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 = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
801 # if set_cookie != set_cookie_escaped:
802 # del response.headers[set_cookie_header]
803 # response.headers[set_cookie_header] = set_cookie_escaped
804 return compat_urllib_request.HTTPCookieProcessor.http_response(self, request, response)
806 https_request = compat_urllib_request.HTTPCookieProcessor.http_request
807 https_response = http_response
810 def parse_iso8601(date_str, delimiter='T', timezone=None):
811 """ Return a UNIX timestamp from the given date """
818 r'(\.[0-9]+)?(?:Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
821 timezone = datetime.timedelta()
823 date_str = date_str[:-len(m.group(0))]
824 if not m.group('sign'):
825 timezone = datetime.timedelta()
827 sign = 1 if m.group('sign') == '+' else -1
828 timezone = datetime.timedelta(
829 hours=sign * int(m.group('hours')),
830 minutes=sign * int(m.group('minutes')))
831 date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
832 dt = datetime.datetime.strptime(date_str, date_format) - timezone
833 return calendar.timegm(dt.timetuple())
836 def unified_strdate(date_str, day_first=True):
837 """Return a string with the date in the format YYYYMMDD"""
843 date_str = date_str.replace(',', ' ')
844 # %z (UTC offset) is only supported in python>=3.2
845 if not re.match(r'^[0-9]{1,2}-[0-9]{1,2}-[0-9]{4}$', date_str):
846 date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
847 # Remove AM/PM + timezone
848 date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
850 format_expressions = [
855 '%b %dst %Y %I:%M%p',
856 '%b %dnd %Y %I:%M%p',
857 '%b %dth %Y %I:%M%p',
863 '%Y-%m-%d %H:%M:%S.%f',
866 '%Y-%m-%dT%H:%M:%SZ',
867 '%Y-%m-%dT%H:%M:%S.%fZ',
868 '%Y-%m-%dT%H:%M:%S.%f0Z',
870 '%Y-%m-%dT%H:%M:%S.%f',
874 format_expressions.extend([
882 format_expressions.extend([
889 for expression in format_expressions:
891 upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
894 if upload_date is None:
895 timetuple = email.utils.parsedate_tz(date_str)
897 upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
901 def determine_ext(url, default_ext='unknown_video'):
904 guess = url.partition('?')[0].rpartition('.')[2]
905 if re.match(r'^[A-Za-z0-9]+$', guess):
911 def subtitles_filename(filename, sub_lang, sub_format):
912 return filename.rsplit('.', 1)[0] + '.' + sub_lang + '.' + sub_format
915 def date_from_str(date_str):
917 Return a datetime object from a string in the format YYYYMMDD or
918 (now|today)[+-][0-9](day|week|month|year)(s)?"""
919 today = datetime.date.today()
920 if date_str in ('now', 'today'):
922 if date_str == 'yesterday':
923 return today - datetime.timedelta(days=1)
924 match = re.match('(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
925 if match is not None:
926 sign = match.group('sign')
927 time = int(match.group('time'))
930 unit = match.group('unit')
931 # A bad aproximation?
939 delta = datetime.timedelta(**{unit: time})
941 return datetime.datetime.strptime(date_str, "%Y%m%d").date()
944 def hyphenate_date(date_str):
946 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
947 match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str)
948 if match is not None:
949 return '-'.join(match.groups())
954 class DateRange(object):
955 """Represents a time interval between two dates"""
957 def __init__(self, start=None, end=None):
958 """start and end must be strings in the format accepted by date"""
959 if start is not None:
960 self.start = date_from_str(start)
962 self.start = datetime.datetime.min.date()
964 self.end = date_from_str(end)
966 self.end = datetime.datetime.max.date()
967 if self.start > self.end:
968 raise ValueError('Date range: "%s" , the start date must be before the end date' % self)
972 """Returns a range that only contains the given day"""
975 def __contains__(self, date):
976 """Check if the date is in the range"""
977 if not isinstance(date, datetime.date):
978 date = date_from_str(date)
979 return self.start <= date <= self.end
982 return '%s - %s' % (self.start.isoformat(), self.end.isoformat())
986 """ Returns the platform name as a compat_str """
987 res = platform.platform()
988 if isinstance(res, bytes):
989 res = res.decode(preferredencoding())
991 assert isinstance(res, compat_str)
995 def _windows_write_string(s, out):
996 """ Returns True if the string was written using special methods,
997 False if it has yet to be written out."""
998 # Adapted from http://stackoverflow.com/a/3259271/35070
1001 import ctypes.wintypes
1009 fileno = out.fileno()
1010 except AttributeError:
1011 # If the output stream doesn't have a fileno, it's virtual
1013 except io.UnsupportedOperation:
1014 # Some strange Windows pseudo files?
