2 # -*- coding: utf-8 -*-
4 from __future__ import unicode_literals
33 import xml.etree.ElementTree
39 compat_etree_fromstring,
44 compat_socket_create_connection,
48 compat_urllib_parse_urlparse,
49 compat_urllib_request,
55 # This is not clearly defined otherwise
56 compiled_regex_type = type(re.compile(''))
59 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)',
60 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
61 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
62 'Accept-Encoding': 'gzip, deflate',
63 'Accept-Language': 'en-us,en;q=0.5',
69 ENGLISH_MONTH_NAMES = [
70 'January', 'February', 'March', 'April', 'May', 'June',
71 'July', 'August', 'September', 'October', 'November', 'December']
74 def preferredencoding():
75 """Get preferred encoding.
77 Returns the best encoding scheme for the system, based on
78 locale.getpreferredencoding() and some further tweaks.
81 pref = locale.getpreferredencoding()
89 def write_json_file(obj, fn):
90 """ Encode obj as JSON and write it to fn, atomically if possible """
92 fn = encodeFilename(fn)
93 if sys.version_info < (3, 0) and sys.platform != 'win32':
94 encoding = get_filesystem_encoding()
95 # os.path.basename returns a bytes object, but NamedTemporaryFile
96 # will fail if the filename contains non ascii characters unless we
97 # use a unicode object
98 path_basename = lambda f: os.path.basename(fn).decode(encoding)
99 # the same for os.path.dirname
100 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
102 path_basename = os.path.basename
103 path_dirname = os.path.dirname
107 'prefix': path_basename(fn) + '.',
108 'dir': path_dirname(fn),
112 # In Python 2.x, json.dump expects a bytestream.
113 # In Python 3.x, it writes to a character stream
114 if sys.version_info < (3, 0):
122 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
127 if sys.platform == 'win32':
128 # Need to remove existing file on Windows, else os.rename raises
129 # WindowsError or FileExistsError.
134 os.rename(tf.name, fn)
143 if sys.version_info >= (2, 7):
144 def find_xpath_attr(node, xpath, key, val=None):
145 """ Find the xpath xpath[@key=val] """
146 assert re.match(r'^[a-zA-Z_-]+$', key)
148 assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val)
149 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
150 return node.find(expr)
152 def find_xpath_attr(node, xpath, key, val=None):
153 # Here comes the crazy part: In 2.6, if the xpath is a unicode,
154 # .//node does not match if a node is a direct child of . !
155 if isinstance(xpath, compat_str):
156 xpath = xpath.encode('ascii')
158 for f in node.findall(xpath):
159 if key not in f.attrib:
161 if val is None or f.attrib.get(key) == val:
165 # On python2.6 the xml.etree.ElementTree.Element methods don't support
166 # the namespace parameter
169 def xpath_with_ns(path, ns_map):
170 components = [c.split(':') for c in path.split('/')]
174 replaced.append(c[0])
177 replaced.append('{%s}%s' % (ns_map[ns], tag))
178 return '/'.join(replaced)
181 def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
182 def _find_xpath(xpath):
183 if sys.version_info < (2, 7): # Crazy 2.6
184 xpath = xpath.encode('ascii')
185 return node.find(xpath)
187 if isinstance(xpath, (str, compat_str)):
188 n = _find_xpath(xpath)
196 if default is not NO_DEFAULT:
199 name = xpath if name is None else name
200 raise ExtractorError('Could not find XML element %s' % name)
206 def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
207 n = xpath_element(node, xpath, name, fatal=fatal, default=default)
208 if n is None or n == default:
211 if default is not NO_DEFAULT:
214 name = xpath if name is None else name
215 raise ExtractorError('Could not find XML element\'s text %s' % name)
221 def xpath_attr(node, xpath, key, name=None, fatal=False, default=NO_DEFAULT):
222 n = find_xpath_attr(node, xpath, key)
224 if default is not NO_DEFAULT:
227 name = '%s[@%s]' % (xpath, key) if name is None else name
228 raise ExtractorError('Could not find XML attribute %s' % name)
234 def get_element_by_id(id, html):
235 """Return the content of the tag with the specified ID in the passed HTML document"""
236 return get_element_by_attribute("id", id, html)
239 def get_element_by_attribute(attribute, value, html):
240 """Return the content of the tag with the specified attribute in the passed HTML document"""
242 m = re.search(r'''(?xs)
244 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
246 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
250 ''' % (re.escape(attribute), re.escape(value)), html)
254 res = m.group('content')
256 if res.startswith('"') or res.startswith("'"):
259 return unescapeHTML(res)
262 def clean_html(html):
263 """Clean an HTML snippet into a readable string"""
265 if html is None: # Convenience for sanitizing descriptions etc.
269 html = html.replace('\n', ' ')
270 html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html)
271 html = re.sub(r'<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
273 html = re.sub('<.*?>', '', html)
274 # Replace html entities
275 html = unescapeHTML(html)
279 def sanitize_open(filename, open_mode):
280 """Try to open the given filename, and slightly tweak it if this fails.
282 Attempts to open the given filename. If this fails, it tries to change
283 the filename slightly, step by step, until it's either able to open it
284 or it fails and raises a final exception, like the standard open()
287 It returns the tuple (stream, definitive_file_name).
291 if sys.platform == 'win32':
293 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
294 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
295 stream = open(encodeFilename(filename), open_mode)
296 return (stream, filename)
297 except (IOError, OSError) as err:
298 if err.errno in (errno.EACCES,):
301 # In case of error, try to remove win32 forbidden chars
302 alt_filename = sanitize_path(filename)
303 if alt_filename == filename:
306 # An exception here should be caught in the caller
307 stream = open(encodeFilename(alt_filename), open_mode)
308 return (stream, alt_filename)
311 def timeconvert(timestr):
312 """Convert RFC 2822 defined time string into system timestamp"""
314 timetuple = email.utils.parsedate_tz(timestr)
315 if timetuple is not None:
316 timestamp = email.utils.mktime_tz(timetuple)
320 def sanitize_filename(s, restricted=False, is_id=False):
321 """Sanitizes a string so it could be used as part of a filename.
322 If restricted is set, use a stricter subset of allowed characters.
323 Set is_id if this is not an arbitrary string, but an ID that should be kept if possible
325 def replace_insane(char):
326 if char == '?' or ord(char) < 32 or ord(char) == 127:
329 return '' if restricted else '\''
331 return '_-' if restricted else ' -'
332 elif char in '\\/|*<>':
334 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
336 if restricted and ord(char) > 127:
341 s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
342 result = ''.join(map(replace_insane, s))
344 while '__' in result:
345 result = result.replace('__', '_')
346 result = result.strip('_')
347 # Common case of "Foreign band name - English song title"
348 if restricted and result.startswith('-_'):
350 if result.startswith('-'):
351 result = '_' + result[len('-'):]
352 result = result.lstrip('.')
