': 'French Polynesia',
3122 'TF': 'French Southern Territories',
3137 'GW': 'Guinea-Bissau',
3140 'HM': 'Heard Island and McDonald Islands',
3141 'VA': 'Holy See (Vatican City State)',
3148 'IR': 'Iran, Islamic Republic of',
3151 'IM': 'Isle of Man',
3161 'KP': 'Korea, Democratic People\'s Republic of',
3162 'KR': 'Korea, Republic of',
3165 'LA': 'Lao People\'s Democratic Republic',
3171 'LI': 'Liechtenstein',
3175 'MK': 'Macedonia, the Former Yugoslav Republic of',
3182 'MH': 'Marshall Islands',
3188 'FM': 'Micronesia, Federated States of',
3189 'MD': 'Moldova, Republic of',
3200 'NL': 'Netherlands',
3201 'NC': 'New Caledonia',
3202 'NZ': 'New Zealand',
3207 'NF': 'Norfolk Island',
3208 'MP': 'Northern Mariana Islands',
3213 'PS': 'Palestine, State of',
3215 'PG': 'Papua New Guinea',
3218 'PH': 'Philippines',
3222 'PR': 'Puerto Rico',
3226 'RU': 'Russian Federation',
3228 'BL': 'Saint Barthélemy',
3229 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
3230 'KN': 'Saint Kitts and Nevis',
3231 'LC': 'Saint Lucia',
3232 'MF': 'Saint Martin (French part)',
3233 'PM': 'Saint Pierre and Miquelon',
3234 'VC': 'Saint Vincent and the Grenadines',
3237 'ST': 'Sao Tome and Principe',
3238 'SA': 'Saudi Arabia',
3242 'SL': 'Sierra Leone',
3244 'SX': 'Sint Maarten (Dutch part)',
3247 'SB': 'Solomon Islands',
3249 'ZA': 'South Africa',
3250 'GS': 'South Georgia and the South Sandwich Islands',
3251 'SS': 'South Sudan',
3256 'SJ': 'Svalbard and Jan Mayen',
3259 'CH': 'Switzerland',
3260 'SY': 'Syrian Arab Republic',
3261 'TW': 'Taiwan, Province of China',
3263 'TZ': 'Tanzania, United Republic of',
3265 'TL': 'Timor-Leste',
3269 'TT': 'Trinidad and Tobago',
3272 'TM': 'Turkmenistan',
3273 'TC': 'Turks and Caicos Islands',
3277 'AE': 'United Arab Emirates',
3278 'GB': 'United Kingdom',
3279 'US': 'United States',
3280 'UM': 'United States Minor Outlying Islands',
3284 'VE': 'Venezuela, Bolivarian Republic of',
3286 'VG': 'Virgin Islands, British',
3287 'VI': 'Virgin Islands, U.S.',
3288 'WF': 'Wallis and Futuna',
3289 'EH': 'Western Sahara',
3296 def short2full(cls, code):
3297 """Convert an ISO 3166-2 country code to the corresponding full name"""
3298 return cls._country_map.get(code.upper())
3301 class GeoUtils(object):
3302 # Major IPv4 address blocks per country
3304 'AD': '85.94.160.0/19',
3305 'AE': '94.200.0.0/13',
3306 'AF': '149.54.0.0/17',
3307 'AG': '209.59.64.0/18',
3308 'AI': '204.14.248.0/21',
3309 'AL': '46.99.0.0/16',
3310 'AM': '46.70.0.0/15',
3311 'AO': '105.168.0.0/13',
3312 'AP': '159.117.192.0/21',
3313 'AR': '181.0.0.0/12',
3314 'AS': '202.70.112.0/20',
3315 'AT': '84.112.0.0/13',
3316 'AU': '1.128.0.0/11',
3317 'AW': '181.41.0.0/18',
3318 'AZ': '5.191.0.0/16',
3319 'BA': '31.176.128.0/17',
