2 # -*- coding: utf-8 -*-
4 from __future__ import unicode_literals
32 import xml.etree.ElementTree
42 compat_socket_create_connection,
46 compat_urllib_parse_urlparse,
47 compat_urllib_request,
53 # This is not clearly defined otherwise
54 compiled_regex_type = type(re.compile(''))
57 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)',
58 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
59 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
60 'Accept-Encoding': 'gzip, deflate',
61 'Accept-Language': 'en-us,en;q=0.5',
67 ENGLISH_MONTH_NAMES = [
68 'January', 'February', 'March', 'April', 'May', 'June',
69 'July', 'August', 'September', 'October', 'November', 'December']
72 def preferredencoding():
73 """Get preferred encoding.
75 Returns the best encoding scheme for the system, based on
76 locale.getpreferredencoding() and some further tweaks.
79 pref = locale.getpreferredencoding()
87 def write_json_file(obj, fn):
88 """ Encode obj as JSON and write it to fn, atomically if possible """
90 fn = encodeFilename(fn)
91 if sys.version_info < (3, 0) and sys.platform != 'win32':
92 encoding = get_filesystem_encoding()
93 # os.path.basename returns a bytes object, but NamedTemporaryFile
94 # will fail if the filename contains non ascii characters unless we
95 # use a unicode object
96 path_basename = lambda f: os.path.basename(fn).decode(encoding)
97 # the same for os.path.dirname
98 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
100 path_basename = os.path.basename
101 path_dirname = os.path.dirname
105 'prefix': path_basename(fn) + '.',
106 'dir': path_dirname(fn),
110 # In Python 2.x, json.dump expects a bytestream.
111 # In Python 3.x, it writes to a character stream
112 if sys.version_info < (3, 0):
120 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
125 if sys.platform == 'win32':
126 # Need to remove existing file on Windows, else os.rename raises
127 # WindowsError or FileExistsError.
132 os.rename(tf.name, fn)
141 if sys.version_info >= (2, 7):
142 def find_xpath_attr(node, xpath, key, val=None):
143 """ Find the xpath xpath[@key=val] """
144 assert re.match(r'^[a-zA-Z_-]+$', key)
146 assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val)
147 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
148 return node.find(expr)
150 def find_xpath_attr(node, xpath, key, val=None):
151 # Here comes the crazy part: In 2.6, if the xpath is a unicode,
152 # .//node does not match if a node is a direct child of . !
153 if isinstance(xpath, compat_str):
154 xpath = xpath.encode('ascii')
156 for f in node.findall(xpath):
157 if key not in f.attrib:
159 if val is None or f.attrib.get(key) == val:
163 # On python2.6 the xml.etree.ElementTree.Element methods don't support
164 # the namespace parameter
167 def xpath_with_ns(path, ns_map):
168 components = [c.split(':') for c in path.split('/')]
172 replaced.append(c[0])
175 replaced.append('{%s}%s' % (ns_map[ns], tag))
176 return '/'.join(replaced)
179 def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
180 if sys.version_info < (2, 7): # Crazy 2.6
181 xpath = xpath.encode('ascii')
185 if default is not NO_DEFAULT:
188 name = xpath if name is None else name
189 raise ExtractorError('Could not find XML element %s' % name)
195 def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
196 n = xpath_element(node, xpath, name, fatal=fatal, default=default)
197 if n is None or n == default:
200 if default is not NO_DEFAULT:
203 name = xpath if name is None else name
204 raise ExtractorError('Could not find XML element\'s text %s' % name)
210 def xpath_attr(node, xpath, key, name=None, fatal=False, default=NO_DEFAULT):
211 n = find_xpath_attr(node, xpath, key)
213 if default is not NO_DEFAULT:
216 name = '%s[@%s]' % (xpath, key) if name is None else name
217 raise ExtractorError('Could not find XML attribute %s' % name)
223 def get_element_by_id(id, html):
224 """Return the content of the tag with the specified ID in the passed HTML document"""
225 return get_element_by_attribute("id", id, html)
228 def get_element_by_attribute(attribute, value, html):
229 """Return the content of the tag with the specified attribute in the passed HTML document"""
231 m = re.search(r'''(?xs)
233 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
235 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
239 ''' % (re.escape(attribute), re.escape(value)), html)
243 res = m.group('content')
245 if res.startswith('"') or res.startswith("'"):
248 return unescapeHTML(res)
251 def clean_html(html):
252 """Clean an HTML snippet into a readable string"""
254 if html is None: # Convenience for sanitizing descriptions etc.
258 html = html.replace('\n', ' ')
259 html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html)
260 html = re.sub(r'<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
262 html = re.sub('<.*?>', '', html)
263 # Replace html entities
264 html = unescapeHTML(html)
268 def sanitize_open(filename, open_mode):
269 """Try to open the given filename, and slightly tweak it if this fails.
271 Attempts to open the given filename. If this fails, it tries to change
272 the filename slightly, step by step, until it's either able to open it
273 or it fails and raises a final exception, like the standard open()
276 It returns the tuple (stream, definitive_file_name).
280 if sys.platform == 'win32':
282 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
283 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
284 stream = open(encodeFilename(filename), open_mode)
285 return (stream, filename)
286 except (IOError, OSError) as err:
287 if err.errno in (errno.EACCES,):
290 # In case of error, try to remove win32 forbidden chars
291 alt_filename = sanitize_path(filename)
292 if alt_filename == filename:
295 # An exception here should be caught in the caller
296 stream = open(encodeFilename(alt_filename), open_mode)
297 return (stream, alt_filename)
300 def timeconvert(timestr):
301 """Convert RFC 2822 defined time string into system timestamp"""
303 timetuple = email.utils.parsedate_tz(timestr)
304 if timetuple is not None:
305 timestamp = email.utils.mktime_tz(timetuple)
309 def sanitize_filename(s, restricted=False, is_id=False):
310 """Sanitizes a string so it could be used as part of a filename.
311 If restricted is set, use a stricter subset of allowed characters.
312 Set is_id if this is not an arbitrary string, but an ID that should be kept if possible
314 def replace_insane(char):
315 if char == '?' or ord(char) < 32 or ord(char) == 127:
318 return '' if restricted else '\''
320 return '_-' if restricted else ' -'
321 elif char in '\\/|*<>':
323 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
325 if restricted and ord(char) > 127:
330 s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
331 result = ''.join(map(replace_insane, s))
333 while '__' in result:
334 result = result.replace('__', '_')
335 result = result.strip('_')
336 # Common case of "Foreign band name - English song title"
337 if restricted and result.startswith('-_'):
339 if result.startswith('-'):
340 result = '_' + result[len('-'):]
341 result = result.lstrip('.')