358 def sanitize_path(s):
359 """Sanitizes and normalizes path on Windows"""
360 if sys.platform != 'win32':
362 drive_or_unc, _ = os.path.splitdrive(s)
363 if sys.version_info < (2, 7) and not drive_or_unc:
364 drive_or_unc, _ = os.path.splitunc(s)
365 norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
369 path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|[\s.]$)', '#', path_part)
370 for path_part in norm_path]
372 sanitized_path.insert(0, drive_or_unc + os.path.sep)
373 return os.path.join(*sanitized_path)
376 # Prepend protocol-less URLs with `http:` scheme in order to mitigate the number of
377 # unwanted failures due to missing protocol
378 def sanitized_Request(url, *args, **kwargs):
379 return compat_urllib_request.Request(
380 'http:%s' % url if url.startswith('//') else url, *args, **kwargs)
383 def orderedSet(iterable):
384 """ Remove all duplicates from the input iterable """
392 def _htmlentity_transform(entity):
393 """Transforms an HTML entity to a character."""
394 # Known non-numeric HTML entity
395 if entity in compat_html_entities.name2codepoint:
396 return compat_chr(compat_html_entities.name2codepoint[entity])
398 mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
400 numstr = mobj.group(1)
401 if numstr.startswith('x'):
403 numstr = '0%s' % numstr
406 # See https://github.com/rg3/youtube-dl/issues/7518
408 return compat_chr(int(numstr, base))
412 # Unknown entity in name, return its literal representation
413 return '&%s;' % entity
419 assert type(s) == compat_str
422 r'&([^;]+);', lambda m: _htmlentity_transform(m.group(1)), s)
425 def get_subprocess_encoding():
426 if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
427 # For subprocess calls, encode with locale encoding
428 # Refer to http://stackoverflow.com/a/9951851/35070
429 encoding = preferredencoding()
431 encoding = sys.getfilesystemencoding()
437 def encodeFilename(s, for_subprocess=False):
439 @param s The name of the file
442 assert type(s) == compat_str
444 # Python 3 has a Unicode API
445 if sys.version_info >= (3, 0):
448 # Pass '' directly to use Unicode APIs on Windows 2000 and up
449 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
450 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
451 if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
454 return s.encode(get_subprocess_encoding(), 'ignore')
457 def decodeFilename(b, for_subprocess=False):
459 if sys.version_info >= (3, 0):
462 if not isinstance(b, bytes):
465 return b.decode(get_subprocess_encoding(), 'ignore')
468 def encodeArgument(s):
469 if not isinstance(s, compat_str):
470 # Legacy code that uses byte strings
471 # Uncomment the following line after fixing all post processors
472 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
473 s = s.decode('ascii')
474 return encodeFilename(s, True)
477 def decodeArgument(b):
478 return decodeFilename(b, True)
481 def decodeOption(optval):
484 if isinstance(optval, bytes):
485 optval = optval.decode(preferredencoding())
487 assert isinstance(optval, compat_str)
491 def formatSeconds(secs):
493 return '%d:%02d:%02d' % (secs // 3600, (secs % 3600) // 60, secs % 60)
495 return '%d:%02d' % (secs // 60, secs % 60)
500 def make_HTTPS_handler(params, **kwargs):
501 opts_no_check_certificate = params.get('nocheckcertificate', False)
502 if hasattr(ssl, 'create_default_context'): # Python >= 3.4 or 2.7.9
503 context = ssl.create_default_context(ssl.Purpose.SERVER_AUTH)
504 if opts_no_check_certificate:
505 context.check_hostname = False
506 context.verify_mode = ssl.CERT_NONE
508 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
511 # (create_default_context present but HTTPSHandler has no context=)
514 if sys.version_info < (3, 2):
515 return YoutubeDLHTTPSHandler(params, **kwargs)
517 context = ssl.SSLContext(ssl.PROTOCOL_TLSv1)
518 context.verify_mode = (ssl.CERT_NONE
519 if opts_no_check_certificate
520 else ssl.CERT_REQUIRED)
521 context.set_default_verify_paths()
522 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
525 def bug_reports_message():
526 if ytdl_is_updateable():
527 update_cmd = 'type youtube-dl -U to update'
529 update_cmd = 'see https://yt-dl.org/update on how to update'
530 msg = '; please report this issue on https://yt-dl.org/bug .'
531 msg += ' Make sure you are using the latest version; %s.' % update_cmd
532 msg += ' Be sure to call youtube-dl with the --verbose flag and include its complete output.'
536 class ExtractorError(Exception):
537 """Error during info extraction."""
539 def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):
540 """ tb, if given, is the original traceback (so that it can be printed out).
541 If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
544 if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError):
546 if video_id is not None:
547 msg = video_id + ': ' + msg
549 msg += ' (caused by %r)' % cause
551 msg += bug_reports_message()
552 super(ExtractorError, self).__init__(msg)
555 self.exc_info = sys.exc_info() # preserve original exception
557 self.video_id = video_id
559 def format_traceback(self):
560 if self.traceback is None:
562 return ''.join(traceback.format_tb(self.traceback))
565 class UnsupportedError(ExtractorError):
566 def __init__(self, url):
567 super(UnsupportedError, self).__init__(
568 'Unsupported URL: %s' % url, expected=True)
572 class RegexNotFoundError(ExtractorError):
573 """Error when a regex didn't match"""
577 class DownloadError(Exception):
578 """Download Error exception.
580 This exception may be thrown by FileDownloader objects if they are not
581 configured to continue on errors. They will contain the appropriate
585 def __init__(self, msg, exc_info=None):
586 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
587 super(DownloadError, self).__init__(msg)
588 self.exc_info = exc_info
591 class SameFileError(Exception):
592 """Same File exception.
594 This exception will be thrown by FileDownloader objects if they detect
595 multiple files would have to be downloaded to the same file on disk.
600 class PostProcessingError(Exception):
601 """Post Processing exception.
603 This exception may be raised by PostProcessor's .run() method to
604 indicate an error in the postprocessing task.
607 def __init__(self, msg):
611 class MaxDownloadsReached(Exception):
612 """ --max-downloads limit has been reached. """
616 class UnavailableVideoError(Exception):
617 """Unavailable Format exception.
619 This exception will be thrown when a video is requested
620 in a format that is not available for that video.
625 class ContentTooShortError(Exception):
626 """Content Too Short exception.
628 This exception may be raised by FileDownloader objects when a file they
629 download is too small for what the server announced first, indicating
630 the connection was probably interrupted.