1016 if fileno not in WIN_OUTPUT_IDS:
1019 GetStdHandle = ctypes.WINFUNCTYPE(
1020 ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
1021 (b"GetStdHandle", ctypes.windll.kernel32))
1022 h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
1024 WriteConsoleW = ctypes.WINFUNCTYPE(
1025 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
1026 ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
1027 ctypes.wintypes.LPVOID)((b"WriteConsoleW", ctypes.windll.kernel32))
1028 written = ctypes.wintypes.DWORD(0)
1030 GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b"GetFileType", ctypes.windll.kernel32))
1031 FILE_TYPE_CHAR = 0x0002
1032 FILE_TYPE_REMOTE = 0x8000
1033 GetConsoleMode = ctypes.WINFUNCTYPE(
1034 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
1035 ctypes.POINTER(ctypes.wintypes.DWORD))(
1036 (b"GetConsoleMode", ctypes.windll.kernel32))
1037 INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
1039 def not_a_console(handle):
1040 if handle == INVALID_HANDLE_VALUE or handle is None:
1042 return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR or
1043 GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0)
1045 if not_a_console(h):
1048 def next_nonbmp_pos(s):
1050 return next(i for i, c in enumerate(s) if ord(c) > 0xffff)
1051 except StopIteration:
1055 count = min(next_nonbmp_pos(s), 1024)
1057 ret = WriteConsoleW(
1058 h, s, count if count else 2, ctypes.byref(written), None)
1060 raise OSError('Failed to write string')
1061 if not count: # We just wrote a non-BMP character
1062 assert written.value == 2
1065 assert written.value > 0
1066 s = s[written.value:]
1070 def write_string(s, out=None, encoding=None):
1073 assert type(s) == compat_str
1075 if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'):
1076 if _windows_write_string(s, out):
1079 if ('b' in getattr(out, 'mode', '') or
1080 sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr
1081 byt = s.encode(encoding or preferredencoding(), 'ignore')
1083 elif hasattr(out, 'buffer'):
1084 enc = encoding or getattr(out, 'encoding', None) or preferredencoding()
1085 byt = s.encode(enc, 'ignore')
1086 out.buffer.write(byt)
1092 def bytes_to_intlist(bs):
1095 if isinstance(bs[0], int): # Python 3
1098 return [ord(c) for c in bs]
1101 def intlist_to_bytes(xs):
1104 return struct_pack('%dB' % len(xs), *xs)
1107 # Cross-platform file locking
1108 if sys.platform == 'win32':
1109 import ctypes.wintypes
1112 class OVERLAPPED(ctypes.Structure):
1114 ('Internal', ctypes.wintypes.LPVOID),
1115 ('InternalHigh', ctypes.wintypes.LPVOID),
1116 ('Offset', ctypes.wintypes.DWORD),
1117 ('OffsetHigh', ctypes.wintypes.DWORD),
1118 ('hEvent', ctypes.wintypes.HANDLE),
1121 kernel32 = ctypes.windll.kernel32
1122 LockFileEx = kernel32.LockFileEx
1123 LockFileEx.argtypes = [
1124 ctypes.wintypes.HANDLE, # hFile
1125 ctypes.wintypes.DWORD, # dwFlags
1126 ctypes.wintypes.DWORD, # dwReserved
1127 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1128 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1129 ctypes.POINTER(OVERLAPPED) # Overlapped
1131 LockFileEx.restype = ctypes.wintypes.BOOL
1132 UnlockFileEx = kernel32.UnlockFileEx
1133 UnlockFileEx.argtypes = [
1134 ctypes.wintypes.HANDLE, # hFile
1135 ctypes.wintypes.DWORD, # dwReserved
1136 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1137 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1138 ctypes.POINTER(OVERLAPPED) # Overlapped
1140 UnlockFileEx.restype = ctypes.wintypes.BOOL
1141 whole_low = 0xffffffff
1142 whole_high = 0x7fffffff
1144 def _lock_file(f, exclusive):
1145 overlapped = OVERLAPPED()
1146 overlapped.Offset = 0
1147 overlapped.OffsetHigh = 0
1148 overlapped.hEvent = 0
1149 f._lock_file_overlapped_p = ctypes.pointer(overlapped)
1150 handle = msvcrt.get_osfhandle(f.fileno())
1151 if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0,
1152 whole_low, whole_high, f._lock_file_overlapped_p):
1153 raise OSError('Locking file failed: %r' % ctypes.FormatError())
1155 def _unlock_file(f):
1156 assert f._lock_file_overlapped_p
1157 handle = msvcrt.get_osfhandle(f.fileno())
1158 if not UnlockFileEx(handle, 0,
1159 whole_low, whole_high, f._lock_file_overlapped_p):
1160 raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