347 def sanitize_path(s):
348 """Sanitizes and normalizes path on Windows"""
349 if sys.platform != 'win32':
351 drive_or_unc, _ = os.path.splitdrive(s)
352 if sys.version_info < (2, 7) and not drive_or_unc:
353 drive_or_unc, _ = os.path.splitunc(s)
354 norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
358 path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|\.$)', '#', path_part)
359 for path_part in norm_path]
361 sanitized_path.insert(0, drive_or_unc + os.path.sep)
362 return os.path.join(*sanitized_path)
365 def orderedSet(iterable):
366 """ Remove all duplicates from the input iterable """
374 def _htmlentity_transform(entity):
375 """Transforms an HTML entity to a character."""
376 # Known non-numeric HTML entity
377 if entity in compat_html_entities.name2codepoint:
378 return compat_chr(compat_html_entities.name2codepoint[entity])
380 mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
382 numstr = mobj.group(1)
383 if numstr.startswith('x'):
385 numstr = '0%s' % numstr
388 return compat_chr(int(numstr, base))
390 # Unknown entity in name, return its literal representation
391 return ('&%s;' % entity)
397 assert type(s) == compat_str
400 r'&([^;]+);', lambda m: _htmlentity_transform(m.group(1)), s)
403 def get_subprocess_encoding():
404 if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
405 # For subprocess calls, encode with locale encoding
406 # Refer to http://stackoverflow.com/a/9951851/35070
407 encoding = preferredencoding()
409 encoding = sys.getfilesystemencoding()
415 def encodeFilename(s, for_subprocess=False):
417 @param s The name of the file
420 assert type(s) == compat_str
422 # Python 3 has a Unicode API
423 if sys.version_info >= (3, 0):
426 # Pass '' directly to use Unicode APIs on Windows 2000 and up
427 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
428 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
429 if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
432 return s.encode(get_subprocess_encoding(), 'ignore')
435 def decodeFilename(b, for_subprocess=False):
437 if sys.version_info >= (3, 0):
440 if not isinstance(b, bytes):
443 return b.decode(get_subprocess_encoding(), 'ignore')
446 def encodeArgument(s):
447 if not isinstance(s, compat_str):
448 # Legacy code that uses byte strings
449 # Uncomment the following line after fixing all post processors
450 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
451 s = s.decode('ascii')
452 return encodeFilename(s, True)
455 def decodeArgument(b):
456 return decodeFilename(b, True)
459 def decodeOption(optval):
462 if isinstance(optval, bytes):
463 optval = optval.decode(preferredencoding())
465 assert isinstance(optval, compat_str)
469 def formatSeconds(secs):
471 return '%d:%02d:%02d' % (secs // 3600, (secs % 3600) // 60, secs % 60)
473 return '%d:%02d' % (secs // 60, secs % 60)
478 def make_HTTPS_handler(params, **kwargs):
479 opts_no_check_certificate = params.get('nocheckcertificate', False)
480 if hasattr(ssl, 'create_default_context'): # Python >= 3.4 or 2.7.9
481 context = ssl.create_default_context(ssl.Purpose.SERVER_AUTH)
482 if opts_no_check_certificate:
483 context.check_hostname = False
484 context.verify_mode = ssl.CERT_NONE
486 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
489 # (create_default_context present but HTTPSHandler has no context=)
492 if sys.version_info < (3, 2):
493 return YoutubeDLHTTPSHandler(params, **kwargs)
495 context = ssl.SSLContext(ssl.PROTOCOL_TLSv1)
496 context.verify_mode = (ssl.CERT_NONE
497 if opts_no_check_certificate
498 else ssl.CERT_REQUIRED)
499 context.set_default_verify_paths()
500 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
503 def bug_reports_message():
504 if ytdl_is_updateable():
505 update_cmd = 'type youtube-dl -U to update'
507 update_cmd = 'see https://yt-dl.org/update on how to update'
508 msg = '; please report this issue on https://yt-dl.org/bug .'
509 msg += ' Make sure you are using the latest version; %s.' % update_cmd
510 msg += ' Be sure to call youtube-dl with the --verbose flag and include its complete output.'
514 class ExtractorError(Exception):
515 """Error during info extraction."""
517 def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):
518 """ tb, if given, is the original traceback (so that it can be printed out).
519 If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
522 if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError):
524 if video_id is not None:
525 msg = video_id + ': ' + msg
527 msg += ' (caused by %r)' % cause
529 msg += bug_reports_message()
530 super(ExtractorError, self).__init__(msg)
533 self.exc_info = sys.exc_info() # preserve original exception
535 self.video_id = video_id
537 def format_traceback(self):
538 if self.traceback is None:
540 return ''.join(traceback.format_tb(self.traceback))
543 class UnsupportedError(ExtractorError):
544 def __init__(self, url):
545 super(UnsupportedError, self).__init__(
546 'Unsupported URL: %s' % url, expected=True)
550 class RegexNotFoundError(ExtractorError):
551 """Error when a regex didn't match"""
555 class DownloadError(Exception):
556 """Download Error exception.
558 This exception may be thrown by FileDownloader objects if they are not
559 configured to continue on errors. They will contain the appropriate
563 def __init__(self, msg, exc_info=None):
564 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
565 super(DownloadError, self).__init__(msg)
566 self.exc_info = exc_info
569 class SameFileError(Exception):
570 """Same File exception.
572 This exception will be thrown by FileDownloader objects if they detect
573 multiple files would have to be downloaded to the same file on disk.
578 class PostProcessingError(Exception):
579 """Post Processing exception.
581 This exception may be raised by PostProcessor's .run() method to
582 indicate an error in the postprocessing task.
585 def __init__(self, msg):
589 class MaxDownloadsReached(Exception):
590 """ --max-downloads limit has been reached. """
594 class UnavailableVideoError(Exception):
595 """Unavailable Format exception.
597 This exception will be thrown when a video is requested
598 in a format that is not available for that video.
603 class ContentTooShortError(Exception):
604 """Content Too Short exception.
606 This exception may be raised by FileDownloader objects when a file they
607 download is too small for what the server announced first, indicating
608 the connection was probably interrupted.
611 def __init__(self, downloaded, expected):
613 self.downloaded = downloaded
614 self.expected = expected
617 def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
618 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
619 # expected HTTP responses to meet HTTP/1.0 or later (see also
620 # https://github.com/rg3/youtube-dl/issues/6727)
621 if sys.version_info < (3, 0):
622 kwargs['strict'] = True
623 hc = http_class(*args, **kwargs)
624 source_address = ydl_handler._params.get('source_address')
625 if source_address is not None:
626 sa = (source_address, 0)
627 if hasattr(hc, 'source_address'): # Python 2.7+
628 hc.source_address = sa
630 def _hc_connect(self, *args, **kwargs):
631 sock = compat_socket_create_connection(
632 (self.host, self.port), self.timeout, sa)
634 self.sock = ssl.wrap_socket(
635 sock, self.key_file, self.cert_file,
636 ssl_version=ssl.PROTOCOL_TLSv1)
639 hc.connect = functools.partial(_hc_connect, hc)
644 class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
645 """Handler for HTTP requests and responses.