633 def __init__(self, downloaded, expected):
635 self.downloaded = downloaded
636 self.expected = expected
639 def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
640 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
641 # expected HTTP responses to meet HTTP/1.0 or later (see also
642 # https://github.com/rg3/youtube-dl/issues/6727)
643 if sys.version_info < (3, 0):
644 kwargs[b'strict'] = True
645 hc = http_class(*args, **kwargs)
646 source_address = ydl_handler._params.get('source_address')
647 if source_address is not None:
648 sa = (source_address, 0)
649 if hasattr(hc, 'source_address'): # Python 2.7+
650 hc.source_address = sa
652 def _hc_connect(self, *args, **kwargs):
653 sock = compat_socket_create_connection(
654 (self.host, self.port), self.timeout, sa)
656 self.sock = ssl.wrap_socket(
657 sock, self.key_file, self.cert_file,
658 ssl_version=ssl.PROTOCOL_TLSv1)
661 hc.connect = functools.partial(_hc_connect, hc)
666 def handle_youtubedl_headers(headers):
667 filtered_headers = headers
669 if 'Youtubedl-no-compression' in filtered_headers:
670 filtered_headers = dict((k, v) for k, v in filtered_headers.items() if k.lower() != 'accept-encoding')
671 del filtered_headers['Youtubedl-no-compression']
673 return filtered_headers
676 class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
677 """Handler for HTTP requests and responses.
679 This class, when installed with an OpenerDirector, automatically adds
680 the standard headers to every HTTP request and handles gzipped and
681 deflated responses from web servers. If compression is to be avoided in
682 a particular request, the original request in the program code only has
683 to include the HTTP header "Youtubedl-no-compression", which will be
684 removed before making the real request.
686 Part of this code was copied from:
688 http://techknack.net/python-urllib2-handlers/
690 Andrew Rowls, the author of that code, agreed to release it to the
694 def __init__(self, params, *args, **kwargs):
695 compat_urllib_request.HTTPHandler.__init__(self, *args, **kwargs)
696 self._params = params
698 def http_open(self, req):
699 return self.do_open(functools.partial(
700 _create_http_connection, self, compat_http_client.HTTPConnection, False),
706 return zlib.decompress(data, -zlib.MAX_WBITS)
708 return zlib.decompress(data)
711 def addinfourl_wrapper(stream, headers, url, code):
712 if hasattr(compat_urllib_request.addinfourl, 'getcode'):
713 return compat_urllib_request.addinfourl(stream, headers, url, code)
714 ret = compat_urllib_request.addinfourl(stream, headers, url)
718 def http_request(self, req):
719 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
720 # always respected by websites, some tend to give out URLs with non percent-encoded
721 # non-ASCII characters (see telemb.py, ard.py [#3412])
722 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
723 # To work around aforementioned issue we will replace request's original URL with
724 # percent-encoded one
725 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
726 # the code of this workaround has been moved here from YoutubeDL.urlopen()
727 url = req.get_full_url()
728 url_escaped = escape_url(url)
730 # Substitute URL if any change after escaping
731 if url != url_escaped:
732 req_type = HEADRequest if req.get_method() == 'HEAD' else compat_urllib_request.Request
734 url_escaped, data=req.data, headers=req.headers,
735 origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
736 new_req.timeout = req.timeout
739 for h, v in std_headers.items():
740 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
741 # The dict keys are capitalized because of this bug by urllib
742 if h.capitalize() not in req.headers:
745 req.headers = handle_youtubedl_headers(req.headers)
747 if sys.version_info < (2, 7) and '#' in req.get_full_url():
748 # Python 2.6 is brain-dead when it comes to fragments
749 req._Request__original = req._Request__original.partition('#')[0]
750 req._Request__r_type = req._Request__r_type.partition('#')[0]
754 def http_response(self, req, resp):
757 if resp.headers.get('Content-encoding', '') == 'gzip':
758 content = resp.read()
759 gz = gzip.GzipFile(fileobj=io.BytesIO(content), mode='rb')
761 uncompressed = io.BytesIO(gz.read())
762 except IOError as original_ioerror:
763 # There may be junk add the end of the file
764 # See http://stackoverflow.com/q/4928560/35070 for details
765 for i in range(1, 1024):
767 gz = gzip.GzipFile(fileobj=io.BytesIO(content[:-i]), mode='rb')
768 uncompressed = io.BytesIO(gz.read())
773 raise original_ioerror
774 resp = self.addinfourl_wrapper(uncompressed, old_resp.headers, old_resp.url, old_resp.code)
775 resp.msg = old_resp.msg
777 if resp.headers.get('Content-encoding', '') == 'deflate':
778 gz = io.BytesIO(self.deflate(resp.read()))
779 resp = self.addinfourl_wrapper(gz, old_resp.headers, old_resp.url, old_resp.code)
780 resp.msg = old_resp.msg
781 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
782 # https://github.com/rg3/youtube-dl/issues/6457).
783 if 300 <= resp.code < 400:
784 location = resp.headers.get('Location')
786 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
787 if sys.version_info >= (3, 0):
788 location = location.encode('iso-8859-1').decode('utf-8')
789 location_escaped = escape_url(location)
790 if location != location_escaped:
791 del resp.headers['Location']
792 resp.headers['Location'] = location_escaped
795 https_request = http_request
796 https_response = http_response
799 class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
800 def __init__(self, params, https_conn_class=None, *args, **kwargs):
801 compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)
802 self._https_conn_class = https_conn_class or compat_http_client.HTTPSConnection
803 self._params = params
805 def https_open(self, req):
807 if hasattr(self, '_context'): # python > 2.6
808 kwargs['context'] = self._context
809 if hasattr(self, '_check_hostname'): # python 3.x
810 kwargs['check_hostname'] = self._check_hostname
811 return self.do_open(functools.partial(
812 _create_http_connection, self, self._https_conn_class, True),
816 class YoutubeDLCookieProcessor(compat_urllib_request.HTTPCookieProcessor):
817 def __init__(self, cookiejar=None):
818 compat_urllib_request.HTTPCookieProcessor.__init__(self, cookiejar)
820 def http_response(self, request, response):
821 # Python 2 will choke on next HTTP request in row if there are non-ASCII
822 # characters in Set-Cookie HTTP header of last response (see
823 # https://github.com/rg3/youtube-dl/issues/6769).
824 # In order to at least prevent crashing we will percent encode Set-Cookie
825 # header before HTTPCookieProcessor starts processing it.