1165 def _lock_file(f, exclusive):
1166 fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
1168 def _unlock_file(f):
1169 fcntl.flock(f, fcntl.LOCK_UN)
1172 class locked_file(object):
1173 def __init__(self, filename, mode, encoding=None):
1174 assert mode in ['r', 'a', 'w']
1175 self.f = io.open(filename, mode, encoding=encoding)
1178 def __enter__(self):
1179 exclusive = self.mode != 'r'
1181 _lock_file(self.f, exclusive)
1187 def __exit__(self, etype, value, traceback):
1189 _unlock_file(self.f)
1196 def write(self, *args):
1197 return self.f.write(*args)
1199 def read(self, *args):
1200 return self.f.read(*args)
1203 def get_filesystem_encoding():
1204 encoding = sys.getfilesystemencoding()
1205 return encoding if encoding is not None else 'utf-8'
1208 def shell_quote(args):
1210 encoding = get_filesystem_encoding()
1212 if isinstance(a, bytes):
1213 # We may get a filename encoded with 'encodeFilename'
1214 a = a.decode(encoding)
1215 quoted_args.append(pipes.quote(a))
1216 return ' '.join(quoted_args)
1219 def smuggle_url(url, data):
1220 """ Pass additional data in a URL for internal use. """
1222 sdata = compat_urllib_parse.urlencode(
1223 {'__youtubedl_smuggle': json.dumps(data)})
1224 return url + '#' + sdata
1227 def unsmuggle_url(smug_url, default=None):
1228 if '#__youtubedl_smuggle' not in smug_url:
1229 return smug_url, default
1230 url, _, sdata = smug_url.rpartition('#')
1231 jsond = compat_parse_qs(sdata)['__youtubedl_smuggle'][0]
1232 data = json.loads(jsond)
1236 def format_bytes(bytes):
1239 if type(bytes) is str:
1240 bytes = float(bytes)
1244 exponent = int(math.log(bytes, 1024.0))
1245 suffix = ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent]
1246 converted = float(bytes) / float(1024 ** exponent)
1247 return '%.2f%s' % (converted, suffix)
1250 def parse_filesize(s):
1254 # The lower-case forms are of course incorrect and inofficial,
1255 # but we support those too
1293 units_re = '|'.join(re.escape(u) for u in _UNIT_TABLE)
1295 r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)' % units_re, s)
1299 num_str = m.group('num').replace(',', '.')
1300 mult = _UNIT_TABLE[m.group('unit')]
1301 return int(float(num_str) * mult)
1304 def month_by_name(name):
1305 """ Return the number of a month by (locale-independently) English name """
1308 return ENGLISH_MONTH_NAMES.index(name) + 1
1313 def month_by_abbreviation(abbrev):
1314 """ Return the number of a month by (locale-independently) English
1318 return [s[:3] for s in ENGLISH_MONTH_NAMES].index(abbrev) + 1
1323 def fix_xml_ampersands(xml_str):
1324 """Replace all the '&' by '&' in XML"""
1326 r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1331 def setproctitle(title):
1332 assert isinstance(title, compat_str)
1334 libc = ctypes.cdll.LoadLibrary("libc.so.6")
1337 title_bytes = title.encode('utf-8')
1338 buf = ctypes.create_string_buffer(len(title_bytes))
1339 buf.value = title_bytes
1341 libc.prctl(15, buf, 0, 0, 0)
1342 except AttributeError:
1343 return # Strange libc, just skip this
1346 def remove_start(s, start):
1347 if s.startswith(start):
1348 return s[len(start):]
1352 def remove_end(s, end):
1354 return s[:-len(end)]
1358 def url_basename(url):
1359 path = compat_urlparse.urlparse(url).path
1360 return path.strip('/').split('/')[-1]
1363 class HEADRequest(compat_urllib_request.Request):
1364 def get_method(self):
1368 def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
1371 v = getattr(v, get_attr, None)
1374 return default if v is None else (int(v) * invscale // scale)
1377 def str_or_none(v, default=None):
1378 return default if v is None else compat_str(v)
1381 def str_to_int(int_str):
1382 """ A more relaxed version of int_or_none """
1385 int_str = re.sub(r'[,\.\+]', '', int_str)
1389 def float_or_none(v, scale=1, invscale=1, default=None):
1390 return default if v is None else (float(v) * invscale / scale)
1393 def parse_duration(s):
1394 if not isinstance(s, compat_basestring):
1402 (?P<only_mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*|
1403 (?P<only_hours>[0-9.]+)\s*(?:hours?)|
1405 \s*(?P<hours_reversed>[0-9]+)\s*(?:[:h]|hours?)\s*(?P<mins_reversed>[0-9]+)\s*(?:[:m]|mins?\.?|minutes?)\s*|
1408 (?:(?P<days>[0-9]+)\s*(?:[:d]|days?)\s*)?