647 This class, when installed with an OpenerDirector, automatically adds
648 the standard headers to every HTTP request and handles gzipped and
649 deflated responses from web servers. If compression is to be avoided in
650 a particular request, the original request in the program code only has
651 to include the HTTP header "Youtubedl-No-Compression", which will be
652 removed before making the real request.
654 Part of this code was copied from:
656 http://techknack.net/python-urllib2-handlers/
658 Andrew Rowls, the author of that code, agreed to release it to the
662 def __init__(self, params, *args, **kwargs):
663 compat_urllib_request.HTTPHandler.__init__(self, *args, **kwargs)
664 self._params = params
666 def http_open(self, req):
667 return self.do_open(functools.partial(
668 _create_http_connection, self, compat_http_client.HTTPConnection, False),
674 return zlib.decompress(data, -zlib.MAX_WBITS)
676 return zlib.decompress(data)
679 def addinfourl_wrapper(stream, headers, url, code):
680 if hasattr(compat_urllib_request.addinfourl, 'getcode'):
681 return compat_urllib_request.addinfourl(stream, headers, url, code)
682 ret = compat_urllib_request.addinfourl(stream, headers, url)
686 def http_request(self, req):
687 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
688 # always respected by websites, some tend to give out URLs with non percent-encoded
689 # non-ASCII characters (see telemb.py, ard.py [#3412])
690 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
691 # To work around aforementioned issue we will replace request's original URL with
692 # percent-encoded one
693 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
694 # the code of this workaround has been moved here from YoutubeDL.urlopen()
695 url = req.get_full_url()
696 url_escaped = escape_url(url)
698 # Substitute URL if any change after escaping
699 if url != url_escaped:
700 req_type = HEADRequest if req.get_method() == 'HEAD' else compat_urllib_request.Request
702 url_escaped, data=req.data, headers=req.headers,
703 origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
704 new_req.timeout = req.timeout
707 for h, v in std_headers.items():
708 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
709 # The dict keys are capitalized because of this bug by urllib
710 if h.capitalize() not in req.headers:
712 if 'Youtubedl-no-compression' in req.headers:
713 if 'Accept-encoding' in req.headers:
714 del req.headers['Accept-encoding']
715 del req.headers['Youtubedl-no-compression']
717 if sys.version_info < (2, 7) and '#' in req.get_full_url():
718 # Python 2.6 is brain-dead when it comes to fragments
719 req._Request__original = req._Request__original.partition('#')[0]
720 req._Request__r_type = req._Request__r_type.partition('#')[0]
724 def http_response(self, req, resp):
727 if resp.headers.get('Content-encoding', '') == 'gzip':
728 content = resp.read()
729 gz = gzip.GzipFile(fileobj=io.BytesIO(content), mode='rb')
731 uncompressed = io.BytesIO(gz.read())
732 except IOError as original_ioerror:
733 # There may be junk add the end of the file
734 # See http://stackoverflow.com/q/4928560/35070 for details
735 for i in range(1, 1024):
737 gz = gzip.GzipFile(fileobj=io.BytesIO(content[:-i]), mode='rb')
738 uncompressed = io.BytesIO(gz.read())
743 raise original_ioerror
744 resp = self.addinfourl_wrapper(uncompressed, old_resp.headers, old_resp.url, old_resp.code)
745 resp.msg = old_resp.msg
747 if resp.headers.get('Content-encoding', '') == 'deflate':
748 gz = io.BytesIO(self.deflate(resp.read()))
749 resp = self.addinfourl_wrapper(gz, old_resp.headers, old_resp.url, old_resp.code)
750 resp.msg = old_resp.msg
751 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
752 # https://github.com/rg3/youtube-dl/issues/6457).
753 if 300 <= resp.code < 400:
754 location = resp.headers.get('Location')
756 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
757 if sys.version_info >= (3, 0):
758 location = location.encode('iso-8859-1').decode('utf-8')
759 location_escaped = escape_url(location)
760 if location != location_escaped:
761 del resp.headers['Location']
762 resp.headers['Location'] = location_escaped
765 https_request = http_request
766 https_response = http_response
769 class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
770 def __init__(self, params, https_conn_class=None, *args, **kwargs):
771 compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)
772 self._https_conn_class = https_conn_class or compat_http_client.HTTPSConnection
773 self._params = params
775 def https_open(self, req):
777 if hasattr(self, '_context'): # python > 2.6
778 kwargs['context'] = self._context
779 if hasattr(self, '_check_hostname'): # python 3.x
780 kwargs['check_hostname'] = self._check_hostname
781 return self.do_open(functools.partial(
782 _create_http_connection, self, self._https_conn_class, True),
786 class YoutubeDLCookieProcessor(compat_urllib_request.HTTPCookieProcessor):
787 def __init__(self, cookiejar=None):
788 compat_urllib_request.HTTPCookieProcessor.__init__(self, cookiejar)
790 def http_response(self, request, response):
791 # Python 2 will choke on next HTTP request in row if there are non-ASCII
792 # characters in Set-Cookie HTTP header of last response (see
793 # https://github.com/rg3/youtube-dl/issues/6769).
794 # In order to at least prevent crashing we will percent encode Set-Cookie
795 # header before HTTPCookieProcessor starts processing it.