826 # if sys.version_info < (3, 0) and response.headers:
827 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
828 # set_cookie = response.headers.get(set_cookie_header)
830 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
831 # if set_cookie != set_cookie_escaped:
832 # del response.headers[set_cookie_header]
833 # response.headers[set_cookie_header] = set_cookie_escaped
834 return compat_urllib_request.HTTPCookieProcessor.http_response(self, request, response)
836 https_request = compat_urllib_request.HTTPCookieProcessor.http_request
837 https_response = http_response
840 def parse_iso8601(date_str, delimiter='T', timezone=None):
841 """ Return a UNIX timestamp from the given date """
846 date_str = re.sub(r'\.[0-9]+', '', date_str)
850 r'(?:Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
853 timezone = datetime.timedelta()
855 date_str = date_str[:-len(m.group(0))]
856 if not m.group('sign'):
857 timezone = datetime.timedelta()
859 sign = 1 if m.group('sign') == '+' else -1
860 timezone = datetime.timedelta(
861 hours=sign * int(m.group('hours')),
862 minutes=sign * int(m.group('minutes')))
864 date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
865 dt = datetime.datetime.strptime(date_str, date_format) - timezone
866 return calendar.timegm(dt.timetuple())
871 def unified_strdate(date_str, day_first=True):
872 """Return a string with the date in the format YYYYMMDD"""
878 date_str = date_str.replace(',', ' ')
879 # %z (UTC offset) is only supported in python>=3.2
880 if not re.match(r'^[0-9]{1,2}-[0-9]{1,2}-[0-9]{4}$', date_str):
881 date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
882 # Remove AM/PM + timezone
883 date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
885 format_expressions = [
890 '%b %dst %Y %I:%M%p',
891 '%b %dnd %Y %I:%M%p',
892 '%b %dth %Y %I:%M%p',
898 '%Y-%m-%d %H:%M:%S.%f',
901 '%Y-%m-%dT%H:%M:%SZ',
902 '%Y-%m-%dT%H:%M:%S.%fZ',
903 '%Y-%m-%dT%H:%M:%S.%f0Z',
905 '%Y-%m-%dT%H:%M:%S.%f',
909 format_expressions.extend([
917 format_expressions.extend([
924 for expression in format_expressions:
926 upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
929 if upload_date is None:
930 timetuple = email.utils.parsedate_tz(date_str)
932 upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
933 if upload_date is not None:
934 return compat_str(upload_date)
937 def determine_ext(url, default_ext='unknown_video'):
940 guess = url.partition('?')[0].rpartition('.')[2]
941 if re.match(r'^[A-Za-z0-9]+$', guess):
943 elif guess.rstrip('/') in (
944 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
945 'flv', 'f4v', 'f4a', 'f4b',
946 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
947 'mkv', 'mka', 'mk3d',
956 'f4f', 'f4m', 'm3u8', 'smil'):
957 return guess.rstrip('/')
962 def subtitles_filename(filename, sub_lang, sub_format):
963 return filename.rsplit('.', 1)[0] + '.' + sub_lang + '.' + sub_format
966 def date_from_str(date_str):
968 Return a datetime object from a string in the format YYYYMMDD or
969 (now|today)[+-][0-9](day|week|month|year)(s)?"""
970 today = datetime.date.today()
971 if date_str in ('now', 'today'):
973 if date_str == 'yesterday':
974 return today - datetime.timedelta(days=1)
975 match = re.match('(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
976 if match is not None:
977 sign = match.group('sign')
978 time = int(match.group('time'))
981 unit = match.group('unit')
982 # A bad aproximation?
990 delta = datetime.timedelta(**{unit: time})
992 return datetime.datetime.strptime(date_str, "%Y%m%d").date()
995 def hyphenate_date(date_str):
997 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
998 match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str)
999 if match is not None:
1000 return '-'.join(match.groups())
1005 class DateRange(object):
1006 """Represents a time interval between two dates"""
1008 def __init__(self, start=None, end=None):
1009 """start and end must be strings in the format accepted by date"""
1010 if start is not None:
1011 self.start = date_from_str(start)
1013 self.start = datetime.datetime.min.date()
1015 self.end = date_from_str(end)
1017 self.end = datetime.datetime.max.date()
1018 if self.start > self.end:
1019 raise ValueError('Date range: "%s" , the start date must be before the end date' % self)
1023 """Returns a range that only contains the given day"""
1024 return cls(day, day)
1026 def __contains__(self, date):
1027 """Check if the date is in the range"""
1028 if not isinstance(date, datetime.date):
1029 date = date_from_str(date)
1030 return self.start <= date <= self.end
1033 return '%s - %s' % (self.start.isoformat(), self.end.isoformat())
1036 def platform_name():
1037 """ Returns the platform name as a compat_str """
1038 res = platform.platform()
1039 if isinstance(res, bytes):
1040 res = res.decode(preferredencoding())
1042 assert isinstance(res, compat_str)
1046 def _windows_write_string(s, out):
1047 """ Returns True if the string was written using special methods,
1048 False if it has yet to be written out."""
1049 # Adapted from http://stackoverflow.com/a/3259271/35070
1052 import ctypes.wintypes
1060 fileno = out.fileno()
1061 except AttributeError:
1062 # If the output stream doesn't have a fileno, it's virtual
1064 except io.UnsupportedOperation:
1065 # Some strange Windows pseudo files?
1067 if fileno not in WIN_OUTPUT_IDS:
1070 GetStdHandle = ctypes.WINFUNCTYPE(
1071 ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
1072 (b"GetStdHandle", ctypes.windll.kernel32))
1073 h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
1075 WriteConsoleW = ctypes.WINFUNCTYPE(
1076 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
1077 ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
1078 ctypes.wintypes.LPVOID)((b"WriteConsoleW", ctypes.windll.kernel32))
1079 written = ctypes.wintypes.DWORD(0)
1081 GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b"GetFileType", ctypes.windll.kernel32))
1082 FILE_TYPE_CHAR = 0x0002
1083 FILE_TYPE_REMOTE = 0x8000
1084 GetConsoleMode = ctypes.WINFUNCTYPE(
1085 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
1086 ctypes.POINTER(ctypes.wintypes.DWORD))(
1087 (b"GetConsoleMode", ctypes.windll.kernel32))
1088 INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
1090 def not_a_console(handle):
1091 if handle == INVALID_HANDLE_VALUE or handle is None:
1093 return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR or
1094 GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0)
1096 if not_a_console(h):
1099 def next_nonbmp_pos(s):
1101 return next(i for i, c in enumerate(s) if ord(c) > 0xffff)
1102 except StopIteration:
1106 count = min(next_nonbmp_pos(s), 1024)
1108 ret = WriteConsoleW(
1109 h, s, count if count else 2, ctypes.byref(written), None)
1111 raise OSError('Failed to write string')
1112 if not count: # We just wrote a non-BMP character
1113 assert written.value == 2