1409 (?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*
1411 (?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*
1413 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?
1418 if m.group('only_mins'):
1419 return float_or_none(m.group('only_mins'), invscale=60)
1420 if m.group('only_hours'):
1421 return float_or_none(m.group('only_hours'), invscale=60 * 60)
1423 res += int(m.group('secs'))
1424 if m.group('mins_reversed'):
1425 res += int(m.group('mins_reversed')) * 60
1427 res += int(m.group('mins')) * 60
1428 if m.group('hours'):
1429 res += int(m.group('hours')) * 60 * 60
1430 if m.group('hours_reversed'):
1431 res += int(m.group('hours_reversed')) * 60 * 60
1433 res += int(m.group('days')) * 24 * 60 * 60
1435 res += float(m.group('ms'))
1439 def prepend_extension(filename, ext, expected_real_ext=None):
1440 name, real_ext = os.path.splitext(filename)
1442 '{0}.{1}{2}'.format(name, ext, real_ext)
1443 if not expected_real_ext or real_ext[1:] == expected_real_ext
1444 else '{0}.{1}'.format(filename, ext))
1447 def replace_extension(filename, ext, expected_real_ext=None):
1448 name, real_ext = os.path.splitext(filename)
1449 return '{0}.{1}'.format(
1450 name if not expected_real_ext or real_ext[1:] == expected_real_ext else filename,
1454 def check_executable(exe, args=[]):
1455 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
1456 args can be a list of arguments for a short output (like -version) """
1458 subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
1464 def get_exe_version(exe, args=['--version'],
1465 version_re=None, unrecognized='present'):
1466 """ Returns the version of the specified executable,
1467 or False if the executable is not present """
1469 out, _ = subprocess.Popen(
1470 [encodeArgument(exe)] + args,
1471 stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
1474 if isinstance(out, bytes): # Python 2.x
1475 out = out.decode('ascii', 'ignore')
1476 return detect_exe_version(out, version_re, unrecognized)
1479 def detect_exe_version(output, version_re=None, unrecognized='present'):
1480 assert isinstance(output, compat_str)
1481 if version_re is None:
1482 version_re = r'version\s+([-0-9._a-zA-Z]+)'
1483 m = re.search(version_re, output)
1490 class PagedList(object):
1492 # This is only useful for tests
1493 return len(self.getslice())
1496 class OnDemandPagedList(PagedList):
1497 def __init__(self, pagefunc, pagesize):
1498 self._pagefunc = pagefunc
1499 self._pagesize = pagesize
1501 def getslice(self, start=0, end=None):
1503 for pagenum in itertools.count(start // self._pagesize):
1504 firstid = pagenum * self._pagesize
1505 nextfirstid = pagenum * self._pagesize + self._pagesize
1506 if start >= nextfirstid:
1509 page_results = list(self._pagefunc(pagenum))
1512 start % self._pagesize
1513 if firstid <= start < nextfirstid
1517 ((end - 1) % self._pagesize) + 1
1518 if (end is not None and firstid <= end <= nextfirstid)
1521 if startv != 0 or endv is not None:
1522 page_results = page_results[startv:endv]
1523 res.extend(page_results)
1525 # A little optimization - if current page is not "full", ie. does
1526 # not contain page_size videos then we can assume that this page
1527 # is the last one - there are no more ids on further pages -
1528 # i.e. no need to query again.