3320 'BB': '65.48.128.0/17',
3321 'BD': '114.130.0.0/16',
3323 'BF': '129.45.128.0/17',
3324 'BG': '95.42.0.0/15',
3325 'BH': '37.131.0.0/17',
3326 'BI': '154.117.192.0/18',
3327 'BJ': '137.255.0.0/16',
3328 'BL': '192.131.134.0/24',
3329 'BM': '196.12.64.0/18',
3330 'BN': '156.31.0.0/16',
3331 'BO': '161.56.0.0/16',
3332 'BQ': '161.0.80.0/20',
3333 'BR': '152.240.0.0/12',
3334 'BS': '24.51.64.0/18',
3335 'BT': '119.2.96.0/19',
3336 'BW': '168.167.0.0/16',
3337 'BY': '178.120.0.0/13',
3338 'BZ': '179.42.192.0/18',
3339 'CA': '99.224.0.0/11',
3340 'CD': '41.243.0.0/16',
3341 'CF': '196.32.200.0/21',
3342 'CG': '197.214.128.0/17',
3343 'CH': '85.0.0.0/13',
3344 'CI': '154.232.0.0/14',
3345 'CK': '202.65.32.0/19',
3346 'CL': '152.172.0.0/14',
3347 'CM': '165.210.0.0/15',
3348 'CN': '36.128.0.0/10',
3349 'CO': '181.240.0.0/12',
3350 'CR': '201.192.0.0/12',
3351 'CU': '152.206.0.0/15',
3352 'CV': '165.90.96.0/19',
3353 'CW': '190.88.128.0/17',
3354 'CY': '46.198.0.0/15',
3355 'CZ': '88.100.0.0/14',
3357 'DJ': '197.241.0.0/17',
3358 'DK': '87.48.0.0/12',
3359 'DM': '192.243.48.0/20',
3360 'DO': '152.166.0.0/15',
3361 'DZ': '41.96.0.0/12',
3362 'EC': '186.68.0.0/15',
3363 'EE': '90.190.0.0/15',
3364 'EG': '156.160.0.0/11',
3365 'ER': '196.200.96.0/20',
3366 'ES': '88.0.0.0/11',
3367 'ET': '196.188.0.0/14',
3368 'EU': '2.16.0.0/13',
3369 'FI': '91.152.0.0/13',
3370 'FJ': '144.120.0.0/16',
3371 'FM': '119.252.112.0/20',
3372 'FO': '88.85.32.0/19',
3374 'GA': '41.158.0.0/15',
3376 'GD': '74.122.88.0/21',
3377 'GE': '31.146.0.0/16',
3378 'GF': '161.22.64.0/18',
3379 'GG': '62.68.160.0/19',
3380 'GH': '45.208.0.0/14',
3381 'GI': '85.115.128.0/19',
3382 'GL': '88.83.0.0/19',
3383 'GM': '160.182.0.0/15',
3384 'GN': '197.149.192.0/18',
3385 'GP': '104.250.0.0/19',
3386 'GQ': '105.235.224.0/20',
3387 'GR': '94.64.0.0/13',
3388 'GT': '168.234.0.0/16',
3389 'GU': '168.123.0.0/16',
3390 'GW': '197.214.80.0/20',
3391 'GY': '181.41.64.0/18',
3392 'HK': '113.252.0.0/14',
3393 'HN': '181.210.0.0/16',
3394 'HR': '93.136.0.0/13',
3395 'HT': '148.102.128.0/17',
3396 'HU': '84.0.0.0/14',
3397 'ID': '39.192.0.0/10',
3398 'IE': '87.32.0.0/12',
3399 'IL': '79.176.0.0/13',
3400 'IM': '5.62.80.0/20',
3401 'IN': '117.192.0.0/10',
3402 'IO': '203.83.48.0/21',
3403 'IQ': '37.236.0.0/14',
3404 'IR': '2.176.0.0/12',
3405 'IS': '82.221.0.0/16',
3406 'IT': '79.0.0.0/10',
3407 'JE': '87.244.64.0/18',
3408 'JM': '72.27.0.0/17',
3409 'JO': '176.29.0.0/16',
3410 'JP': '126.0.0.0/8',
3411 'KE': '105.48.0.0/12',
3412 'KG': '158.181.128.0/17',
3413 'KH': '36.37.128.0/17',
3414 'KI': '103.25.140.0/22',
3415 'KM': '197.255.224.0/20',
3416 'KN': '198.32.32.0/19',
3417 'KP': '175.45.176.0/22',
3418 'KR': '175.192.0.0/10',
3419 'KW': '37.36.0.0/14',
