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
752 if 300 <= resp.code < 400:
753 location = resp.headers.get('Location')
755 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
756 if sys.version_info >= (3, 0):
757 location = location.encode('iso-8859-1').decode('utf-8')
758 location_escaped = escape_url(location)
759 if location != location_escaped:
760 del resp.headers['Location']
761 resp.headers['Location'] = location_escaped
764 https_request = http_request
765 https_response = http_response
768 class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
769 def __init__(self, params, https_conn_class=None, *args, **kwargs):
770 compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)
771 self._https_conn_class = https_conn_class or compat_http_client.HTTPSConnection
772 self._params = params
774 def https_open(self, req):
776 if hasattr(self, '_context'): # python > 2.6
777 kwargs['context'] = self._context
778 if hasattr(self, '_check_hostname'): # python 3.x
779 kwargs['check_hostname'] = self._check_hostname
780 return self.do_open(functools.partial(
781 _create_http_connection, self, self._https_conn_class, True),
785 class YoutubeDLCookieProcessor(compat_urllib_request.HTTPCookieProcessor):
786 def __init__(self, cookiejar=None):
787 compat_urllib_request.HTTPCookieProcessor.__init__(self, cookiejar)
789 def http_response(self, request, response):
790 # Python 2 will choke on next HTTP request in row if there are non-ASCII
791 # characters in Set-Cookie HTTP header of last response (see
792 # https://github.com/rg3/youtube-dl/issues/6769).
793 # In order to at least prevent crashing we will percent encode Set-Cookie
794 # header before HTTPCookieProcessor starts processing it.
795 if sys.version_info < (3, 0) and response.headers:
796 for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
797 set_cookie = response.headers.get(set_cookie_header)
799 set_cookie_escaped = '; '.join([
800 escape_rfc3986(cookie_attr.strip())
801 for cookie_attr in set_cookie.decode('iso-8859-1').split(';')]).encode('iso-8859-1')
802 if set_cookie != set_cookie_escaped:
803 del response.headers[set_cookie_header]
804 response.headers[set_cookie_header] = set_cookie_escaped
805 return compat_urllib_request.HTTPCookieProcessor.http_response(self, request, response)
807 https_request = compat_urllib_request.HTTPCookieProcessor.http_request
808 https_response = http_response
811 def parse_iso8601(date_str, delimiter='T', timezone=None):
812 """ Return a UNIX timestamp from the given date """
819 r'(\.[0-9]+)?(?:Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
822 timezone = datetime.timedelta()
824 date_str = date_str[:-len(m.group(0))]
825 if not m.group('sign'):
826 timezone = datetime.timedelta()
828 sign = 1 if m.group('sign') == '+' else -1
829 timezone = datetime.timedelta(
830 hours=sign * int(m.group('hours')),
831 minutes=sign * int(m.group('minutes')))
832 date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
833 dt = datetime.datetime.strptime(date_str, date_format) - timezone
834 return calendar.timegm(dt.timetuple())
837 def unified_strdate(date_str, day_first=True):
838 """Return a string with the date in the format YYYYMMDD"""
844 date_str = date_str.replace(',', ' ')
845 # %z (UTC offset) is only supported in python>=3.2
846 if not re.match(r'^[0-9]{1,2}-[0-9]{1,2}-[0-9]{4}$', date_str):
847 date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
848 # Remove AM/PM + timezone
849 date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
851 format_expressions = [
856 '%b %dst %Y %I:%M%p',
857 '%b %dnd %Y %I:%M%p',
858 '%b %dth %Y %I:%M%p',
864 '%Y-%m-%d %H:%M:%S.%f',
867 '%Y-%m-%dT%H:%M:%SZ',
868 '%Y-%m-%dT%H:%M:%S.%fZ',
869 '%Y-%m-%dT%H:%M:%S.%f0Z',
871 '%Y-%m-%dT%H:%M:%S.%f',
875 format_expressions.extend([
883 format_expressions.extend([
890 for expression in format_expressions:
892 upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
895 if upload_date is None:
896 timetuple = email.utils.parsedate_tz(date_str)
898 upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
902 def determine_ext(url, default_ext='unknown_video'):
905 guess = url.partition('?')[0].rpartition('.')[2]
906 if re.match(r'^[A-Za-z0-9]+$', guess):
912 def subtitles_filename(filename, sub_lang, sub_format):
913 return filename.rsplit('.', 1)[0] + '.' + sub_lang + '.' + sub_format
916 def date_from_str(date_str):
918 Return a datetime object from a string in the format YYYYMMDD or
919 (now|today)[+-][0-9](day|week|month|year)(s)?"""