1529 if len(page_results) + startv < self._pagesize:
1532 # If we got the whole page, but the next page is not interesting,
1533 # break out early as well
1534 if end == nextfirstid:
1539 class InAdvancePagedList(PagedList):
1540 def __init__(self, pagefunc, pagecount, pagesize):
1541 self._pagefunc = pagefunc
1542 self._pagecount = pagecount
1543 self._pagesize = pagesize
1545 def getslice(self, start=0, end=None):
1547 start_page = start // self._pagesize
1549 self._pagecount if end is None else (end // self._pagesize + 1))
1550 skip_elems = start - start_page * self._pagesize
1551 only_more = None if end is None else end - start
1552 for pagenum in range(start_page, end_page):
1553 page = list(self._pagefunc(pagenum))
1555 page = page[skip_elems:]
1557 if only_more is not None:
1558 if len(page) < only_more:
1559 only_more -= len(page)
1561 page = page[:only_more]
1568 def uppercase_escape(s):
1569 unicode_escape = codecs.getdecoder('unicode_escape')
1571 r'\\U[0-9a-fA-F]{8}',
1572 lambda m: unicode_escape(m.group(0))[0],
1576 def lowercase_escape(s):
1577 unicode_escape = codecs.getdecoder('unicode_escape')
1579 r'\\u[0-9a-fA-F]{4}',
1580 lambda m: unicode_escape(m.group(0))[0],
1584 def escape_rfc3986(s):
1585 """Escape non-ASCII characters as suggested by RFC 3986"""
1586 if sys.version_info < (3, 0) and isinstance(s, compat_str):
1587 s = s.encode('utf-8')
1588 return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]")
1591 def escape_url(url):
1592 """Escape URL as suggested by RFC 3986"""
1593 url_parsed = compat_urllib_parse_urlparse(url)
1594 return url_parsed._replace(
1595 path=escape_rfc3986(url_parsed.path),
1596 params=escape_rfc3986(url_parsed.params),
1597 query=escape_rfc3986(url_parsed.query),
1598 fragment=escape_rfc3986(url_parsed.fragment)
1602 struct.pack('!I', 0)
1604 # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
1605 def struct_pack(spec, *args):
1606 if isinstance(spec, compat_str):
1607 spec = spec.encode('ascii')
1608 return struct.pack(spec, *args)
1610 def struct_unpack(spec, *args):
1611 if isinstance(spec, compat_str):
1612 spec = spec.encode('ascii')
1613 return struct.unpack(spec, *args)
1615 struct_pack = struct.pack
1616 struct_unpack = struct.unpack
1619 def read_batch_urls(batch_fd):
1621 if not isinstance(url, compat_str):
1622 url = url.decode('utf-8', 'replace')
1623 BOM_UTF8 = '\xef\xbb\xbf'
1624 if url.startswith(BOM_UTF8):
1625 url = url[len(BOM_UTF8):]
1627 if url.startswith(('#', ';', ']')):
1631 with contextlib.closing(batch_fd) as fd:
1632 return [url for url in map(fixup, fd) if url]
1635 def urlencode_postdata(*args, **kargs):
1636 return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
1639 def encode_dict(d, encoding='utf-8'):
1640 return dict((k.encode(encoding), v.encode(encoding)) for k, v in d.items())
1644 etree_iter = xml.etree.ElementTree.Element.iter
1645 except AttributeError: # Python <=2.6
1646 etree_iter = lambda n: n.findall('.//*')
1650 class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
1651 def doctype(self, name, pubid, system):
1652 pass # Ignore doctypes
1654 parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
1655 kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
1656 tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
1657 # Fix up XML parser in Python 2.x
1658 if sys.version_info < (3, 0):
1659 for n in etree_iter(tree):
1660 if n.text is not None:
1661 if not isinstance(n.text, compat_str):
1662 n.text = n.text.decode('utf-8')
1675 def parse_age_limit(s):
1678 m = re.match(r'^(?P<age>\d{1,2})\+?$', s)
1679 return int(m.group('age')) if m else US_RATINGS.get(s, None)
1682 def strip_jsonp(code):
1684 r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code)
1687 def js_to_json(code):
1690 if v in ('true', 'false', 'null'):
1692 if v.startswith('"'):
1694 if v.startswith("'"):