1116 assert written.value > 0
1117 s = s[written.value:]
1121 def write_string(s, out=None, encoding=None):
1124 assert type(s) == compat_str
1126 if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'):
1127 if _windows_write_string(s, out):
1130 if ('b' in getattr(out, 'mode', '') or
1131 sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr
1132 byt = s.encode(encoding or preferredencoding(), 'ignore')
1134 elif hasattr(out, 'buffer'):
1135 enc = encoding or getattr(out, 'encoding', None) or preferredencoding()
1136 byt = s.encode(enc, 'ignore')
1137 out.buffer.write(byt)
1143 def bytes_to_intlist(bs):
1146 if isinstance(bs[0], int): # Python 3
1149 return [ord(c) for c in bs]
1152 def intlist_to_bytes(xs):
1155 return struct_pack('%dB' % len(xs), *xs)
1158 # Cross-platform file locking
1159 if sys.platform == 'win32':
1160 import ctypes.wintypes
1163 class OVERLAPPED(ctypes.Structure):
1165 ('Internal', ctypes.wintypes.LPVOID),
1166 ('InternalHigh', ctypes.wintypes.LPVOID),
1167 ('Offset', ctypes.wintypes.DWORD),
1168 ('OffsetHigh', ctypes.wintypes.DWORD),
1169 ('hEvent', ctypes.wintypes.HANDLE),
1172 kernel32 = ctypes.windll.kernel32
1173 LockFileEx = kernel32.LockFileEx
1174 LockFileEx.argtypes = [
1175 ctypes.wintypes.HANDLE, # hFile
1176 ctypes.wintypes.DWORD, # dwFlags
1177 ctypes.wintypes.DWORD, # dwReserved
1178 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1179 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1180 ctypes.POINTER(OVERLAPPED) # Overlapped
1182 LockFileEx.restype = ctypes.wintypes.BOOL
1183 UnlockFileEx = kernel32.UnlockFileEx
1184 UnlockFileEx.argtypes = [
1185 ctypes.wintypes.HANDLE, # hFile
1186 ctypes.wintypes.DWORD, # dwReserved
1187 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1188 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1189 ctypes.POINTER(OVERLAPPED) # Overlapped
1191 UnlockFileEx.restype = ctypes.wintypes.BOOL
1192 whole_low = 0xffffffff
1193 whole_high = 0x7fffffff
1195 def _lock_file(f, exclusive):
1196 overlapped = OVERLAPPED()
1197 overlapped.Offset = 0
1198 overlapped.OffsetHigh = 0
1199 overlapped.hEvent = 0
1200 f._lock_file_overlapped_p = ctypes.pointer(overlapped)
1201 handle = msvcrt.get_osfhandle(f.fileno())
1202 if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0,
1203 whole_low, whole_high, f._lock_file_overlapped_p):
1204 raise OSError('Locking file failed: %r' % ctypes.FormatError())
1206 def _unlock_file(f):
1207 assert f._lock_file_overlapped_p
1208 handle = msvcrt.get_osfhandle(f.fileno())
1209 if not UnlockFileEx(handle, 0,
1210 whole_low, whole_high, f._lock_file_overlapped_p):
1211 raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
1216 def _lock_file(f, exclusive):
1217 fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
1219 def _unlock_file(f):
1220 fcntl.flock(f, fcntl.LOCK_UN)
1223 class locked_file(object):
1224 def __init__(self, filename, mode, encoding=None):
1225 assert mode in ['r', 'a', 'w']
1226 self.f = io.open(filename, mode, encoding=encoding)
1229 def __enter__(self):
1230 exclusive = self.mode != 'r'
1232 _lock_file(self.f, exclusive)
1238 def __exit__(self, etype, value, traceback):
1240 _unlock_file(self.f)
1247 def write(self, *args):
1248 return self.f.write(*args)
1250 def read(self, *args):
1251 return self.f.read(*args)
1254 def get_filesystem_encoding():
1255 encoding = sys.getfilesystemencoding()
1256 return encoding if encoding is not None else 'utf-8'
1259 def shell_quote(args):
1261 encoding = get_filesystem_encoding()
1263 if isinstance(a, bytes):
1264 # We may get a filename encoded with 'encodeFilename'
1265 a = a.decode(encoding)
1266 quoted_args.append(pipes.quote(a))
1267 return ' '.join(quoted_args)
1270 def smuggle_url(url, data):
1271 """ Pass additional data in a URL for internal use. """
1273 sdata = compat_urllib_parse.urlencode(
1274 {'__youtubedl_smuggle': json.dumps(data)})
1275 return url + '#' + sdata
1278 def unsmuggle_url(smug_url, default=None):
1279 if '#__youtubedl_smuggle' not in smug_url:
1280 return smug_url, default
1281 url, _, sdata = smug_url.rpartition('#')
1282 jsond = compat_parse_qs(sdata)['__youtubedl_smuggle'][0]
1283 data = json.loads(jsond)
1287 def format_bytes(bytes):
1290 if type(bytes) is str:
1291 bytes = float(bytes)
1295 exponent = int(math.log(bytes, 1024.0))
1296 suffix = ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent]
1297 converted = float(bytes) / float(1024 ** exponent)
1298 return '%.2f%s' % (converted, suffix)
1301 def parse_filesize(s):
1305 # The lower-case forms are of course incorrect and inofficial,
1306 # but we support those too
1344 units_re = '|'.join(re.escape(u) for u in _UNIT_TABLE)
1346 r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)' % units_re, s)
1350 num_str = m.group('num').replace(',', '.')
1351 mult = _UNIT_TABLE[m.group('unit')]
1352 return int(float(num_str) * mult)
1355 def month_by_name(name):
1356 """ Return the number of a month by (locale-independently) English name """
1359 return ENGLISH_MONTH_NAMES.index(name) + 1
1364 def month_by_abbreviation(abbrev):
1365 """ Return the number of a month by (locale-independently) English
1369 return [s[:3] for s in ENGLISH_MONTH_NAMES].index(abbrev) + 1
1374 def fix_xml_ampersands(xml_str):
1375 """Replace all the '&' by '&' in XML"""
1377 r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1382 def setproctitle(title):
1383 assert isinstance(title, compat_str)
1385 libc = ctypes.cdll.LoadLibrary("libc.so.6")
1388 title_bytes = title.encode('utf-8')
1389 buf = ctypes.create_string_buffer(len(title_bytes))
1390 buf.value = title_bytes
1392 libc.prctl(15, buf, 0, 0, 0)
1393 except AttributeError:
1394 return # Strange libc, just skip this
1397 def remove_start(s, start):
1398 if s.startswith(start):
1399 return s[len(start):]
1403 def remove_end(s, end):
1405 return s[:-len(end)]
1409 def url_basename(url):
1410 path = compat_urlparse.urlparse(url).path
1411 return path.strip('/').split('/')[-1]
1414 class HEADRequest(compat_urllib_request.Request):
1415 def get_method(self):
1419 def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
1422 v = getattr(v, get_attr, None)
1428 return int(v) * invscale // scale
1433 def str_or_none(v, default=None):
1434 return default if v is None else compat_str(v)
1437 def str_to_int(int_str):
1438 """ A more relaxed version of int_or_none """
1441 int_str = re.sub(r'[,\.\+]', '', int_str)
1445 def float_or_none(v, scale=1, invscale=1, default=None):
1449 return float(v) * invscale / scale
1454 def parse_duration(s):
1455 if not isinstance(s, compat_basestring):
1463 (?P<only_mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*|
1464 (?P<only_hours>[0-9.]+)\s*(?:hours?)|
1466 \s*(?P<hours_reversed>[0-9]+)\s*(?:[:h]|hours?)\s*(?P<mins_reversed>[0-9]+)\s*(?:[:m]|mins?\.?|minutes?)\s*|
1469 (?:(?P<days>[0-9]+)\s*(?:[:d]|days?)\s*)?