796 if sys.version_info < (3, 0) and response.headers:
797 for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
798 set_cookie = response.headers.get(set_cookie_header)
800 set_cookie_escaped = '; '.join([
801 escape_rfc3986(cookie_attr.strip())
802 for cookie_attr in set_cookie.decode('iso-8859-1').split(';')]).encode('iso-8859-1')
803 if set_cookie != set_cookie_escaped:
804 del response.headers[set_cookie_header]
805 response.headers[set_cookie_header] = set_cookie_escaped
806 return compat_urllib_request.HTTPCookieProcessor.http_response(self, request, response)
808 https_request = compat_urllib_request.HTTPCookieProcessor.http_request
809 https_response = http_response
812 def parse_iso8601(date_str, delimiter='T', timezone=None):
813 """ Return a UNIX timestamp from the given date """
820 r'(\.[0-9]+)?(?:Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
823 timezone = datetime.timedelta()
825 date_str = date_str[:-len(m.group(0))]
826 if not m.group('sign'):
827 timezone = datetime.timedelta()
829 sign = 1 if m.group('sign') == '+' else -1
830 timezone = datetime.timedelta(
831 hours=sign * int(m.group('hours')),
832 minutes=sign * int(m.group('minutes')))
833 date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
834 dt = datetime.datetime.strptime(date_str, date_format) - timezone
835 return calendar.timegm(dt.timetuple())
838 def unified_strdate(date_str, day_first=True):
839 """Return a string with the date in the format YYYYMMDD"""
845 date_str = date_str.replace(',', ' ')
846 # %z (UTC offset) is only supported in python>=3.2
847 if not re.match(r'^[0-9]{1,2}-[0-9]{1,2}-[0-9]{4}$', date_str):
848 date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
849 # Remove AM/PM + timezone
850 date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
852 format_expressions = [
857 '%b %dst %Y %I:%M%p',
858 '%b %dnd %Y %I:%M%p',
859 '%b %dth %Y %I:%M%p',
865 '%Y-%m-%d %H:%M:%S.%f',
868 '%Y-%m-%dT%H:%M:%SZ',
869 '%Y-%m-%dT%H:%M:%S.%fZ',
870 '%Y-%m-%dT%H:%M:%S.%f0Z',
872 '%Y-%m-%dT%H:%M:%S.%f',
876 format_expressions.extend([
884 format_expressions.extend([
891 for expression in format_expressions:
893 upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
896 if upload_date is None:
897 timetuple = email.utils.parsedate_tz(date_str)
899 upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
903 def determine_ext(url, default_ext='unknown_video'):
906 guess = url.partition('?')[0].rpartition('.')[2]
907 if re.match(r'^[A-Za-z0-9]+$', guess):
913 def subtitles_filename(filename, sub_lang, sub_format):
914 return filename.rsplit('.', 1)[0] + '.' + sub_lang + '.' + sub_format
917 def date_from_str(date_str):
919 Return a datetime object from a string in the format YYYYMMDD or
920 (now|today)[+-][0-9](day|week|month|year)(s)?"""
921 today = datetime.date.today()
922 if date_str in ('now', 'today'):
924 if date_str == 'yesterday':
925 return today - datetime.timedelta(days=1)
926 match = re.match('(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
927 if match is not None:
928 sign = match.group('sign')
929 time = int(match.group('time'))
932 unit = match.group('unit')
933 # A bad aproximation?
941 delta = datetime.timedelta(**{unit: time})
943 return datetime.datetime.strptime(date_str, "%Y%m%d").date()
946 def hyphenate_date(date_str):
948 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
949 match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str)
950 if match is not None:
951 return '-'.join(match.groups())
956 class DateRange(object):
957 """Represents a time interval between two dates"""
959 def __init__(self, start=None, end=None):
960 """start and end must be strings in the format accepted by date"""
961 if start is not None:
962 self.start = date_from_str(start)
964 self.start = datetime.datetime.min.date()
966 self.end = date_from_str(end)
968 self.end = datetime.datetime.max.date()
969 if self.start > self.end:
970 raise ValueError('Date range: "%s" , the start date must be before the end date' % self)
974 """Returns a range that only contains the given day"""
977 def __contains__(self, date):
978 """Check if the date is in the range"""
979 if not isinstance(date, datetime.date):
980 date = date_from_str(date)
981 return self.start <= date <= self.end
984 return '%s - %s' % (self.start.isoformat(), self.end.isoformat())
988 """ Returns the platform name as a compat_str """
989 res = platform.platform()
990 if isinstance(res, bytes):
991 res = res.decode(preferredencoding())
993 assert isinstance(res, compat_str)
997 def _windows_write_string(s, out):
998 """ Returns True if the string was written using special methods,
999 False if it has yet to be written out."""
1000 # Adapted from http://stackoverflow.com/a/3259271/35070
1003 import ctypes.wintypes
1011 fileno = out.fileno()
1012 except AttributeError:
1013 # If the output stream doesn't have a fileno, it's virtual
1015 except io.UnsupportedOperation:
1016 # Some strange Windows pseudo files?
1018 if fileno not in WIN_OUTPUT_IDS:
1021 GetStdHandle = ctypes.WINFUNCTYPE(
1022 ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
1023 (b"GetStdHandle", ctypes.windll.kernel32))
1024 h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
1026 WriteConsoleW = ctypes.WINFUNCTYPE(
1027 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
1028 ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
1029 ctypes.wintypes.LPVOID)((b"WriteConsoleW", ctypes.windll.kernel32))
1030 written = ctypes.wintypes.DWORD(0)
1032 GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b"GetFileType", ctypes.windll.kernel32))
1033 FILE_TYPE_CHAR = 0x0002
1034 FILE_TYPE_REMOTE = 0x8000
1035 GetConsoleMode = ctypes.WINFUNCTYPE(
1036 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
1037 ctypes.POINTER(ctypes.wintypes.DWORD))(
1038 (b"GetConsoleMode", ctypes.windll.kernel32))
1039 INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
1041 def not_a_console(handle):
1042 if handle == INVALID_HANDLE_VALUE or handle is None:
1044 return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR or
1045 GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0)
1047 if not_a_console(h):
1050 def next_nonbmp_pos(s):
1052 return next(i for i, c in enumerate(s) if ord(c) > 0xffff)
1053 except StopIteration:
1057 count = min(next_nonbmp_pos(s), 1024)
1059 ret = WriteConsoleW(
1060 h, s, count if count else 2, ctypes.byref(written), None)
1062 raise OSError('Failed to write string')
1063 if not count: # We just wrote a non-BMP character
1064 assert written.value == 2
1067 assert written.value > 0
1068 s = s[written.value:]
1072 def write_string(s, out=None, encoding=None):
1075 assert type(s) == compat_str
1077 if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'):
1078 if _windows_write_string(s, out):
1081 if ('b' in getattr(out, 'mode', '') or