920 today = datetime.date.today()
921 if date_str in ('now', 'today'):
923 if date_str == 'yesterday':
924 return today - datetime.timedelta(days=1)
925 match = re.match('(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
926 if match is not None:
927 sign = match.group('sign')
928 time = int(match.group('time'))
931 unit = match.group('unit')
932 # A bad aproximation?
940 delta = datetime.timedelta(**{unit: time})
942 return datetime.datetime.strptime(date_str, "%Y%m%d").date()
945 def hyphenate_date(date_str):
947 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
948 match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str)
949 if match is not None:
950 return '-'.join(match.groups())
955 class DateRange(object):
956 """Represents a time interval between two dates"""
958 def __init__(self, start=None, end=None):
959 """start and end must be strings in the format accepted by date"""
960 if start is not None:
961 self.start = date_from_str(start)
963 self.start = datetime.datetime.min.date()
965 self.end = date_from_str(end)
967 self.end = datetime.datetime.max.date()
968 if self.start > self.end:
969 raise ValueError('Date range: "%s" , the start date must be before the end date' % self)
973 """Returns a range that only contains the given day"""
976 def __contains__(self, date):
977 """Check if the date is in the range"""
978 if not isinstance(date, datetime.date):
979 date = date_from_str(date)
980 return self.start <= date <= self.end
983 return '%s - %s' % (self.start.isoformat(), self.end.isoformat())
987 """ Returns the platform name as a compat_str """
988 res = platform.platform()
989 if isinstance(res, bytes):
990 res = res.decode(preferredencoding())
992 assert isinstance(res, compat_str)
996 def _windows_write_string(s, out):
997 """ Returns True if the string was written using special methods,
998 False if it has yet to be written out."""
999 # Adapted from http://stackoverflow.com/a/3259271/35070
1002 import ctypes.wintypes
1010 fileno = out.fileno()
1011 except AttributeError:
1012 # If the output stream doesn't have a fileno, it's virtual
1014 except io.UnsupportedOperation:
1015 # Some strange Windows pseudo files?
1017 if fileno not in WIN_OUTPUT_IDS:
1020 GetStdHandle = ctypes.WINFUNCTYPE(
1021 ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
1022 (b"GetStdHandle", ctypes.windll.kernel32))
1023 h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
1025 WriteConsoleW = ctypes.WINFUNCTYPE(
1026 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
1027 ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
1028 ctypes.wintypes.LPVOID)((b"WriteConsoleW", ctypes.windll.kernel32))
1029 written = ctypes.wintypes.DWORD(0)
1031 GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b"GetFileType", ctypes.windll.kernel32))
1032 FILE_TYPE_CHAR = 0x0002
1033 FILE_TYPE_REMOTE = 0x8000
1034 GetConsoleMode = ctypes.WINFUNCTYPE(
1035 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
1036 ctypes.POINTER(ctypes.wintypes.DWORD))(
1037 (b"GetConsoleMode", ctypes.windll.kernel32))
1038 INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
1040 def not_a_console(handle):
1041 if handle == INVALID_HANDLE_VALUE or handle is None:
1043 return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR or
1044 GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0)
1046 if not_a_console(h):
1049 def next_nonbmp_pos(s):
1051 return next(i for i, c in enumerate(s) if ord(c) > 0xffff)
1052 except StopIteration:
1056 count = min(next_nonbmp_pos(s), 1024)
1058 ret = WriteConsoleW(
1059 h, s, count if count else 2, ctypes.byref(written), None)
1061 raise OSError('Failed to write string')
1062 if not count: # We just wrote a non-BMP character
1063 assert written.value == 2
1066 assert written.value > 0
1067 s = s[written.value:]
1071 def write_string(s, out=None, encoding=None):
1074 assert type(s) == compat_str
1076 if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'):
1077 if _windows_write_string(s, out):
1080 if ('b' in getattr(out, 'mode', '') or
1081 sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr
1082 byt = s.encode(encoding or preferredencoding(), 'ignore')
1084 elif hasattr(out, 'buffer'):
1085 enc = encoding or getattr(out, 'encoding', None) or preferredencoding()