3420 'KY': '64.96.0.0/15',
3421 'KZ': '2.72.0.0/13',
3422 'LA': '115.84.64.0/18',
3423 'LB': '178.135.0.0/16',
3424 'LC': '192.147.231.0/24',
3425 'LI': '82.117.0.0/19',
3426 'LK': '112.134.0.0/15',
3427 'LR': '41.86.0.0/19',
3428 'LS': '129.232.0.0/17',
3429 'LT': '78.56.0.0/13',
3430 'LU': '188.42.0.0/16',
3431 'LV': '46.109.0.0/16',
3432 'LY': '41.252.0.0/14',
3433 'MA': '105.128.0.0/11',
3434 'MC': '88.209.64.0/18',
3435 'MD': '37.246.0.0/16',
3436 'ME': '178.175.0.0/17',
3437 'MF': '74.112.232.0/21',
3438 'MG': '154.126.0.0/17',
3439 'MH': '117.103.88.0/21',
3440 'MK': '77.28.0.0/15',
3441 'ML': '154.118.128.0/18',
3442 'MM': '37.111.0.0/17',
3443 'MN': '49.0.128.0/17',
3444 'MO': '60.246.0.0/16',
3445 'MP': '202.88.64.0/20',
3446 'MQ': '109.203.224.0/19',
3447 'MR': '41.188.64.0/18',
3448 'MS': '208.90.112.0/22',
3449 'MT': '46.11.0.0/16',
3450 'MU': '105.16.0.0/12',
3451 'MV': '27.114.128.0/18',
3452 'MW': '105.234.0.0/16',
3453 'MX': '187.192.0.0/11',
3454 'MY': '175.136.0.0/13',
3455 'MZ': '197.218.0.0/15',
3456 'NA': '41.182.0.0/16',
3457 'NC': '101.101.0.0/18',
3458 'NE': '197.214.0.0/18',
3459 'NF': '203.17.240.0/22',
3460 'NG': '105.112.0.0/12',
3461 'NI': '186.76.0.0/15',
3462 'NL': '145.96.0.0/11',
3463 'NO': '84.208.0.0/13',
3464 'NP': '36.252.0.0/15',
3465 'NR': '203.98.224.0/19',
3466 'NU': '49.156.48.0/22',
3467 'NZ': '49.224.0.0/14',
3468 'OM': '5.36.0.0/15',
3469 'PA': '186.72.0.0/15',
3470 'PE': '186.160.0.0/14',
3471 'PF': '123.50.64.0/18',
3472 'PG': '124.240.192.0/19',
3473 'PH': '49.144.0.0/13',
3474 'PK': '39.32.0.0/11',
3475 'PL': '83.0.0.0/11',
3476 'PM': '70.36.0.0/20',
3477 'PR': '66.50.0.0/16',
3478 'PS': '188.161.0.0/16',
3479 'PT': '85.240.0.0/13',
3480 'PW': '202.124.224.0/20',
3481 'PY': '181.120.0.0/14',
3482 'QA': '37.210.0.0/15',
3483 'RE': '139.26.0.0/16',
3484 'RO': '79.112.0.0/13',
3485 'RS': '178.220.0.0/14',
3486 'RU': '5.136.0.0/13',
3487 'RW': '105.178.0.0/15',
3488 'SA': '188.48.0.0/13',
3489 'SB': '202.1.160.0/19',
3490 'SC': '154.192.0.0/11',
3491 'SD': '154.96.0.0/13',
3492 'SE': '78.64.0.0/12',
3493 'SG': '152.56.0.0/14',
3494 'SI': '188.196.0.0/14',
3495 'SK': '78.98.0.0/15',
3496 'SL': '197.215.0.0/17',
3497 'SM': '89.186.32.0/19',
3498 'SN': '41.82.0.0/15',
3499 'SO': '197.220.64.0/19',
3500 'SR': '186.179.128.0/17',
3501 'SS': '105.235.208.0/21',
3502 'ST': '197.159.160.0/19',
3503 'SV': '168.243.0.0/16',
3504 'SX': '190.102.0.0/20',
3506 'SZ': '41.84.224.0/19',
3507 'TC': '65.255.48.0/20',
3508 'TD': '154.68.128.0/19',
3509 'TG': '196.168.0.0/14',
3510 'TH': '171.96.0.0/13',
3511 'TJ': '85.9.128.0/18',
3512 'TK': '27.96.24.0/21',
3513 'TL': '180.189.160.0/20',
3514 'TM': '95.85.96.0/19',
3515 'TN': '197.0.0.0/11',
3516 'TO': '175.176.144.0/21',