1082 sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr
1083 byt = s.encode(encoding or preferredencoding(), 'ignore')
1085 elif hasattr(out, 'buffer'):
1086 enc = encoding or getattr(out, 'encoding', None) or preferredencoding()
1087 byt = s.encode(enc, 'ignore')
1088 out.buffer.write(byt)
1094 def bytes_to_intlist(bs):
1097 if isinstance(bs[0], int): # Python 3
1100 return [ord(c) for c in bs]
1103 def intlist_to_bytes(xs):
1106 return struct_pack('%dB' % len(xs), *xs)
1109 # Cross-platform file locking
1110 if sys.platform == 'win32':
1111 import ctypes.wintypes
1114 class OVERLAPPED(ctypes.Structure):
1116 ('Internal', ctypes.wintypes.LPVOID),
1117 ('InternalHigh', ctypes.wintypes.LPVOID),
1118 ('Offset', ctypes.wintypes.DWORD),
1119 ('OffsetHigh', ctypes.wintypes.DWORD),
1120 ('hEvent', ctypes.wintypes.HANDLE),
1123 kernel32 = ctypes.windll.kernel32
1124 LockFileEx = kernel32.LockFileEx
1125 LockFileEx.argtypes = [
1126 ctypes.wintypes.HANDLE, # hFile
1127 ctypes.wintypes.DWORD, # dwFlags
1128 ctypes.wintypes.DWORD, # dwReserved
1129 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1130 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1131 ctypes.POINTER(OVERLAPPED) # Overlapped
1133 LockFileEx.restype = ctypes.wintypes.BOOL
1134 UnlockFileEx = kernel32.UnlockFileEx
1135 UnlockFileEx.argtypes = [
1136 ctypes.wintypes.HANDLE, # hFile
1137 ctypes.wintypes.DWORD, # dwReserved
1138 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1139 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1140 ctypes.POINTER(OVERLAPPED) # Overlapped
1142 UnlockFileEx.restype = ctypes.wintypes.BOOL
1143 whole_low = 0xffffffff
1144 whole_high = 0x7fffffff
1146 def _lock_file(f, exclusive):
1147 overlapped = OVERLAPPED()
1148 overlapped.Offset = 0
1149 overlapped.OffsetHigh = 0
1150 overlapped.hEvent = 0
1151 f._lock_file_overlapped_p = ctypes.pointer(overlapped)
1152 handle = msvcrt.get_osfhandle(f.fileno())
1153 if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0,
1154 whole_low, whole_high, f._lock_file_overlapped_p):
1155 raise OSError('Locking file failed: %r' % ctypes.FormatError())
1157 def _unlock_file(f):
1158 assert f._lock_file_overlapped_p
1159 handle = msvcrt.get_osfhandle(f.fileno())
1160 if not UnlockFileEx(handle, 0,
1161 whole_low, whole_high, f._lock_file_overlapped_p):
1162 raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
1167 def _lock_file(f, exclusive):
1168 fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
1170 def _unlock_file(f):
1171 fcntl.flock(f, fcntl.LOCK_UN)
1174 class locked_file(object):
1175 def __init__(self, filename, mode, encoding=None):
1176 assert mode in ['r', 'a', 'w']
1177 self.f = io.open(filename, mode, encoding=encoding)
1180 def __enter__(self):
1181 exclusive = self.mode != 'r'
1183 _lock_file(self.f, exclusive)
1189 def __exit__(self, etype, value, traceback):
1191 _unlock_file(self.f)
1198 def write(self, *args):
1199 return self.f.write(*args)
1201 def read(self, *args):
1202 return self.f.read(*args)
1205 def get_filesystem_encoding():
1206 encoding = sys.getfilesystemencoding()
1207 return encoding if encoding is not None else 'utf-8'
1210 def shell_quote(args):
1212 encoding = get_filesystem_encoding()
1214 if isinstance(a, bytes):
1215 # We may get a filename encoded with 'encodeFilename'
1216 a = a.decode(encoding)
1217 quoted_args.append(pipes.quote(a))
1218 return ' '.join(quoted_args)
1221 def smuggle_url(url, data):
1222 """ Pass additional data in a URL for internal use. """
1224 sdata = compat_urllib_parse.urlencode(
1225 {'__youtubedl_smuggle': json.dumps(data)})
1226 return url + '#' + sdata
1229 def unsmuggle_url(smug_url, default=None):
1230 if '#__youtubedl_smuggle' not in smug_url:
1231 return smug_url, default
1232 url, _, sdata = smug_url.rpartition('#')
1233 jsond = compat_parse_qs(sdata)['__youtubedl_smuggle'][0]
1234 data = json.loads(jsond)
1238 def format_bytes(bytes):
1241 if type(bytes) is str:
1242 bytes = float(bytes)
1246 exponent = int(math.log(bytes, 1024.0))
1247 suffix = ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent]
1248 converted = float(bytes) / float(1024 ** exponent)
1249 return '%.2f%s' % (converted, suffix)
1252 def parse_filesize(s):
1256 # The lower-case forms are of course incorrect and inofficial,
1257 # but we support those too
1295 units_re = '|'.join(re.escape(u) for u in _UNIT_TABLE)
1297 r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)' % units_re, s)
1301 num_str = m.group('num').replace(',', '.')
1302 mult = _UNIT_TABLE[m.group('unit')]
1303 return int(float(num_str) * mult)
1306 def month_by_name(name):
1307 """ Return the number of a month by (locale-independently) English name """
1310 return ENGLISH_MONTH_NAMES.index(name) + 1
1315 def month_by_abbreviation(abbrev):
1316 """ Return the number of a month by (locale-independently) English
1320 return [s[:3] for s in ENGLISH_MONTH_NAMES].index(abbrev) + 1
1325 def fix_xml_ampersands(xml_str):
1326 """Replace all the '&' by '&' in XML"""
1328 r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1333 def setproctitle(title):
1334 assert isinstance(title, compat_str)
1336 libc = ctypes.cdll.LoadLibrary("libc.so.6")
1339 title_bytes = title.encode('utf-8')
1340 buf = ctypes.create_string_buffer(len(title_bytes))
1341 buf.value = title_bytes
1343 libc.prctl(15, buf, 0, 0, 0)
1344 except AttributeError:
1345 return # Strange libc, just skip this
1348 def remove_start(s, start):
1349 if s.startswith(start):
1350 return s[len(start):]
1354 def remove_end(s, end):
1356 return s[:-len(end)]
1360 def url_basename(url):
1361 path = compat_urlparse.urlparse(url).path
1362 return path.strip('/').split('/')[-1]
1365 class HEADRequest(compat_urllib_request.Request):
1366 def get_method(self):
1370 def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
1373 v = getattr(v, get_attr, None)
1376 return default if v is None else (int(v) * invscale // scale)
1379 def str_or_none(v, default=None):
1380 return default if v is None else compat_str(v)
1383 def str_to_int(int_str):
1384 """ A more relaxed version of int_or_none """
1387 int_str = re.sub(r'[,\.\+]', '', int_str)
1391 def float_or_none(v, scale=1, invscale=1, default=None):
1392 return default if v is None else (float(v) * invscale / scale)
1395 def parse_duration(s):
1396 if not isinstance(s, compat_basestring):
1404 (?P<only_mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*|
1405 (?P<only_hours>[0-9.]+)\s*(?:hours?)|
1407 \s*(?P<hours_reversed>[0-9]+)\s*(?:[:h]|hours?)\s*(?P<mins_reversed>[0-9]+)\s*(?:[:m]|mins?\.?|minutes?)\s*|
1410 (?:(?P<days>[0-9]+)\s*(?:[:d]|days?)\s*)?
1411 (?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*
1413 (?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*
1415 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?