1696 v = re.sub(r"\\\\|\\'|\"", lambda m: {
1703 res = re.sub(r'''(?x)
1704 "(?:[^"\\]*(?:\\\\|\\['"nu]))*[^"\\]*"|
1705 '(?:[^'\\]*(?:\\\\|\\['"nu]))*[^'\\]*'|
1706 [a-zA-Z_][.a-zA-Z_0-9]*
1708 res = re.sub(r',(\s*[\]}])', lambda m: m.group(1), res)
1712 def qualities(quality_ids):
1713 """ Get a numeric quality value out of a list of possible values """
1716 return quality_ids.index(qid)
1722 DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s'
1725 def limit_length(s, length):
1726 """ Add ellipses to overly long strings """
1731 return s[:length - len(ELLIPSES)] + ELLIPSES
1735 def version_tuple(v):
1736 return tuple(int(e) for e in re.split(r'[-.]', v))
1739 def is_outdated_version(version, limit, assume_new=True):
1741 return not assume_new
1743 return version_tuple(version) < version_tuple(limit)
1745 return not assume_new
1748 def ytdl_is_updateable():
1749 """ Returns if youtube-dl can be updated with -U """
1750 from zipimport import zipimporter
1752 return isinstance(globals().get('__loader__'), zipimporter) or hasattr(sys, 'frozen')
1755 def args_to_str(args):
1756 # Get a short string representation for a subprocess command
1757 return ' '.join(shlex_quote(a) for a in args)
1760 def mimetype2ext(mt):
1761 _, _, res = mt.rpartition('/')
1765 'x-mp4-fragmented': 'mp4',
1770 def urlhandle_detect_ext(url_handle):
1773 getheader = lambda h: url_handle.headers[h]
1774 except AttributeError: # Python < 3
1775 getheader = url_handle.info().getheader
1777 cd = getheader('Content-Disposition')
1779 m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd)
1781 e = determine_ext(m.group('filename'), default_ext=None)
1785 return mimetype2ext(getheader('Content-Type'))
1788 def age_restricted(content_limit, age_limit):
1789 """ Returns True iff the content should be blocked """
1791 if age_limit is None: # No limit set
1793 if content_limit is None:
1794 return False # Content available for everyone
1795 return age_limit < content_limit
1798 def is_html(first_bytes):
1799 """ Detect whether a file contains HTML by examining its first bytes. """
1802 (b'\xef\xbb\xbf', 'utf-8'),
1803 (b'\x00\x00\xfe\xff', 'utf-32-be'),
1804 (b'\xff\xfe\x00\x00', 'utf-32-le'),
1805 (b'\xff\xfe', 'utf-16-le'),
1806 (b'\xfe\xff', 'utf-16-be'),
1808 for bom, enc in BOMS:
1809 if first_bytes.startswith(bom):
1810 s = first_bytes[len(bom):].decode(enc, 'replace')
1813 s = first_bytes.decode('utf-8', 'replace')
1815 return re.match(r'^\s*<', s)
1818 def determine_protocol(info_dict):
1819 protocol = info_dict.get('protocol')
1820 if protocol is not None:
1823 url = info_dict['url']
1824 if url.startswith('rtmp'):
1826 elif url.startswith('mms'):
1828 elif url.startswith('rtsp'):
1831 ext = determine_ext(url)
1837 return compat_urllib_parse_urlparse(url).scheme
1840 def render_table(header_row, data):
1841 """ Render a list of rows, each as a list of values """
1842 table = [header_row] + data
1843 max_lens = [max(len(compat_str(v)) for v in col) for col in zip(*table)]
1844 format_str = ' '.join('%-' + compat_str(ml + 1) + 's' for ml in max_lens[:-1]) + '%s'
1845 return '\n'.join(format_str % tuple(row) for row in table)
1848 def _match_one(filter_part, dct):
1849 COMPARISON_OPERATORS = {
1857 operator_rex = re.compile(r'''(?x)\s*
1859 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1861 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
1862 (?P<strval>(?![0-9.])[a-z0-9A-Z]*)
1865 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
1866 m = operator_rex.search(filter_part)
1868 op = COMPARISON_OPERATORS[m.group('op')]
1869 if m.group('strval') is not None:
1870 if m.group('op') not in ('=', '!='):
1872 'Operator %s does not support string values!' % m.group('op'))
1873 comparison_value = m.group('strval')
1876 comparison_value = int(m.group('intval'))
1878 comparison_value = parse_filesize(m.group('intval'))