1470 (?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*
1472 (?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*
1474 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?
1479 if m.group('only_mins'):
1480 return float_or_none(m.group('only_mins'), invscale=60)
1481 if m.group('only_hours'):
1482 return float_or_none(m.group('only_hours'), invscale=60 * 60)
1484 res += int(m.group('secs'))
1485 if m.group('mins_reversed'):
1486 res += int(m.group('mins_reversed')) * 60
1488 res += int(m.group('mins')) * 60
1489 if m.group('hours'):
1490 res += int(m.group('hours')) * 60 * 60
1491 if m.group('hours_reversed'):
1492 res += int(m.group('hours_reversed')) * 60 * 60
1494 res += int(m.group('days')) * 24 * 60 * 60
1496 res += float(m.group('ms'))
1500 def prepend_extension(filename, ext, expected_real_ext=None):
1501 name, real_ext = os.path.splitext(filename)
1503 '{0}.{1}{2}'.format(name, ext, real_ext)
1504 if not expected_real_ext or real_ext[1:] == expected_real_ext
1505 else '{0}.{1}'.format(filename, ext))
1508 def replace_extension(filename, ext, expected_real_ext=None):
1509 name, real_ext = os.path.splitext(filename)
1510 return '{0}.{1}'.format(
1511 name if not expected_real_ext or real_ext[1:] == expected_real_ext else filename,
1515 def check_executable(exe, args=[]):
1516 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
1517 args can be a list of arguments for a short output (like -version) """
1519 subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
1525 def get_exe_version(exe, args=['--version'],
1526 version_re=None, unrecognized='present'):
1527 """ Returns the version of the specified executable,
1528 or False if the executable is not present """
1530 out, _ = subprocess.Popen(
1531 [encodeArgument(exe)] + args,
1532 stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
1535 if isinstance(out, bytes): # Python 2.x
1536 out = out.decode('ascii', 'ignore')
1537 return detect_exe_version(out, version_re, unrecognized)
1540 def detect_exe_version(output, version_re=None, unrecognized='present'):
1541 assert isinstance(output, compat_str)
1542 if version_re is None:
1543 version_re = r'version\s+([-0-9._a-zA-Z]+)'
1544 m = re.search(version_re, output)
1551 class PagedList(object):
1553 # This is only useful for tests
1554 return len(self.getslice())
1557 class OnDemandPagedList(PagedList):
1558 def __init__(self, pagefunc, pagesize):
1559 self._pagefunc = pagefunc
1560 self._pagesize = pagesize
1562 def getslice(self, start=0, end=None):
1564 for pagenum in itertools.count(start // self._pagesize):
1565 firstid = pagenum * self._pagesize
1566 nextfirstid = pagenum * self._pagesize + self._pagesize
1567 if start >= nextfirstid:
1570 page_results = list(self._pagefunc(pagenum))
1573 start % self._pagesize
1574 if firstid <= start < nextfirstid
1578 ((end - 1) % self._pagesize) + 1
1579 if (end is not None and firstid <= end <= nextfirstid)
1582 if startv != 0 or endv is not None:
1583 page_results = page_results[startv:endv]
1584 res.extend(page_results)
1586 # A little optimization - if current page is not "full", ie. does
1587 # not contain page_size videos then we can assume that this page
1588 # is the last one - there are no more ids on further pages -
1589 # i.e. no need to query again.
1590 if len(page_results) + startv < self._pagesize:
1593 # If we got the whole page, but the next page is not interesting,
1594 # break out early as well
1595 if end == nextfirstid:
1600 class InAdvancePagedList(PagedList):
1601 def __init__(self, pagefunc, pagecount, pagesize):
1602 self._pagefunc = pagefunc
1603 self._pagecount = pagecount
1604 self._pagesize = pagesize
1606 def getslice(self, start=0, end=None):
1608 start_page = start // self._pagesize
1610 self._pagecount if end is None else (end // self._pagesize + 1))
1611 skip_elems = start - start_page * self._pagesize
1612 only_more = None if end is None else end - start
1613 for pagenum in range(start_page, end_page):
1614 page = list(self._pagefunc(pagenum))
1616 page = page[skip_elems:]
1618 if only_more is not None:
1619 if len(page) < only_more:
1620 only_more -= len(page)
1622 page = page[:only_more]
1629 def uppercase_escape(s):
1630 unicode_escape = codecs.getdecoder('unicode_escape')
1632 r'\\U[0-9a-fA-F]{8}',
1633 lambda m: unicode_escape(m.group(0))[0],
1637 def lowercase_escape(s):
1638 unicode_escape = codecs.getdecoder('unicode_escape')
1640 r'\\u[0-9a-fA-F]{4}',
1641 lambda m: unicode_escape(m.group(0))[0],
1645 def escape_rfc3986(s):
1646 """Escape non-ASCII characters as suggested by RFC 3986"""
1647 if sys.version_info < (3, 0) and isinstance(s, compat_str):
1648 s = s.encode('utf-8')
1649 return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]")
1652 def escape_url(url):
1653 """Escape URL as suggested by RFC 3986"""
1654 url_parsed = compat_urllib_parse_urlparse(url)
1655 return url_parsed._replace(
1656 path=escape_rfc3986(url_parsed.path),
1657 params=escape_rfc3986(url_parsed.params),
1658 query=escape_rfc3986(url_parsed.query),
1659 fragment=escape_rfc3986(url_parsed.fragment)
1663 struct.pack('!I', 0)
1665 # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
1666 def struct_pack(spec, *args):
1667 if isinstance(spec, compat_str):
1668 spec = spec.encode('ascii')
1669 return struct.pack(spec, *args)
1671 def struct_unpack(spec, *args):
1672 if isinstance(spec, compat_str):
1673 spec = spec.encode('ascii')
1674 return struct.unpack(spec, *args)
1676 struct_pack = struct.pack
1677 struct_unpack = struct.unpack
1680 def read_batch_urls(batch_fd):
1682 if not isinstance(url, compat_str):
1683 url = url.decode('utf-8', 'replace')
1684 BOM_UTF8 = '\xef\xbb\xbf'
1685 if url.startswith(BOM_UTF8):
1686 url = url[len(BOM_UTF8):]
1688 if url.startswith(('#', ';', ']')):
1692 with contextlib.closing(batch_fd) as fd:
1693 return [url for url in map(fixup, fd) if url]
1696 def urlencode_postdata(*args, **kargs):
1697 return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
1700 def encode_dict(d, encoding='utf-8'):
1702 return v.encode(encoding) if isinstance(v, compat_basestring) else v