1420 if m.group('only_mins'):
1421 return float_or_none(m.group('only_mins'), invscale=60)
1422 if m.group('only_hours'):
1423 return float_or_none(m.group('only_hours'), invscale=60 * 60)
1425 res += int(m.group('secs'))
1426 if m.group('mins_reversed'):
1427 res += int(m.group('mins_reversed')) * 60
1429 res += int(m.group('mins')) * 60
1430 if m.group('hours'):
1431 res += int(m.group('hours')) * 60 * 60
1432 if m.group('hours_reversed'):
1433 res += int(m.group('hours_reversed')) * 60 * 60
1435 res += int(m.group('days')) * 24 * 60 * 60
1437 res += float(m.group('ms'))
1441 def prepend_extension(filename, ext, expected_real_ext=None):
1442 name, real_ext = os.path.splitext(filename)
1444 '{0}.{1}{2}'.format(name, ext, real_ext)
1445 if not expected_real_ext or real_ext[1:] == expected_real_ext
1446 else '{0}.{1}'.format(filename, ext))
1449 def replace_extension(filename, ext, expected_real_ext=None):
1450 name, real_ext = os.path.splitext(filename)
1451 return '{0}.{1}'.format(
1452 name if not expected_real_ext or real_ext[1:] == expected_real_ext else filename,
1456 def check_executable(exe, args=[]):
1457 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
1458 args can be a list of arguments for a short output (like -version) """
1460 subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
1466 def get_exe_version(exe, args=['--version'],
1467 version_re=None, unrecognized='present'):
1468 """ Returns the version of the specified executable,
1469 or False if the executable is not present """
1471 out, _ = subprocess.Popen(
1472 [encodeArgument(exe)] + args,
1473 stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
1476 if isinstance(out, bytes): # Python 2.x
1477 out = out.decode('ascii', 'ignore')
1478 return detect_exe_version(out, version_re, unrecognized)
1481 def detect_exe_version(output, version_re=None, unrecognized='present'):
1482 assert isinstance(output, compat_str)
1483 if version_re is None:
1484 version_re = r'version\s+([-0-9._a-zA-Z]+)'
1485 m = re.search(version_re, output)
1492 class PagedList(object):
1494 # This is only useful for tests
1495 return len(self.getslice())
1498 class OnDemandPagedList(PagedList):
1499 def __init__(self, pagefunc, pagesize):
1500 self._pagefunc = pagefunc
1501 self._pagesize = pagesize
1503 def getslice(self, start=0, end=None):
1505 for pagenum in itertools.count(start // self._pagesize):
1506 firstid = pagenum * self._pagesize
1507 nextfirstid = pagenum * self._pagesize + self._pagesize
1508 if start >= nextfirstid:
1511 page_results = list(self._pagefunc(pagenum))
1514 start % self._pagesize
1515 if firstid <= start < nextfirstid
1519 ((end - 1) % self._pagesize) + 1
1520 if (end is not None and firstid <= end <= nextfirstid)
1523 if startv != 0 or endv is not None:
1524 page_results = page_results[startv:endv]
1525 res.extend(page_results)
1527 # A little optimization - if current page is not "full", ie. does
1528 # not contain page_size videos then we can assume that this page
1529 # is the last one - there are no more ids on further pages -
1530 # i.e. no need to query again.
1531 if len(page_results) + startv < self._pagesize:
1534 # If we got the whole page, but the next page is not interesting,
1535 # break out early as well
1536 if end == nextfirstid:
1541 class InAdvancePagedList(PagedList):
1542 def __init__(self, pagefunc, pagecount, pagesize):
1543 self._pagefunc = pagefunc
1544 self._pagecount = pagecount
1545 self._pagesize = pagesize
1547 def getslice(self, start=0, end=None):
1549 start_page = start // self._pagesize
1551 self._pagecount if end is None else (end // self._pagesize + 1))
1552 skip_elems = start - start_page * self._pagesize
1553 only_more = None if end is None else end - start
1554 for pagenum in range(start_page, end_page):
1555 page = list(self._pagefunc(pagenum))
1557 page = page[skip_elems:]
1559 if only_more is not None:
1560 if len(page) < only_more:
1561 only_more -= len(page)
1563 page = page[:only_more]
1570 def uppercase_escape(s):
1571 unicode_escape = codecs.getdecoder('unicode_escape')
1573 r'\\U[0-9a-fA-F]{8}',
1574 lambda m: unicode_escape(m.group(0))[0],
1578 def lowercase_escape(s):
1579 unicode_escape = codecs.getdecoder('unicode_escape')
1581 r'\\u[0-9a-fA-F]{4}',
1582 lambda m: unicode_escape(m.group(0))[0],
1586 def escape_rfc3986(s):
1587 """Escape non-ASCII characters as suggested by RFC 3986"""
1588 if sys.version_info < (3, 0) and isinstance(s, compat_str):
1589 s = s.encode('utf-8')
1590 return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]")
1593 def escape_url(url):
1594 """Escape URL as suggested by RFC 3986"""
1595 url_parsed = compat_urllib_parse_urlparse(url)
1596 return url_parsed._replace(
1597 path=escape_rfc3986(url_parsed.path),
1598 params=escape_rfc3986(url_parsed.params),
1599 query=escape_rfc3986(url_parsed.query),
1600 fragment=escape_rfc3986(url_parsed.fragment)
1604 struct.pack('!I', 0)
1606 # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
1607 def struct_pack(spec, *args):
1608 if isinstance(spec, compat_str):
1609 spec = spec.encode('ascii')
1610 return struct.pack(spec, *args)
1612 def struct_unpack(spec, *args):
1613 if isinstance(spec, compat_str):
1614 spec = spec.encode('ascii')
1615 return struct.unpack(spec, *args)
1617 struct_pack = struct.pack
1618 struct_unpack = struct.unpack
1621 def read_batch_urls(batch_fd):
1623 if not isinstance(url, compat_str):
1624 url = url.decode('utf-8', 'replace')
1625 BOM_UTF8 = '\xef\xbb\xbf'
1626 if url.startswith(BOM_UTF8):
1627 url = url[len(BOM_UTF8):]
1629 if url.startswith(('#', ';', ']')):
1633 with contextlib.closing(batch_fd) as fd:
1634 return [url for url in map(fixup, fd) if url]
1637 def urlencode_postdata(*args, **kargs):
1638 return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
1641 def encode_dict(d, encoding='utf-8'):
1642 return dict((k.encode(encoding), v.encode(encoding)) for k, v in d.items())
1646 etree_iter = xml.etree.ElementTree.Element.iter
1647 except AttributeError: # Python <=2.6
1648 etree_iter = lambda n: n.findall('.//*')
1652 class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
1653 def doctype(self, name, pubid, system):