1879 if comparison_value is None:
1880 comparison_value = parse_filesize(m.group('intval') + 'B')
1881 if comparison_value is None:
1883 'Invalid integer value %r in filter part %r' % (
1884 m.group('intval'), filter_part))
1885 actual_value = dct.get(m.group('key'))
1886 if actual_value is None:
1887 return m.group('none_inclusive')
1888 return op(actual_value, comparison_value)
1891 '': lambda v: v is not None,
1892 '!': lambda v: v is None,
1894 operator_rex = re.compile(r'''(?x)\s*
1895 (?P<op>%s)\s*(?P<key>[a-z_]+)
1897 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
1898 m = operator_rex.search(filter_part)
1900 op = UNARY_OPERATORS[m.group('op')]
1901 actual_value = dct.get(m.group('key'))
1902 return op(actual_value)
1904 raise ValueError('Invalid filter part %r' % filter_part)
1907 def match_str(filter_str, dct):
1908 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """
1911 _match_one(filter_part, dct) for filter_part in filter_str.split('&'))
1914 def match_filter_func(filter_str):
1915 def _match_func(info_dict):
1916 if match_str(filter_str, info_dict):
1919 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1920 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
1924 def parse_dfxp_time_expr(time_expr):
1928 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
1930 return float(mobj.group('time_offset'))
1932 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:\.\d+)?)$', time_expr)
1934 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3))
1937 def srt_subtitles_timecode(seconds):
1938 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
1941 def dfxp2srt(dfxp_data):
1942 _x = functools.partial(xpath_with_ns, ns_map={
1943 'ttml': 'http://www.w3.org/ns/ttml',
1944 'ttaf1': 'http://www.w3.org/2006/10/ttaf1',
1947 def parse_node(node):
1948 str_or_empty = functools.partial(str_or_none, default='')
1950 out = str_or_empty(node.text)
1953 if child.tag in (_x('ttml:br'), _x('ttaf1:br'), 'br'):
1954 out += '\n' + str_or_empty(child.tail)
1955 elif child.tag in (_x('ttml:span'), _x('ttaf1:span'), 'span'):
1956 out += str_or_empty(parse_node(child))
1958 out += str_or_empty(xml.etree.ElementTree.tostring(child))
1962 dfxp = xml.etree.ElementTree.fromstring(dfxp_data.encode('utf-8'))
1964 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall('.//p')
1967 raise ValueError('Invalid dfxp/TTML subtitle')
1969 for para, index in zip(paras, itertools.count(1)):
1970 begin_time = parse_dfxp_time_expr(para.attrib['begin'])
1971 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
1973 end_time = begin_time + parse_dfxp_time_expr(para.attrib['dur'])
1974 out.append('%d\n%s --> %s\n%s\n\n' % (
1976 srt_subtitles_timecode(begin_time),
1977 srt_subtitles_timecode(end_time),
1983 def cli_option(params, command_option, param):
1984 param = params.get(param)
1985 return [command_option, param] if param is not None else []
1988 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
1989 param = params.get(param)
1990 assert isinstance(param, bool)
1992 return [command_option + separator + (true_value if param else false_value)]
1993 return [command_option, true_value if param else false_value]
1996 def cli_valueless_option(params, command_option, param, expected_value=True):
1997 param = params.get(param)
1998 return [command_option] if param == expected_value else []
2001 def cli_configuration_args(params, param, default=[]):
2002 ex_args = params.get(param)
2005 assert isinstance(ex_args, list)
2009 class ISO639Utils(object):
2010 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
2199 def short2long(cls, code):
2200 """Convert language code from ISO 639-1 to ISO 639-2/T"""
2201 return cls._lang_map.get(code[:2])
2204 def long2short(cls, code):
2205 """Convert language code from ISO 639-2/T to ISO 639-1"""
2206 for short_name, long_name in cls._lang_map.items():