1703 return dict((encode(k), encode(v)) for k, v in d.items())
1715 def parse_age_limit(s):
1718 m = re.match(r'^(?P<age>\d{1,2})\+?$', s)
1719 return int(m.group('age')) if m else US_RATINGS.get(s, None)
1722 def strip_jsonp(code):
1724 r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code)
1727 def js_to_json(code):
1730 if v in ('true', 'false', 'null'):
1732 if v.startswith('"'):
1733 v = re.sub(r"\\'", "'", v[1:-1])
1734 elif v.startswith("'"):
1736 v = re.sub(r"\\\\|\\'|\"", lambda m: {
1743 res = re.sub(r'''(?x)
1744 "(?:[^"\\]*(?:\\\\|\\['"nu]))*[^"\\]*"|
1745 '(?:[^'\\]*(?:\\\\|\\['"nu]))*[^'\\]*'|
1746 [a-zA-Z_][.a-zA-Z_0-9]*
1748 res = re.sub(r',(\s*[\]}])', lambda m: m.group(1), res)
1752 def qualities(quality_ids):
1753 """ Get a numeric quality value out of a list of possible values """
1756 return quality_ids.index(qid)
1762 DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s'
1765 def limit_length(s, length):
1766 """ Add ellipses to overly long strings """
1771 return s[:length - len(ELLIPSES)] + ELLIPSES
1775 def version_tuple(v):
1776 return tuple(int(e) for e in re.split(r'[-.]', v))
1779 def is_outdated_version(version, limit, assume_new=True):
1781 return not assume_new
1783 return version_tuple(version) < version_tuple(limit)
1785 return not assume_new
1788 def ytdl_is_updateable():
1789 """ Returns if youtube-dl can be updated with -U """
1790 from zipimport import zipimporter
1792 return isinstance(globals().get('__loader__'), zipimporter) or hasattr(sys, 'frozen')
1795 def args_to_str(args):
1796 # Get a short string representation for a subprocess command
1797 return ' '.join(shlex_quote(a) for a in args)
1800 def mimetype2ext(mt):
1801 _, _, res = mt.rpartition('/')
1805 'x-mp4-fragmented': 'mp4',
1810 def urlhandle_detect_ext(url_handle):
1813 getheader = lambda h: url_handle.headers[h]
1814 except AttributeError: # Python < 3
1815 getheader = url_handle.info().getheader
1817 cd = getheader('Content-Disposition')
1819 m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd)
1821 e = determine_ext(m.group('filename'), default_ext=None)
1825 return mimetype2ext(getheader('Content-Type'))
1828 def encode_data_uri(data, mime_type):
1829 return 'data:%s;base64,%s' % (mime_type, base64.b64encode(data).decode('ascii'))
1832 def age_restricted(content_limit, age_limit):
1833 """ Returns True iff the content should be blocked """
1835 if age_limit is None: # No limit set
1837 if content_limit is None:
1838 return False # Content available for everyone
1839 return age_limit < content_limit
1842 def is_html(first_bytes):
1843 """ Detect whether a file contains HTML by examining its first bytes. """
1846 (b'\xef\xbb\xbf', 'utf-8'),
1847 (b'\x00\x00\xfe\xff', 'utf-32-be'),
1848 (b'\xff\xfe\x00\x00', 'utf-32-le'),
1849 (b'\xff\xfe', 'utf-16-le'),
1850 (b'\xfe\xff', 'utf-16-be'),
1852 for bom, enc in BOMS:
1853 if first_bytes.startswith(bom):
1854 s = first_bytes[len(bom):].decode(enc, 'replace')
1857 s = first_bytes.decode('utf-8', 'replace')
1859 return re.match(r'^\s*<', s)
1862 def determine_protocol(info_dict):
1863 protocol = info_dict.get('protocol')
1864 if protocol is not None:
1867 url = info_dict['url']
1868 if url.startswith('rtmp'):
1870 elif url.startswith('mms'):
1872 elif url.startswith('rtsp'):
1875 ext = determine_ext(url)
1881 return compat_urllib_parse_urlparse(url).scheme
1884 def render_table(header_row, data):
1885 """ Render a list of rows, each as a list of values """
1886 table = [header_row] + data
1887 max_lens = [max(len(compat_str(v)) for v in col) for col in zip(*table)]
1888 format_str = ' '.join('%-' + compat_str(ml + 1) + 's' for ml in max_lens[:-1]) + '%s'
1889 return '\n'.join(format_str % tuple(row) for row in table)
1892 def _match_one(filter_part, dct):
1893 COMPARISON_OPERATORS = {
1901 operator_rex = re.compile(r'''(?x)\s*
1903 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1905 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
1906 (?P<strval>(?![0-9.])[a-z0-9A-Z]*)
1909 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
1910 m = operator_rex.search(filter_part)
1912 op = COMPARISON_OPERATORS[m.group('op')]
1913 if m.group('strval') is not None:
1914 if m.group('op') not in ('=', '!='):
1916 'Operator %s does not support string values!' % m.group('op'))
1917 comparison_value = m.group('strval')
1920 comparison_value = int(m.group('intval'))
1922 comparison_value = parse_filesize(m.group('intval'))
1923 if comparison_value is None:
1924 comparison_value = parse_filesize(m.group('intval') + 'B')
1925 if comparison_value is None:
1927 'Invalid integer value %r in filter part %r' % (
1928 m.group('intval'), filter_part))
1929 actual_value = dct.get(m.group('key'))
1930 if actual_value is None:
1931 return m.group('none_inclusive')
1932 return op(actual_value, comparison_value)
1935 '': lambda v: v is not None,
1936 '!': lambda v: v is None,
1938 operator_rex = re.compile(r'''(?x)\s*
1939 (?P<op>%s)\s*(?P<key>[a-z_]+)
1941 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
1942 m = operator_rex.search(filter_part)
1944 op = UNARY_OPERATORS[m.group('op')]
1945 actual_value = dct.get(m.group('key'))
1946 return op(actual_value)
1948 raise ValueError('Invalid filter part %r' % filter_part)
1951 def match_str(filter_str, dct):
1952 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """
1955 _match_one(filter_part, dct) for filter_part in filter_str.split('&'))
1958 def match_filter_func(filter_str):
1959 def _match_func(info_dict):
1960 if match_str(filter_str, info_dict):
1963 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1964 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
1968 def parse_dfxp_time_expr(time_expr):
1972 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
1974 return float(mobj.group('time_offset'))
1976 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:\.\d+)?)$', time_expr)
1978 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3))
1981 def srt_subtitles_timecode(seconds):
1982 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
1985 def dfxp2srt(dfxp_data):