1654 pass # Ignore doctypes
1656 parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
1657 kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
1658 tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
1659 # Fix up XML parser in Python 2.x
1660 if sys.version_info < (3, 0):
1661 for n in etree_iter(tree):
1662 if n.text is not None:
1663 if not isinstance(n.text, compat_str):
1664 n.text = n.text.decode('utf-8')
1677 def parse_age_limit(s):
1680 m = re.match(r'^(?P<age>\d{1,2})\+?$', s)
1681 return int(m.group('age')) if m else US_RATINGS.get(s, None)
1684 def strip_jsonp(code):
1686 r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code)
1689 def js_to_json(code):
1692 if v in ('true', 'false', 'null'):
1694 if v.startswith('"'):
1696 if v.startswith("'"):
1698 v = re.sub(r"\\\\|\\'|\"", lambda m: {
1705 res = re.sub(r'''(?x)
1706 "(?:[^"\\]*(?:\\\\|\\['"nu]))*[^"\\]*"|
1707 '(?:[^'\\]*(?:\\\\|\\['"nu]))*[^'\\]*'|
1708 [a-zA-Z_][.a-zA-Z_0-9]*
1710 res = re.sub(r',(\s*[\]}])', lambda m: m.group(1), res)
1714 def qualities(quality_ids):
1715 """ Get a numeric quality value out of a list of possible values """
1718 return quality_ids.index(qid)
1724 DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s'
1727 def limit_length(s, length):
1728 """ Add ellipses to overly long strings """
1733 return s[:length - len(ELLIPSES)] + ELLIPSES
1737 def version_tuple(v):
1738 return tuple(int(e) for e in re.split(r'[-.]', v))
1741 def is_outdated_version(version, limit, assume_new=True):
1743 return not assume_new
1745 return version_tuple(version) < version_tuple(limit)
1747 return not assume_new
1750 def ytdl_is_updateable():
1751 """ Returns if youtube-dl can be updated with -U """
1752 from zipimport import zipimporter
1754 return isinstance(globals().get('__loader__'), zipimporter) or hasattr(sys, 'frozen')
1757 def args_to_str(args):
1758 # Get a short string representation for a subprocess command
1759 return ' '.join(shlex_quote(a) for a in args)
1762 def mimetype2ext(mt):
1763 _, _, res = mt.rpartition('/')
1767 'x-mp4-fragmented': 'mp4',
1772 def urlhandle_detect_ext(url_handle):
1775 getheader = lambda h: url_handle.headers[h]
1776 except AttributeError: # Python < 3
1777 getheader = url_handle.info().getheader
1779 cd = getheader('Content-Disposition')
1781 m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd)
1783 e = determine_ext(m.group('filename'), default_ext=None)
1787 return mimetype2ext(getheader('Content-Type'))
1790 def age_restricted(content_limit, age_limit):
1791 """ Returns True iff the content should be blocked """
1793 if age_limit is None: # No limit set
1795 if content_limit is None:
1796 return False # Content available for everyone
1797 return age_limit < content_limit
1800 def is_html(first_bytes):
1801 """ Detect whether a file contains HTML by examining its first bytes. """
1804 (b'\xef\xbb\xbf', 'utf-8'),
1805 (b'\x00\x00\xfe\xff', 'utf-32-be'),
1806 (b'\xff\xfe\x00\x00', 'utf-32-le'),
1807 (b'\xff\xfe', 'utf-16-le'),
1808 (b'\xfe\xff', 'utf-16-be'),
1810 for bom, enc in BOMS:
1811 if first_bytes.startswith(bom):
1812 s = first_bytes[len(bom):].decode(enc, 'replace')
1815 s = first_bytes.decode('utf-8', 'replace')
1817 return re.match(r'^\s*<', s)
1820 def determine_protocol(info_dict):
1821 protocol = info_dict.get('protocol')
1822 if protocol is not None:
1825 url = info_dict['url']
1826 if url.startswith('rtmp'):
1828 elif url.startswith('mms'):
1830 elif url.startswith('rtsp'):
1833 ext = determine_ext(url)
1839 return compat_urllib_parse_urlparse(url).scheme
1842 def render_table(header_row, data):
1843 """ Render a list of rows, each as a list of values """
1844 table = [header_row] + data
1845 max_lens = [max(len(compat_str(v)) for v in col) for col in zip(*table)]
1846 format_str = ' '.join('%-' + compat_str(ml + 1) + 's' for ml in max_lens[:-1]) + '%s'
1847 return '\n'.join(format_str % tuple(row) for row in table)
1850 def _match_one(filter_part, dct):
1851 COMPARISON_OPERATORS = {
1859 operator_rex = re.compile(r'''(?x)\s*
1861 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1863 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
1864 (?P<strval>(?![0-9.])[a-z0-9A-Z]*)
1867 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
1868 m = operator_rex.search(filter_part)
1870 op = COMPARISON_OPERATORS[m.group('op')]
1871 if m.group('strval') is not None:
1872 if m.group('op') not in ('=', '!='):
1874 'Operator %s does not support string values!' % m.group('op'))
1875 comparison_value = m.group('strval')
1878 comparison_value = int(m.group('intval'))
1880 comparison_value = parse_filesize(m.group('intval'))
1881 if comparison_value is None:
1882 comparison_value = parse_filesize(m.group('intval') + 'B')
1883 if comparison_value is None:
1885 'Invalid integer value %r in filter part %r' % (
1886 m.group('intval'), filter_part))
1887 actual_value = dct.get(m.group('key'))
1888 if actual_value is None:
1889 return m.group('none_inclusive')
1890 return op(actual_value, comparison_value)
1893 '': lambda v: v is not None,
1894 '!': lambda v: v is None,
1896 operator_rex = re.compile(r'''(?x)\s*
1897 (?P<op>%s)\s*(?P<key>[a-z_]+)
1899 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
1900 m = operator_rex.search(filter_part)
1902 op = UNARY_OPERATORS[m.group('op')]
1903 actual_value = dct.get(m.group('key'))
1904 return op(actual_value)
1906 raise ValueError('Invalid filter part %r' % filter_part)
1909 def match_str(filter_str, dct):
1910 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """
1913 _match_one(filter_part, dct) for filter_part in filter_str.split('&'))
1916 def match_filter_func(filter_str):
1917 def _match_func(info_dict):
1918 if match_str(filter_str, info_dict):
1921 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1922 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
1926 def parse_dfxp_time_expr(time_expr):
1930 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
1932 return float(mobj.group('time_offset'))
1934 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:\.\d+)?)$', time_expr)
1936 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3))
1939 def srt_subtitles_timecode(seconds):