2207 if long_name == code:
2211 class ISO3166Utils(object):
2212 # From http://data.okfn.org/data/core/country-list
2214 'AF': 'Afghanistan',
2215 'AX': 'Åland Islands',
2218 'AS': 'American Samoa',
2223 'AG': 'Antigua and Barbuda',
2240 'BO': 'Bolivia, Plurinational State of',
2241 'BQ': 'Bonaire, Sint Eustatius and Saba',
2242 'BA': 'Bosnia and Herzegovina',
2244 'BV': 'Bouvet Island',
2246 'IO': 'British Indian Ocean Territory',
2247 'BN': 'Brunei Darussalam',
2249 'BF': 'Burkina Faso',
2255 'KY': 'Cayman Islands',
2256 'CF': 'Central African Republic',
2260 'CX': 'Christmas Island',
2261 'CC': 'Cocos (Keeling) Islands',
2265 'CD': 'Congo, the Democratic Republic of the',
2266 'CK': 'Cook Islands',
2268 'CI': 'Côte d\'Ivoire',
2273 'CZ': 'Czech Republic',
2277 'DO': 'Dominican Republic',
2280 'SV': 'El Salvador',
2281 'GQ': 'Equatorial Guinea',
2285 'FK': 'Falkland Islands (Malvinas)',
2286 'FO': 'Faroe Islands',
2290 'GF': 'French Guiana',
2291 'PF': 'French Polynesia',
2292 'TF': 'French Southern Territories',
2307 'GW': 'Guinea-Bissau',
2310 'HM': 'Heard Island and McDonald Islands',
2311 'VA': 'Holy See (Vatican City State)',
2318 'IR': 'Iran, Islamic Republic of',
2321 'IM': 'Isle of Man',
2331 'KP': 'Korea, Democratic People\'s Republic of',
2332 'KR': 'Korea, Republic of',
2335 'LA': 'Lao People\'s Democratic Republic',
2341 'LI': 'Liechtenstein',
2345 'MK': 'Macedonia, the Former Yugoslav Republic of',
2352 'MH': 'Marshall Islands',
2358 'FM': 'Micronesia, Federated States of',
2359 'MD': 'Moldova, Republic of',
2370 'NL': 'Netherlands',
2371 'NC': 'New Caledonia',
2372 'NZ': 'New Zealand',
2377 'NF': 'Norfolk Island',
2378 'MP': 'Northern Mariana Islands',
2383 'PS': 'Palestine, State of',
2385 'PG': 'Papua New Guinea',
2388 'PH': 'Philippines',
2392 'PR': 'Puerto Rico',
2396 'RU': 'Russian Federation',
2398 'BL': 'Saint Barthélemy',
2399 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
2400 'KN': 'Saint Kitts and Nevis',
2401 'LC': 'Saint Lucia',
2402 'MF': 'Saint Martin (French part)',
2403 'PM': 'Saint Pierre and Miquelon',
2404 'VC': 'Saint Vincent and the Grenadines',
2407 'ST': 'Sao Tome and Principe',
2408 'SA': 'Saudi Arabia',
2412 'SL': 'Sierra Leone',
2414 'SX': 'Sint Maarten (Dutch part)',
2417 'SB': 'Solomon Islands',
2419 'ZA': 'South Africa',
2420 'GS': 'South Georgia and the South Sandwich Islands',
2421 'SS': 'South Sudan',
2426 'SJ': 'Svalbard and Jan Mayen',
2429 'CH': 'Switzerland',
2430 'SY': 'Syrian Arab Republic',
2431 'TW': 'Taiwan, Province of China',
2433 'TZ': 'Tanzania, United Republic of',
2435 'TL': 'Timor-Leste',
2439 'TT': 'Trinidad and Tobago',
2442 'TM': 'Turkmenistan',
2443 'TC': 'Turks and Caicos Islands',
2447 'AE': 'United Arab Emirates',
2448 'GB': 'United Kingdom',
2449 'US': 'United States',
2450 'UM': 'United States Minor Outlying Islands',
2454 'VE': 'Venezuela, Bolivarian Republic of',
2456 'VG': 'Virgin Islands, British',
2457 'VI': 'Virgin Islands, U.S.',
2458 'WF': 'Wallis and Futuna',
2459 'EH': 'Western Sahara',
2466 def short2full(cls, code):
2467 """Convert an ISO 3166-2 country code to the corresponding full name"""
2468 return cls._country_map.get(code.upper())
2471 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
2472 def __init__(self, proxies=None):
2473 # Set default handlers
2474 for type in ('http', 'https'):
2475 setattr(self, '%s_open' % type,
2476 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
2477 meth(r, proxy, type))
2478 return compat_urllib_request.ProxyHandler.__init__(self, proxies)
2480 def proxy_open(self, req, proxy, type):
2481 req_proxy = req.headers.get('Ytdl-request-proxy')
2482 if req_proxy is not None:
2484 del req.headers['Ytdl-request-proxy']
2486 if proxy == '__noproxy__':
2487 return None # No Proxy
2488 return compat_urllib_request.ProxyHandler.proxy_open(
2489 self, req, proxy, type)