1986 _x = functools.partial(xpath_with_ns, ns_map={
1987 'ttml': 'http://www.w3.org/ns/ttml',
1988 'ttaf1': 'http://www.w3.org/2006/10/ttaf1',
1991 def parse_node(node):
1992 str_or_empty = functools.partial(str_or_none, default='')
1994 out = str_or_empty(node.text)
1997 if child.tag in (_x('ttml:br'), _x('ttaf1:br'), 'br'):
1998 out += '\n' + str_or_empty(child.tail)
1999 elif child.tag in (_x('ttml:span'), _x('ttaf1:span'), 'span'):
2000 out += str_or_empty(parse_node(child))
2002 out += str_or_empty(xml.etree.ElementTree.tostring(child))
2006 dfxp = compat_etree_fromstring(dfxp_data.encode('utf-8'))
2008 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall('.//p')
2011 raise ValueError('Invalid dfxp/TTML subtitle')
2013 for para, index in zip(paras, itertools.count(1)):
2014 begin_time = parse_dfxp_time_expr(para.attrib['begin'])
2015 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
2017 end_time = begin_time + parse_dfxp_time_expr(para.attrib['dur'])
2018 out.append('%d\n%s --> %s\n%s\n\n' % (
2020 srt_subtitles_timecode(begin_time),
2021 srt_subtitles_timecode(end_time),
2027 def cli_option(params, command_option, param):
2028 param = params.get(param)
2029 return [command_option, param] if param is not None else []
2032 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
2033 param = params.get(param)
2034 assert isinstance(param, bool)
2036 return [command_option + separator + (true_value if param else false_value)]
2037 return [command_option, true_value if param else false_value]
2040 def cli_valueless_option(params, command_option, param, expected_value=True):
2041 param = params.get(param)
2042 return [command_option] if param == expected_value else []
2045 def cli_configuration_args(params, param, default=[]):
2046 ex_args = params.get(param)
2049 assert isinstance(ex_args, list)
2053 class ISO639Utils(object):
2054 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
2243 def short2long(cls, code):
2244 """Convert language code from ISO 639-1 to ISO 639-2/T"""
2245 return cls._lang_map.get(code[:2])
2248 def long2short(cls, code):
2249 """Convert language code from ISO 639-2/T to ISO 639-1"""
2250 for short_name, long_name in cls._lang_map.items():
2251 if long_name == code:
2255 class ISO3166Utils(object):
2256 # From http://data.okfn.org/data/core/country-list
2258 'AF': 'Afghanistan',
2259 'AX': 'Ã…land Islands',
2262 'AS': 'American Samoa',
2267 'AG': 'Antigua and Barbuda',
2284 'BO': 'Bolivia, Plurinational State of',
2285 'BQ': 'Bonaire, Sint Eustatius and Saba',
2286 'BA': 'Bosnia and Herzegovina',
2288 'BV': 'Bouvet Island',
2290 'IO': 'British Indian Ocean Territory',
2291 'BN': 'Brunei Darussalam',
2293 'BF': 'Burkina Faso',
2299 'KY': 'Cayman Islands',
2300 'CF': 'Central African Republic',
2304 'CX': 'Christmas Island',
2305 'CC': 'Cocos (Keeling) Islands',
2309 'CD': 'Congo, the Democratic Republic of the',
2310 'CK': 'Cook Islands',
2312 'CI': 'Côte d\'Ivoire',
2317 'CZ': 'Czech Republic',
2321 'DO': 'Dominican Republic',
2324 'SV': 'El Salvador',
2325 'GQ': 'Equatorial Guinea',
2329 'FK': 'Falkland Islands (Malvinas)',
2330 'FO': 'Faroe Islands',
2334 'GF': 'French Guiana',
2335 'PF': 'French Polynesia',
2336 'TF': 'French Southern Territories',
2351 'GW': 'Guinea-Bissau',
2354 'HM': 'Heard Island and McDonald Islands',
2355 'VA': 'Holy See (Vatican City State)',
2362 'IR': 'Iran, Islamic Republic of',
2365 'IM': 'Isle of Man',
2375 'KP': 'Korea, Democratic People\'s Republic of',
2376 'KR': 'Korea, Republic of',
2379 'LA': 'Lao People\'s Democratic Republic',
2385 'LI': 'Liechtenstein',
2389 'MK': 'Macedonia, the Former Yugoslav Republic of',
2396 'MH': 'Marshall Islands',
2402 'FM': 'Micronesia, Federated States of',
2403 'MD': 'Moldova, Republic of',
2414 'NL': 'Netherlands',
2415 'NC': 'New Caledonia',
2416 'NZ': 'New Zealand',
2421 'NF': 'Norfolk Island',
2422 'MP': 'Northern Mariana Islands',
2427 'PS': 'Palestine, State of',
2429 'PG': 'Papua New Guinea',
2432 'PH': 'Philippines',
2436 'PR': 'Puerto Rico',
2440 'RU': 'Russian Federation',
2442 'BL': 'Saint Barthélemy',
2443 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
2444 'KN': 'Saint Kitts and Nevis',
2445 'LC': 'Saint Lucia',
2446 'MF': 'Saint Martin (French part)',
2447 'PM': 'Saint Pierre and Miquelon',
2448 'VC': 'Saint Vincent and the Grenadines',
2451 'ST': 'Sao Tome and Principe',
2452 'SA': 'Saudi Arabia',
2456 'SL': 'Sierra Leone',
2458 'SX': 'Sint Maarten (Dutch part)',
2461 'SB': 'Solomon Islands',
2463 'ZA': 'South Africa',
2464 'GS': 'South Georgia and the South Sandwich Islands',
2465 'SS': 'South Sudan',
2470 'SJ': 'Svalbard and Jan Mayen',
2473 'CH': 'Switzerland',
2474 'SY': 'Syrian Arab Republic',
2475 'TW': 'Taiwan, Province of China',
2477 'TZ': 'Tanzania, United Republic of',
2479 'TL': 'Timor-Leste',
2483 'TT': 'Trinidad and Tobago',
2486 'TM': 'Turkmenistan',
2487 'TC': 'Turks and Caicos Islands',
2491 'AE': 'United Arab Emirates',
2492 'GB': 'United Kingdom',
2493 'US': 'United States',
2494 'UM': 'United States Minor Outlying Islands',
2498 'VE': 'Venezuela, Bolivarian Republic of',
2500 'VG': 'Virgin Islands, British',
2501 'VI': 'Virgin Islands, U.S.',
2502 'WF': 'Wallis and Futuna',
2503 'EH': 'Western Sahara',
2510 def short2full(cls, code):
2511 """Convert an ISO 3166-2 country code to the corresponding full name"""
2512 return cls._country_map.get(code.upper())
2515 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
2516 def __init__(self, proxies=None):
2517 # Set default handlers
2518 for type in ('http', 'https'):
2519 setattr(self, '%s_open' % type,
2520 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
2521 meth(r, proxy, type))
2522 return compat_urllib_request.ProxyHandler.__init__(self, proxies)
2524 def proxy_open(self, req, proxy, type):
2525 req_proxy = req.headers.get('Ytdl-request-proxy')
2526 if req_proxy is not None:
2528 del req.headers['Ytdl-request-proxy']
2530 if proxy == '__noproxy__':
2531 return None # No Proxy
2532 return compat_urllib_request.ProxyHandler.proxy_open(
2533 self, req, proxy, type)