3517 'TR': '78.160.0.0/11',
3518 'TT': '186.44.0.0/15',
3519 'TV': '202.2.96.0/19',
3520 'TW': '120.96.0.0/11',
3521 'TZ': '156.156.0.0/14',
3522 'UA': '93.72.0.0/13',
3523 'UG': '154.224.0.0/13',
3525 'UY': '167.56.0.0/13',
3526 'UZ': '82.215.64.0/18',
3527 'VA': '212.77.0.0/19',
3528 'VC': '24.92.144.0/20',
3529 'VE': '186.88.0.0/13',
3530 'VG': '172.103.64.0/18',
3531 'VI': '146.226.0.0/16',
3532 'VN': '14.160.0.0/11',
3533 'VU': '202.80.32.0/20',
3534 'WF': '117.20.32.0/21',
3535 'WS': '202.4.32.0/19',
3536 'YE': '134.35.0.0/16',
3537 'YT': '41.242.116.0/22',
3538 'ZA': '41.0.0.0/11',
3539 'ZM': '165.56.0.0/13',
3540 'ZW': '41.85.192.0/19',
3544 def random_ipv4(cls, code_or_block):
3545 if len(code_or_block) == 2:
3546 block = cls._country_ip_map.get(code_or_block.upper())
3550 block = code_or_block
3551 addr, preflen = block.split('/')
3552 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
3553 addr_max = addr_min | (0xffffffff >> int(preflen))
3554 return compat_str(socket.inet_ntoa(
3555 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
3558 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
3559 def __init__(self, proxies=None):
3560 # Set default handlers
3561 for type in ('http', 'https'):
3562 setattr(self, '%s_open' % type,
3563 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
3564 meth(r, proxy, type))
3565 return compat_urllib_request.ProxyHandler.__init__(self, proxies)
3567 def proxy_open(self, req, proxy, type):
3568 req_proxy = req.headers.get('Ytdl-request-proxy')
3569 if req_proxy is not None:
3571 del req.headers['Ytdl-request-proxy']
3573 if proxy == '__noproxy__':
3574 return None # No Proxy
3575 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
3576 req.add_header('Ytdl-socks-proxy', proxy)
3577 # youtube-dl's http/https handlers do wrapping the socket with socks
3579 return compat_urllib_request.ProxyHandler.proxy_open(
3580 self, req, proxy, type)
3583 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
3584 # released into Public Domain
3585 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
3587 def long_to_bytes(n, blocksize=0):
3588 """long_to_bytes(n:long, blocksize:int) : string
3589 Convert a long integer to a byte string.
3591 If optional blocksize is given and greater than zero, pad the front of the
3592 byte string with binary zeros so that the length is a multiple of
3595 # after much testing, this algorithm was deemed to be the fastest
3599 s = compat_struct_pack('>I', n & 0xffffffff) + s
3601 # strip off leading zeros
3602 for i in range(len(s)):
3603 if s[i] != b'\000'[0]:
3606 # only happens when n == 0
3610 # add back some pad bytes. this could be done more efficiently w.r.t. the
3611 # de-padding being done above, but sigh...