1086 byt = s.encode(enc, 'ignore')
1087 out.buffer.write(byt)
1093 def bytes_to_intlist(bs):
1096 if isinstance(bs[0], int): # Python 3
1099 return [ord(c) for c in bs]
1102 def intlist_to_bytes(xs):
1105 return struct_pack('%dB' % len(xs), *xs)
1108 # Cross-platform file locking
1109 if sys.platform == 'win32':
1110 import ctypes.wintypes
1113 class OVERLAPPED(ctypes.Structure):
1115 ('Internal', ctypes.wintypes.LPVOID),
1116 ('InternalHigh', ctypes.wintypes.LPVOID),
1117 ('Offset', ctypes.wintypes.DWORD),
1118 ('OffsetHigh', ctypes.wintypes.DWORD),
1119 ('hEvent', ctypes.wintypes.HANDLE),
1122 kernel32 = ctypes.windll.kernel32
1123 LockFileEx = kernel32.LockFileEx
1124 LockFileEx.argtypes = [
1125 ctypes.wintypes.HANDLE, # hFile
1126 ctypes.wintypes.DWORD, # dwFlags
1127 ctypes.wintypes.DWORD, # dwReserved
1128 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1129 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1130 ctypes.POINTER(OVERLAPPED) # Overlapped
1132 LockFileEx.restype = ctypes.wintypes.BOOL
1133 UnlockFileEx = kernel32.UnlockFileEx
1134 UnlockFileEx.argtypes = [
1135 ctypes.wintypes.HANDLE, # hFile
1136 ctypes.wintypes.DWORD, # dwReserved
1137 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1138 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1139 ctypes.POINTER(OVERLAPPED) # Overlapped
1141 UnlockFileEx.restype = ctypes.wintypes.BOOL
1142 whole_low = 0xffffffff
1143 whole_high = 0x7fffffff
1145 def _lock_file(f, exclusive):
1146 overlapped = OVERLAPPED()
1147 overlapped.Offset = 0
1148 overlapped.OffsetHigh = 0
1149 overlapped.hEvent = 0
1150 f._lock_file_overlapped_p = ctypes.pointer(overlapped)
1151 handle = msvcrt.get_osfhandle(f.fileno())
1152 if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0,
1153 whole_low, whole_high, f._lock_file_overlapped_p):
1154 raise OSError('Locking file failed: %r' % ctypes.FormatError())
1156 def _unlock_file(f):
1157 assert f._lock_file_overlapped_p
1158 handle = msvcrt.get_osfhandle(f.fileno())
1159 if not UnlockFileEx(handle, 0,
1160 whole_low, whole_high, f._lock_file_overlapped_p):
1161 raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
1166 def _lock_file(f, exclusive):
1167 fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
1169 def _unlock_file(f):
1170 fcntl.flock(f, fcntl.LOCK_UN)
1173 class locked_file(object):
1174 def __init__(self, filename, mode, encoding=None):
1175 assert mode in ['r', 'a', 'w']
1176 self.f = io.open(filename, mode, encoding=encoding)
1179 def __enter__(self):
1180 exclusive = self.mode != 'r'
1182 _lock_file(self.f, exclusive)
1188 def __exit__(self, etype, value, traceback):
1190 _unlock_file(self.f)
1197 def write(self, *args):
1198 return self.f.write(*args)
1200 def read(self, *args):
1201 return self.f.read(*args)
1204 def get_filesystem_encoding():
1205 encoding = sys.getfilesystemencoding()
1206 return encoding if encoding is not None else 'utf-8'
1209 def shell_quote(args):
1211 encoding = get_filesystem_encoding()
1213 if isinstance(a, bytes):
1214 # We may get a filename encoded with 'encodeFilename'
1215 a = a.decode(encoding)
1216 quoted_args.append(pipes.quote(a))
1217 return ' '.join(quoted_args)
1220 def smuggle_url(url, data):
1221 """ Pass additional data in a URL for internal use. """
1223 sdata = compat_urllib_parse.urlencode(
1224 {'__youtubedl_smuggle': json.dumps(data)})
1225 return url + '#' + sdata
1228 def unsmuggle_url(smug_url, default=None):
1229 if '#__youtubedl_smuggle' not in smug_url:
1230 return smug_url, default
1231 url, _, sdata = smug_url.rpartition('#')
1232 jsond = compat_parse_qs(sdata)['__youtubedl_smuggle'][0]
1233 data = json.loads(jsond)
1237 def format_bytes(bytes):
1240 if type(bytes) is str:
1241 bytes = float(bytes)
1245 exponent = int(math.log(bytes, 1024.0))
1246 suffix = ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent]
1247 converted = float(bytes) / float(1024 ** exponent)
1248 return '%.2f%s' % (converted, suffix)
1251 def parse_filesize(s):
1255 # The lower-case forms are of course incorrect and inofficial,
1256 # but we support those too
1294 units_re = '|'.join(re.escape(u) for u in _UNIT_TABLE)
1296 r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)' % units_re, s)
1300 num_str = m.group('num').replace(',', '.')