1940 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
1943 def dfxp2srt(dfxp_data):
1944 _x = functools.partial(xpath_with_ns, ns_map={
1945 'ttml': 'http://www.w3.org/ns/ttml',
1946 'ttaf1': 'http://www.w3.org/2006/10/ttaf1',
1949 def parse_node(node):
1950 str_or_empty = functools.partial(str_or_none, default='')
1952 out = str_or_empty(node.text)
1955 if child.tag in (_x('ttml:br'), _x('ttaf1:br'), 'br'):
1956 out += '\n' + str_or_empty(child.tail)
1957 elif child.tag in (_x('ttml:span'), _x('ttaf1:span'), 'span'):
1958 out += str_or_empty(parse_node(child))
1960 out += str_or_empty(xml.etree.ElementTree.tostring(child))
1964 dfxp = xml.etree.ElementTree.fromstring(dfxp_data.encode('utf-8'))
1966 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall('.//p')
1969 raise ValueError('Invalid dfxp/TTML subtitle')
1971 for para, index in zip(paras, itertools.count(1)):
1972 begin_time = parse_dfxp_time_expr(para.attrib['begin'])
1973 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
1975 end_time = begin_time + parse_dfxp_time_expr(para.attrib['dur'])
1976 out.append('%d\n%s --> %s\n%s\n\n' % (
1978 srt_subtitles_timecode(begin_time),
1979 srt_subtitles_timecode(end_time),
1985 def cli_option(params, command_option, param):
1986 param = params.get(param)
1987 return [command_option, param] if param is not None else []
1990 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
1991 param = params.get(param)
1992 assert isinstance(param, bool)
1994 return [command_option + separator + (true_value if param else false_value)]
1995 return [command_option, true_value if param else false_value]
1998 def cli_valueless_option(params, command_option, param, expected_value=True):
1999 param = params.get(param)
2000 return [command_option] if param == expected_value else []
2003 def cli_configuration_args(params, param, default=[]):
2004 ex_args = params.get(param)
2007 assert isinstance(ex_args, list)
2011 class ISO639Utils(object):
2012 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
2201 def short2long(cls, code):
2202 """Convert language code from ISO 639-1 to ISO 639-2/T"""
2203 return cls._lang_map.get(code[:2])
2206 def long2short(cls, code):
2207 """Convert language code from ISO 639-2/T to ISO 639-1"""
2208 for short_name, long_name in cls._lang_map.items():
2209 if long_name == code:
2213 class ISO3166Utils(object):
2214 # From http://data.okfn.org/data/core/country-list
2216 'AF': 'Afghanistan',
2217 'AX': 'Åland Islands',
2220 'AS': 'American Samoa',
2225 'AG': 'Antigua and Barbuda',
2242 'BO': 'Bolivia, Plurinational State of',
2243 'BQ': 'Bonaire, Sint Eustatius and Saba',
2244 'BA': 'Bosnia and Herzegovina',
2246 'BV': 'Bouvet Island',
2248 'IO': 'British Indian Ocean Territory',
2249 'BN': 'Brunei Darussalam',
2251 'BF': 'Burkina Faso',
2257 'KY': 'Cayman Islands',
2258 'CF': 'Central African Republic',
2262 'CX': 'Christmas Island',
2263 'CC': 'Cocos (Keeling) Islands',
2267 'CD': 'Congo, the Democratic Republic of the',
2268 'CK': 'Cook Islands',
2270 'CI': 'Côte d\'Ivoire',
2275 'CZ': 'Czech Republic',
2279 'DO': 'Dominican Republic',
2282 'SV': 'El Salvador',
2283 'GQ': 'Equatorial Guinea',
2287 'FK': 'Falkland Islands (Malvinas)',
2288 'FO': 'Faroe Islands',
2292 'GF': 'French Guiana',
2293 'PF': 'French Polynesia',
2294 'TF': 'French Southern Territories',
2309 'GW': 'Guinea-Bissau',
2312 'HM': 'Heard Island and McDonald Islands',
2313 'VA': 'Holy See (Vatican City State)',
2320 'IR': 'Iran, Islamic Republic of',
2323 'IM': 'Isle of Man',
2333 'KP': 'Korea, Democratic People\'s Republic of',
2334 'KR': 'Korea, Republic of',
2337 'LA': 'Lao People\'s Democratic Republic',
2343 'LI': 'Liechtenstein',
2347 'MK': 'Macedonia, the Former Yugoslav Republic of',
2354 'MH': 'Marshall Islands',
2360 'FM': 'Micronesia, Federated States of',
2361 'MD': 'Moldova, Republic of',
2372 'NL': 'Netherlands',
2373 'NC': 'New Caledonia',
2374 'NZ': 'New Zealand',
2379 'NF': 'Norfolk Island',
2380 'MP': 'Northern Mariana Islands',
2385 'PS': 'Palestine, State of',
2387 'PG': 'Papua New Guinea',
2390 'PH': 'Philippines',
2394 'PR': 'Puerto Rico',
2398 'RU': 'Russian Federation',
2400 'BL': 'Saint Barthélemy',
2401 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
2402 'KN': 'Saint Kitts and Nevis',
2403 'LC': 'Saint Lucia',
2404 'MF': 'Saint Martin (French part)',
2405 'PM': 'Saint Pierre and Miquelon',
2406 'VC': 'Saint Vincent and the Grenadines',
2409 'ST': 'Sao Tome and Principe',
2410 'SA': 'Saudi Arabia',
2414 'SL': 'Sierra Leone',
2416 'SX': 'Sint Maarten (Dutch part)',
2419 'SB': 'Solomon Islands',
2421 'ZA': 'South Africa',
2422 'GS': 'South Georgia and the South Sandwich Islands',
2423 'SS': 'South Sudan',
2428 'SJ': 'Svalbard and Jan Mayen',
2431 'CH': 'Switzerland',
2432 'SY': 'Syrian Arab Republic',
2433 'TW': 'Taiwan, Province of China',
2435 'TZ': 'Tanzania, United Republic of',
2437 'TL': 'Timor-Leste',
2441 'TT': 'Trinidad and Tobago',
2444 'TM': 'Turkmenistan',
2445 'TC': 'Turks and Caicos Islands',
2449 'AE': 'United Arab Emirates',
2450 'GB': 'United Kingdom',
2451 'US': 'United States',
2452 'UM': 'United States Minor Outlying Islands',
2456 'VE': 'Venezuela, Bolivarian Republic of',
2458 'VG': 'Virgin Islands, British',
2459 'VI': 'Virgin Islands, U.S.',
2460 'WF': 'Wallis and Futuna',
2461 'EH': 'Western Sahara',
2468 def short2full(cls, code):
2469 """Convert an ISO 3166-2 country code to the corresponding full name"""
2470 return cls._country_map.get(code.upper())
2473 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
2474 def __init__(self, proxies=None):
2475 # Set default handlers
2476 for type in ('http', 'https'):
2477 setattr(self, '%s_open' % type,
2478 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
2479 meth(r, proxy, type))
2480 return compat_urllib_request.ProxyHandler.__init__(self, proxies)
2482 def proxy_open(self, req, proxy, type):
2483 req_proxy = req.headers.get('Ytdl-request-proxy')
2484 if req_proxy is not None:
2486 del req.headers['Ytdl-request-proxy']
2488 if proxy == '__noproxy__':
2489 return None # No Proxy
2490 return compat_urllib_request.ProxyHandler.proxy_open(
2491 self, req, proxy, type)