3612 if blocksize > 0 and len(s) % blocksize:
3613 s = (blocksize - len(s) % blocksize) * b'\000' + s
3617 def bytes_to_long(s):
3618 """bytes_to_long(string) : long
3619 Convert a byte string to a long integer.
3621 This is (essentially) the inverse of long_to_bytes().
3626 extra = (4 - length % 4)
3627 s = b'\000' * extra + s
3628 length = length + extra
3629 for i in range(0, length, 4):
3630 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
3634 def ohdave_rsa_encrypt(data, exponent, modulus):
3636 Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
3639 data: data to encrypt, bytes-like object
3640 exponent, modulus: parameter e and N of RSA algorithm, both integer
3641 Output: hex string of encrypted data
3643 Limitation: supports one block encryption only
3646 payload = int(binascii.hexlify(data[::-1]), 16)
3647 encrypted = pow(payload, exponent, modulus)
3648 return '%x' % encrypted
3651 def pkcs1pad(data, length):
3653 Padding input data with PKCS#1 scheme
3655 @param {int[]} data input data
3656 @param {int} length target length
3657 @returns {int[]} padded data
3659 if len(data) > length - 11:
3660 raise ValueError('Input data too long for PKCS#1 padding')
3662 pseudo_random = [random.randint(0, 254) for _ in range(length - len(data) - 3)]
3663 return [0, 2] + pseudo_random + [0] + data
3666 def encode_base_n(num, n, table=None):
3667 FULL_TABLE = '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
3669 table = FULL_TABLE[:n]
3672 raise ValueError('base %d exceeds table length %d' % (n, len(table)))
3679 ret = table[num % n] + ret
3684 def decode_packed_codes(code):
3685 mobj = re.search(PACKED_CODES_RE, code)
3686 obfucasted_code, base, count, symbols = mobj.groups()
3689 symbols = symbols.split('|')
3694 base_n_count = encode_base_n(count, base)
3695 symbol_table[base_n_count] = symbols[count] or base_n_count
3698 r'\b(\w+)\b', lambda mobj: symbol_table[mobj.group(0)],
3702 def parse_m3u8_attributes(attrib):
3704 for (key, val) in re.findall(r'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib):
3705 if val.startswith('"'):
3711 def urshift(val, n):
3712 return val >> n if val >= 0 else (val + 0x100000000) >> n
3715 # Based on png2str() written by @gdkchan and improved by @yokrysty
3716 # Originally posted at https://github.com/rg3/youtube-dl/issues/9706
3717 def decode_png(png_data):
3718 # Reference: https://www.w3.org/TR/PNG/
3719 header = png_data[8:]
3721 if png_data[:8] != b'\x89PNG\x0d\x0a\x1a\x0a' or header[4:8] != b'IHDR':
3722 raise IOError('Not a valid PNG file.')
3724 int_map = {1: '>B', 2: '>H', 4: '>I'}
3725 unpack_integer = lambda x: compat_struct_unpack(int_map[len(x)], x)[0]
3730 length = unpack_integer(header[:4])
3733 chunk_type = header[:4]
3736 chunk_data = header[:length]
3737 header = header[length:]
3739 header = header[4:] # Skip CRC
3747 ihdr = chunks[0]['data']
3749 width = unpack_integer(ihdr[:4])
3750 height = unpack_integer(ihdr[4:8])
3754 for chunk in chunks:
3755 if chunk['type'] == b'IDAT':
3756 idat += chunk['data']
3759 raise IOError('Unable to read PNG data.')