1301 mult = _UNIT_TABLE[m.group('unit')]
1302 return int(float(num_str) * mult)
1305 def month_by_name(name):
1306 """ Return the number of a month by (locale-independently) English name """
1309 return ENGLISH_MONTH_NAMES.index(name) + 1
1314 def month_by_abbreviation(abbrev):
1315 """ Return the number of a month by (locale-independently) English
1319 return [s[:3] for s in ENGLISH_MONTH_NAMES].index(abbrev) + 1
1324 def fix_xml_ampersands(xml_str):
1325 """Replace all the '&' by '&' in XML"""
1327 r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1332 def setproctitle(title):
1333 assert isinstance(title, compat_str)
1335 libc = ctypes.cdll.LoadLibrary("libc.so.6")
1338 title_bytes = title.encode('utf-8')
1339 buf = ctypes.create_string_buffer(len(title_bytes))
1340 buf.value = title_bytes
1342 libc.prctl(15, buf, 0, 0, 0)
1343 except AttributeError:
1344 return # Strange libc, just skip this
1347 def remove_start(s, start):
1348 if s.startswith(start):
1349 return s[len(start):]
1353 def remove_end(s, end):
1355 return s[:-len(end)]
1359 def url_basename(url):
1360 path = compat_urlparse.urlparse(url).path
1361 return path.strip('/').split('/')[-1]
1364 class HEADRequest(compat_urllib_request.Request):
1365 def get_method(self):
1369 def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
1372 v = getattr(v, get_attr, None)
1375 return default if v is None else (int(v) * invscale // scale)
1378 def str_or_none(v, default=None):
1379 return default if v is None else compat_str(v)
1382 def str_to_int(int_str):
1383 """ A more relaxed version of int_or_none """
1386 int_str = re.sub(r'[,\.\+]', '', int_str)
1390 def float_or_none(v, scale=1, invscale=1, default=None):
1391 return default if v is None else (float(v) * invscale / scale)
1394 def parse_duration(s):
1395 if not isinstance(s, compat_basestring):
1403 (?P<only_mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*|
1404 (?P<only_hours>[0-9.]+)\s*(?:hours?)|
1406 \s*(?P<hours_reversed>[0-9]+)\s*(?:[:h]|hours?)\s*(?P<mins_reversed>[0-9]+)\s*(?:[:m]|mins?\.?|minutes?)\s*|
1409 (?:(?P<days>[0-9]+)\s*(?:[:d]|days?)\s*)?
1410 (?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*
1412 (?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*
1414 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?
1419 if m.group('only_mins'):
1420 return float_or_none(m.group('only_mins'), invscale=60)
1421 if m.group('only_hours'):
1422 return float_or_none(m.group('only_hours'), invscale=60 * 60)
1424 res += int(m.group('secs'))
1425 if m.group('mins_reversed'):
1426 res += int(m.group('mins_reversed')) * 60
1428 res += int(m.group('mins')) * 60
1429 if m.group('hours'):
1430 res += int(m.group('hours')) * 60 * 60
1431 if m.group('hours_reversed'):
1432 res += int(m.group('hours_reversed')) * 60 * 60
1434 res += int(m.group('days')) * 24 * 60 * 60
1436 res += float(m.group('ms'))
1440 def prepend_extension(filename, ext, expected_real_ext=None):
1441 name, real_ext = os.path.splitext(filename)
1443 '{0}.{1}{2}'.format(name, ext, real_ext)
1444 if not expected_real_ext or real_ext[1:] == expected_real_ext
1445 else '{0}.{1}'.format(filename, ext))
1448 def replace_extension(filename, ext, expected_real_ext=None):
1449 name, real_ext = os.path.splitext(filename)
1450 return '{0}.{1}'.format(
1451 name if not expected_real_ext or real_ext[1:] == expected_real_ext else filename,
1455 def check_executable(exe, args=[]):
1456 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
1457 args can be a list of arguments for a short output (like -version) """
1459 subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
1465 def get_exe_version(exe, args=['--version'],
1466 version_re=None, unrecognized='present'):
1467 """ Returns the version of the specified executable,
1468 or False if the executable is not present """
1470 out, _ = subprocess.Popen(
1471 [encodeArgument(exe)] + args,
1472 stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
1475 if isinstance(out, bytes): # Python 2.x
1476 out = out.decode('ascii', 'ignore')
1477 return detect_exe_version(out, version_re, unrecognized)
1480 def detect_exe_version(output, version_re=None, unrecognized='present'):
1481 assert isinstance(output, compat_str)
1482 if version_re is None:
1483 version_re = r'version\s+([-0-9._a-zA-Z]+)'
1484 m = re.search(version_re, output)
1491 class PagedList(object):
1493 # This is only useful for tests
1494 return len(self.getslice())
1497 class OnDemandPagedList(PagedList):
1498 def __init__(self, pagefunc, pagesize):
1499 self._pagefunc = pagefunc
1500 self._pagesize = pagesize
1502 def getslice(self, start=0, end=None):
1504 for pagenum in itertools.count(start // self._pagesize):
1505 firstid = pagenum * self._pagesize
1506 nextfirstid = pagenum * self._pagesize + self._pagesize
1507 if start >= nextfirstid:
1510 page_results = list(self._pagefunc(pagenum))
1513 start % self._pagesize
1514 if firstid <= start < nextfirstid
1518 ((end - 1) % self._pagesize) + 1
1519 if (end is not None and firstid <= end <= nextfirstid)
1522 if startv != 0 or endv is not None:
1523 page_results = page_results[startv:endv]
1524 res.extend(page_results)
1526 # A little optimization - if current page is not "full", ie. does
1527 # not contain page_size videos then we can assume that this page
1528 # is the last one - there are no more ids on further pages -
1529 # i.e. no need to query again.
1530 if len(page_results) + startv < self._pagesize:
1533 # If we got the whole page, but the next page is not interesting,
1534 # break out early as well
1535 if end == nextfirstid:
1540 class InAdvancePagedList(PagedList):
1541 def __init__(self, pagefunc, pagecount, pagesize):
1542 self._pagefunc = pagefunc
1543 self._pagecount = pagecount
1544 self._pagesize = pagesize
1546 def getslice(self, start=0, end=None):
1548 start_page = start // self._pagesize
1550 self._pagecount if end is None else (end // self._pagesize + 1))
1551 skip_elems = start - start_page * self._pagesize
1552 only_more = None if end is None else end - start
1553 for pagenum in range(start_page, end_page):
1554 page = list(self._pagefunc(pagenum))
1556 page = page[skip_elems:]
1558 if only_more is not None:
1559 if len(page) < only_more:
1560 only_more -= len(page)
1562 page = page[:only_more]
1569 def uppercase_escape(s):
1570 unicode_escape = codecs.getdecoder('unicode_escape')
1572 r'\\U[0-9a-fA-F]{8}',
1573 lambda m: unicode_escape(m.group(0))[0],
1577 def lowercase_escape(s):
1578 unicode_escape = codecs.getdecoder('unicode_escape')
1580 r'\\u[0-9a-fA-F]{4}',
1581 lambda m: unicode_escape(m.group(0))[0],
1585 def escape_rfc3986(s):
1586 """Escape non-ASCII characters as suggested by RFC 3986"""
1587 if sys.version_info < (3, 0) and isinstance(s, compat_str):
1588 s = s.encode('utf-8')
1589 return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]")
1592 def escape_url(url):
1593 """Escape URL as suggested by RFC 3986"""
1594 url_parsed = compat_urllib_parse_urlparse(url)
1595 return url_parsed._replace(
1596 path=escape_rfc3986(url_parsed.path),
1597 params=escape_rfc3986(url_parsed.params),
1598 query=escape_rfc3986(url_parsed.query),
1599 fragment=escape_rfc3986(url_parsed.fragment)
1603 struct.pack('!I', 0)
1605 # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
1606 def struct_pack(spec, *args):
1607 if isinstance(spec, compat_str):
1608 spec = spec.encode('ascii')
1609 return struct.pack(spec, *args)
1611 def struct_unpack(spec, *args):
1612 if isinstance(spec, compat_str):
1613 spec = spec.encode('ascii')
1614 return struct.unpack(spec, *args)
1616 struct_pack = struct.pack
1617 struct_unpack = struct.unpack
1620 def read_batch_urls(batch_fd):
1622 if not isinstance(url, compat_str):
1623 url = url.decode('utf-8', 'replace')
1624 BOM_UTF8 = '\xef\xbb\xbf'
1625 if url.startswith(BOM_UTF8):
1626 url = url[len(BOM_UTF8):]
1628 if url.startswith(('#', ';', ']')):
1632 with contextlib.closing(batch_fd) as fd:
1633 return [url for url in map(fixup, fd) if url]
1636 def urlencode_postdata(*args, **kargs):
1637 return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
1641 etree_iter = xml.etree.ElementTree.Element.iter
1642 except AttributeError: # Python <=2.6
1643 etree_iter = lambda n: n.findall('.//*')
1647 class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
1648 def doctype(self, name, pubid, system):
1649 pass # Ignore doctypes
1651 parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
1652 kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
1653 tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
1654 # Fix up XML parser in Python 2.x
1655 if sys.version_info < (3, 0):
1656 for n in etree_iter(tree):
1657 if n.text is not None:
1658 if not isinstance(n.text, compat_str):
1659 n.text = n.text.decode('utf-8')
1672 def parse_age_limit(s):
1675 m = re.match(r'^(?P<age>\d{1,2})\+?$', s)
1676 return int(m.group('age')) if m else US_RATINGS.get(s, None)
1679 def strip_jsonp(code):
1681 r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code)
1684 def js_to_json(code):
1687 if v in ('true', 'false', 'null'):
1689 if v.startswith('"'):
1691 if v.startswith("'"):
1693 v = re.sub(r"\\\\|\\'|\"", lambda m: {
1700 res = re.sub(r'''(?x)
1701 "(?:[^"\\]*(?:\\\\|\\['"nu]))*[^"\\]*"|