3761 decompressed_data = bytearray(zlib.decompress(idat))
3766 def _get_pixel(idx):
3771 for y in range(height):
3772 basePos = y * (1 + stride)
3773 filter_type = decompressed_data[basePos]
3777 pixels.append(current_row)
3779 for x in range(stride):
3780 color = decompressed_data[1 + basePos + x]
3781 basex = y * stride + x
3786 left = _get_pixel(basex - 3)
3788 up = _get_pixel(basex - stride)
3790 if filter_type == 1: # Sub
3791 color = (color + left) & 0xff
3792 elif filter_type == 2: # Up
3793 color = (color + up) & 0xff
3794 elif filter_type == 3: # Average
3795 color = (color + ((left + up) >> 1)) & 0xff
3796 elif filter_type == 4: # Paeth
3802 c = _get_pixel(basex - stride - 3)
3810 if pa <= pb and pa <= pc:
3811 color = (color + a) & 0xff
3813 color = (color + b) & 0xff
3815 color = (color + c) & 0xff
3817 current_row.append(color)
3819 return width, height, pixels
3822 def write_xattr(path, key, value):
3823 # This mess below finds the best xattr tool for the job
3825 # try the pyxattr module...
3828 if hasattr(xattr, 'set'): # pyxattr
3829 # Unicode arguments are not supported in python-pyxattr until
3831 # See https://github.com/rg3/youtube-dl/issues/5498
3832 pyxattr_required_version = '0.5.0'
3833 if version_tuple(xattr.__version__) < version_tuple(pyxattr_required_version):
3834 # TODO: fallback to CLI tools
3835 raise XAttrUnavailableError(
3836 'python-pyxattr is detected but is too old. '
3837 'youtube-dl requires %s or above while your version is %s. '
3838 'Falling back to other xattr implementations' % (
3839 pyxattr_required_version, xattr.__version__))
3841 setxattr = xattr.set
3843 setxattr = xattr.setxattr
3846 setxattr(path, key, value)
3847 except EnvironmentError as e:
3848 raise XAttrMetadataError(e.errno, e.strerror)
3851 if compat_os_name == 'nt':
3852 # Write xattrs to NTFS Alternate Data Streams:
3853 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
3854 assert ':' not in key
3855 assert os.path.exists(path)
3857 ads_fn = path + ':' + key
3859 with open(ads_fn, 'wb') as f:
3861 except EnvironmentError as e:
3862 raise XAttrMetadataError(e.errno, e.strerror)
3864 user_has_setfattr = check_executable('setfattr', ['--version'])
3865 user_has_xattr = check_executable('xattr', ['-h'])
3867 if user_has_setfattr or user_has_xattr:
3869 value = value.decode('utf-8')
3870 if user_has_setfattr:
3871 executable = 'setfattr'
3872 opts = ['-n', key, '-v', value]
3873 elif user_has_xattr:
3874 executable = 'xattr'
3875 opts = ['-w', key, value]
3877 cmd = ([encodeFilename(executable, True)] +
3878 [encodeArgument(o) for o in opts] +
3879 [encodeFilename(path, True)])
3882 p = subprocess.Popen(
3883 cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, stdin=subprocess.PIPE)
3884 except EnvironmentError as e:
3885 raise XAttrMetadataError(e.errno, e.strerror)
3886 stdout, stderr = p.communicate()
3887 stderr = stderr.decode('utf-8', 'replace')
3888 if p.returncode != 0:
3889 raise XAttrMetadataError(p.returncode, stderr)
3892 # On Unix, and can't find pyxattr, setfattr, or xattr.
3893 if sys.platform.startswith('linux'):
3894 raise XAttrUnavailableError(
3895 "Couldn't find a tool to set the xattrs. "
3896 "Install either the python 'pyxattr' or 'xattr' "
3897 "modules, or the GNU 'attr' package "
3898 "(which contains the 'setfattr' tool).")
3900 raise XAttrUnavailableError(
3901 "Couldn't find a tool to set the xattrs. "
3902 "Install either the python 'xattr' module, "
3903 "or the 'xattr' binary.")
3906 def random_birthday(year_field, month_field, day_field):
3908 year_field: str(random.randint(1950, 1995)),
3909 month_field: str(random.randint(1, 12)),
3910 day_field: str(random.randint(1, 31)),