1702 '(?:[^'\\]*(?:\\\\|\\['"nu]))*[^'\\]*'|
1703 [a-zA-Z_][.a-zA-Z_0-9]*
1705 res = re.sub(r',(\s*[\]}])', lambda m: m.group(1), res)
1709 def qualities(quality_ids):
1710 """ Get a numeric quality value out of a list of possible values """
1713 return quality_ids.index(qid)
1719 DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s'
1722 def limit_length(s, length):
1723 """ Add ellipses to overly long strings """
1728 return s[:length - len(ELLIPSES)] + ELLIPSES
1732 def version_tuple(v):
1733 return tuple(int(e) for e in re.split(r'[-.]', v))
1736 def is_outdated_version(version, limit, assume_new=True):
1738 return not assume_new
1740 return version_tuple(version) < version_tuple(limit)
1742 return not assume_new
1745 def ytdl_is_updateable():
1746 """ Returns if youtube-dl can be updated with -U """
1747 from zipimport import zipimporter
1749 return isinstance(globals().get('__loader__'), zipimporter) or hasattr(sys, 'frozen')
1752 def args_to_str(args):
1753 # Get a short string representation for a subprocess command
1754 return ' '.join(shlex_quote(a) for a in args)
1757 def mimetype2ext(mt):
1758 _, _, res = mt.rpartition('/')
1762 'x-mp4-fragmented': 'mp4',
1767 def urlhandle_detect_ext(url_handle):
1770 getheader = lambda h: url_handle.headers[h]
1771 except AttributeError: # Python < 3
1772 getheader = url_handle.info().getheader
1774 cd = getheader('Content-Disposition')
1776 m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd)
1778 e = determine_ext(m.group('filename'), default_ext=None)
1782 return mimetype2ext(getheader('Content-Type'))
1785 def age_restricted(content_limit, age_limit):
1786 """ Returns True iff the content should be blocked """
1788 if age_limit is None: # No limit set
1790 if content_limit is None:
1791 return False # Content available for everyone
1792 return age_limit < content_limit
1795 def is_html(first_bytes):
1796 """ Detect whether a file contains HTML by examining its first bytes. """
1799 (b'\xef\xbb\xbf', 'utf-8'),
1800 (b'\x00\x00\xfe\xff', 'utf-32-be'),
1801 (b'\xff\xfe\x00\x00', 'utf-32-le'),
1802 (b'\xff\xfe', 'utf-16-le'),
1803 (b'\xfe\xff', 'utf-16-be'),
1805 for bom, enc in BOMS:
1806 if first_bytes.startswith(bom):
1807 s = first_bytes[len(bom):].decode(enc, 'replace')
1810 s = first_bytes.decode('utf-8', 'replace')
1812 return re.match(r'^\s*<', s)
1815 def determine_protocol(info_dict):
1816 protocol = info_dict.get('protocol')
1817 if protocol is not None:
1820 url = info_dict['url']
1821 if url.startswith('rtmp'):
1823 elif url.startswith('mms'):
1825 elif url.startswith('rtsp'):
1828 ext = determine_ext(url)
1834 return compat_urllib_parse_urlparse(url).scheme
1837 def render_table(header_row, data):
1838 """ Render a list of rows, each as a list of values """
1839 table = [header_row] + data
1840 max_lens = [max(len(compat_str(v)) for v in col) for col in zip(*table)]
1841 format_str = ' '.join('%-' + compat_str(ml + 1) + 's' for ml in max_lens[:-1]) + '%s'
1842 return '\n'.join(format_str % tuple(row) for row in table)
1845 def _match_one(filter_part, dct):
1846 COMPARISON_OPERATORS = {
1854 operator_rex = re.compile(r'''(?x)\s*
1856 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1858 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
1859 (?P<strval>(?![0-9.])[a-z0-9A-Z]*)
1862 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
1863 m = operator_rex.search(filter_part)
1865 op = COMPARISON_OPERATORS[m.group('op')]
1866 if m.group('strval') is not None:
1867 if m.group('op') not in ('=', '!='):
1869 'Operator %s does not support string values!' % m.group('op'))
1870 comparison_value = m.group('strval')
1873 comparison_value = int(m.group('intval'))
1875 comparison_value = parse_filesize(m.group('intval'))
1876 if comparison_value is None:
1877 comparison_value = parse_filesize(m.group('intval') + 'B')
1878 if comparison_value is None:
1880 'Invalid integer value %r in filter part %r' % (
1881 m.group('intval'), filter_part))
1882 actual_value = dct.get(m.group('key'))
1883 if actual_value is None:
1884 return m.group('none_inclusive')
1885 return op(actual_value, comparison_value)
1888 '': lambda v: v is not None,
1889 '!': lambda v: v is None,
1891 operator_rex = re.compile(r'''(?x)\s*
1892 (?P<op>%s)\s*(?P<key>[a-z_]+)
1894 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
1895 m = operator_rex.search(filter_part)
1897 op = UNARY_OPERATORS[m.group('op')]
1898 actual_value = dct.get(m.group('key'))
1899 return op(actual_value)
1901 raise ValueError('Invalid filter part %r' % filter_part)
1904 def match_str(filter_str, dct):
1905 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """
1908 _match_one(filter_part, dct) for filter_part in filter_str.split('&'))
1911 def match_filter_func(filter_str):
1912 def _match_func(info_dict):
1913 if match_str(filter_str, info_dict):
1916 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1917 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
1921 def parse_dfxp_time_expr(time_expr):
1925 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
1927 return float(mobj.group('time_offset'))
1929 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:\.\d+)?)$', time_expr)
1931 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3))
1934 def srt_subtitles_timecode(seconds):
1935 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
1938 def dfxp2srt(dfxp_data):
1939 _x = functools.partial(xpath_with_ns, ns_map={
1940 'ttml': 'http://www.w3.org/ns/ttml',
1941 'ttaf1': 'http://www.w3.org/2006/10/ttaf1',
1944 def parse_node(node):
1945 str_or_empty = functools.partial(str_or_none, default='')
1947 out = str_or_empty(node.text)
1950 if child.tag in (_x('ttml:br'), _x('ttaf1:br'), 'br'):
1951 out += '\n' + str_or_empty(child.tail)
1952 elif child.tag in (_x('ttml:span'), _x('ttaf1:span'), 'span'):
1953 out += str_or_empty(parse_node(child))
1955 out += str_or_empty(xml.etree.ElementTree.tostring(child))
1959 dfxp = xml.etree.ElementTree.fromstring(dfxp_data.encode('utf-8'))
1961 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall('.//p')
1964 raise ValueError('Invalid dfxp/TTML subtitle')
1966 for para, index in zip(paras, itertools.count(1)):
1967 begin_time = parse_dfxp_time_expr(para.attrib['begin'])
1968 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
1970 end_time = begin_time + parse_dfxp_time_expr(para.attrib['dur'])
1971 out.append('%d\n%s --> %s\n%s\n\n' % (
1973 srt_subtitles_timecode(begin_time),
1974 srt_subtitles_timecode(end_time),
1980 def cli_option(params, command_option, param):
1981 param = params.get(param)
1982 return [command_option, param] if param is not None else []
1985 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
1986 param = params.get(param)
1987 assert isinstance(param, bool)
1989 return [command_option + separator + (true_value if param else false_value)]
1990 return [command_option, true_value if param else false_value]
1993 def cli_valueless_option(params, command_option, param, expected_value=True):
1994 param = params.get(param)
1995 return [command_option] if param == expected_value else []
1998 def cli_configuration_args(params, param, default=[]):
1999 ex_args = params.get(param)
2002 assert isinstance(ex_args, list)
2006 class ISO639Utils(object):
2007 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
2196 def short2long(cls, code):
2197 """Convert language code from ISO 639-1 to ISO 639-2/T"""
2198 return cls._lang_map.get(code[:2])
2201 def long2short(cls, code):
2202 """Convert language code from ISO 639-2/T to ISO 639-1"""
2203 for short_name, long_name in cls._lang_map.items():
2204 if long_name == code:
2208 class ISO3166Utils(object):
2209 # From http://data.okfn.org/data/core/country-list
2211 'AF': 'Afghanistan',
2212 'AX': 'Åland Islands',
2215 'AS': 'American Samoa',
2220 'AG': 'Antigua and Barbuda',
2237 'BO': 'Bolivia, Plurinational State of',
2238 'BQ': 'Bonaire, Sint Eustatius and Saba',
2239 'BA': 'Bosnia and Herzegovina',
2241 'BV': 'Bouvet Island',
2243 'IO': 'British Indian Ocean Territory',
2244 'BN': 'Brunei Darussalam',
2246 'BF': 'Burkina Faso',
2252 'KY': 'Cayman Islands',
2253 'CF': 'Central African Republic',
2257 'CX': 'Christmas Island',
2258 'CC': 'Cocos (Keeling) Islands',
2262 'CD': 'Congo, the Democratic Republic of the',
2263 'CK': 'Cook Islands',
2265 'CI': 'Côte d\'Ivoire',
2270 'CZ': 'Czech Republic',
2274 'DO': 'Dominican Republic',
2277 'SV': 'El Salvador',
2278 'GQ': 'Equatorial Guinea',
2282 'FK': 'Falkland Islands (Malvinas)',
2283 'FO': 'Faroe Islands',
2287 'GF': 'French Guiana',
2288 'PF': 'French Polynesia',
2289 'TF': 'French Southern Territories',
2304 'GW': 'Guinea-Bissau',
2307 'HM': 'Heard Island and McDonald Islands',
2308 'VA': 'Holy See (Vatican City State)',
2315 'IR': 'Iran, Islamic Republic of',
2318 'IM': 'Isle of Man',
2328 'KP': 'Korea, Democratic People\'s Republic of',
2329 'KR': 'Korea, Republic of',
2332 'LA': 'Lao People\'s Democratic Republic',
2338 'LI': 'Liechtenstein',
2342 'MK': 'Macedonia, the Former Yugoslav Republic of',
2349 'MH': 'Marshall Islands',
2355 'FM': 'Micronesia, Federated States of',
2356 'MD': 'Moldova, Republic of',
2367 'NL': 'Netherlands',
2368 'NC': 'New Caledonia',
2369 'NZ': 'New Zealand',
2374 'NF': 'Norfolk Island',
2375 'MP': 'Northern Mariana Islands',
2380 'PS': 'Palestine, State of',
2382 'PG': 'Papua New Guinea',
2385 'PH': 'Philippines',
2389 'PR': 'Puerto Rico',
2393 'RU': 'Russian Federation',
2395 'BL': 'Saint Barthélemy',
2396 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
2397 'KN': 'Saint Kitts and Nevis',
2398 'LC': 'Saint Lucia',
2399 'MF': 'Saint Martin (French part)',
2400 'PM': 'Saint Pierre and Miquelon',
2401 'VC': 'Saint Vincent and the Grenadines',
2404 'ST': 'Sao Tome and Principe',
2405 'SA': 'Saudi Arabia',
2409 'SL': 'Sierra Leone',
2411 'SX': 'Sint Maarten (Dutch part)',
2414 'SB': 'Solomon Islands',
2416 'ZA': 'South Africa',
2417 'GS': 'South Georgia and the South Sandwich Islands',
2418 'SS': 'South Sudan',
2423 'SJ': 'Svalbard and Jan Mayen',
2426 'CH': 'Switzerland',
2427 'SY': 'Syrian Arab Republic',
2428 'TW': 'Taiwan, Province of China',
2430 'TZ': 'Tanzania, United Republic of',
2432 'TL': 'Timor-Leste',
2436 'TT': 'Trinidad and Tobago',
2439 'TM': 'Turkmenistan',
2440 'TC': 'Turks and Caicos Islands',
2444 'AE': 'United Arab Emirates',
2445 'GB': 'United Kingdom',
2446 'US': 'United States',
2447 'UM': 'United States Minor Outlying Islands',
2451 'VE': 'Venezuela, Bolivarian Republic of',
2453 'VG': 'Virgin Islands, British',
2454 'VI': 'Virgin Islands, U.S.',
2455 'WF': 'Wallis and Futuna',
2456 'EH': 'Western Sahara',
2463 def short2full(cls, code):
2464 """Convert an ISO 3166-2 country code to the corresponding full name"""
2465 return cls._country_map.get(code.upper())
2468 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
2469 def __init__(self, proxies=None):
2470 # Set default handlers
2471 for type in ('http', 'https'):
2472 setattr(self, '%s_open' % type,
2473 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
2474 meth(r, proxy, type))
2475 return compat_urllib_request.ProxyHandler.__init__(self, proxies)
2477 def proxy_open(self, req, proxy, type):
2478 req_proxy = req.headers.get('Ytdl-request-proxy')
2479 if req_proxy is not None:
2481 del req.headers['Ytdl-request-proxy']
2483 if proxy == '__noproxy__':
2484 return None # No Proxy
2485 return compat_urllib_request.ProxyHandler.proxy_open(
2486 self, req, proxy, type)