4 from __future__
import unicode_literals
35 import xml
.etree
.ElementTree
42 compat_etree_fromstring
,
45 compat_html_entities_html5
,
51 compat_socket_create_connection
,
57 compat_urllib_parse_urlencode
,
58 compat_urllib_parse_urlparse
,
59 compat_urllib_parse_unquote_plus
,
60 compat_urllib_request
,
71 def register_socks_protocols():
72 # "Register" SOCKS protocols
73 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
74 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
75 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
76 if scheme
not in compat_urlparse
.uses_netloc
:
77 compat_urlparse
.uses_netloc
.append(scheme
)
80 # This is not clearly defined otherwise
81 compiled_regex_type
= type(re
.compile(''))
84 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/47.0 (Chrome)',
85 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
86 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
87 'Accept-Encoding': 'gzip, deflate',
88 'Accept-Language': 'en-us,en;q=0.5',
93 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
99 ENGLISH_MONTH_NAMES
= [
100 'January', 'February', 'March', 'April', 'May', 'June',
101 'July', 'August', 'September', 'October', 'November', 'December']
104 'en': ENGLISH_MONTH_NAMES
,
106 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
107 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
111 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
112 'flv', 'f4v', 'f4a', 'f4b',
113 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
114 'mkv', 'mka', 'mk3d',
123 'f4f', 'f4m', 'm3u8', 'smil')
125 # needed for sanitizing filenames in restricted mode
126 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
127 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUYP', ['ss'],
128 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuypy')))
151 '%Y-%m-%d %H:%M:%S.%f',
154 '%Y-%m-%dT%H:%M:%SZ',
155 '%Y-%m-%dT%H:%M:%S.%fZ',
156 '%Y-%m-%dT%H:%M:%S.%f0Z',
158 '%Y-%m-%dT%H:%M:%S.%f',
161 '%b %d %Y at %H:%M:%S',
164 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
165 DATE_FORMATS_DAY_FIRST
.extend([
174 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
175 DATE_FORMATS_MONTH_FIRST
.extend([
183 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
186 def preferredencoding():
187 """Get preferred encoding.
189 Returns the best encoding scheme for the system, based on
190 locale.getpreferredencoding() and some further tweaks.
193 pref
= locale
.getpreferredencoding()
201 def write_json_file(obj
, fn
):
202 """ Encode obj as JSON and write it to fn, atomically if possible """
204 fn
= encodeFilename(fn
)
205 if sys
.version_info
< (3, 0) and sys
.platform
!= 'win32':
206 encoding
= get_filesystem_encoding()
207 # os.path.basename returns a bytes object, but NamedTemporaryFile
208 # will fail if the filename contains non ascii characters unless we
209 # use a unicode object
210 path_basename
= lambda f
: os
.path
.basename(fn
).decode(encoding
)
211 # the same for os.path.dirname
212 path_dirname
= lambda f
: os
.path
.dirname(fn
).decode(encoding
)
214 path_basename
= os
.path
.basename
215 path_dirname
= os
.path
.dirname
219 'prefix': path_basename(fn
) + '.',
220 'dir': path_dirname(fn
),
224 # In Python 2.x, json.dump expects a bytestream.
225 # In Python 3.x, it writes to a character stream
226 if sys
.version_info
< (3, 0):
234 tf
= tempfile
.NamedTemporaryFile(**compat_kwargs(args
))
239 if sys
.platform
== 'win32':
240 # Need to remove existing file on Windows, else os.rename raises
241 # WindowsError or FileExistsError.
246 os
.rename(tf
.name
, fn
)
255 if sys
.version_info
>= (2, 7):
256 def find_xpath_attr(node
, xpath
, key
, val
=None):
257 """ Find the xpath xpath[@key=val] """
258 assert re
.match(r
'^[a-zA-Z_-]+$', key
)
259 expr
= xpath
+ ('[@%s]' % key
if val
is None else "[@%s='%s']" % (key
, val
))
260 return node
.find(expr
)
262 def find_xpath_attr(node
, xpath
, key
, val
=None):
263 for f
in node
.findall(compat_xpath(xpath
)):
264 if key
not in f
.attrib
:
266 if val
is None or f
.attrib
.get(key
) == val
:
270 # On python2.6 the xml.etree.ElementTree.Element methods don't support
271 # the namespace parameter
274 def xpath_with_ns(path
, ns_map
):
275 components
= [c
.split(':') for c
in path
.split('/')]
279 replaced
.append(c
[0])
282 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
283 return '/'.join(replaced
)
286 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
287 def _find_xpath(xpath
):
288 return node
.find(compat_xpath(xpath
))
290 if isinstance(xpath
, (str, compat_str
)):
291 n
= _find_xpath(xpath
)
299 if default
is not NO_DEFAULT
:
302 name
= xpath
if name
is None else name
303 raise ExtractorError('Could not find XML element %s' % name
)
309 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
310 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
311 if n
is None or n
== default
:
314 if default
is not NO_DEFAULT
:
317 name
= xpath
if name
is None else name
318 raise ExtractorError('Could not find XML element\'s text %s' % name
)
324 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
325 n
= find_xpath_attr(node
, xpath
, key
)
327 if default
is not NO_DEFAULT
:
330 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
331 raise ExtractorError('Could not find XML attribute %s' % name
)
337 def get_element_by_id(id, html
):
338 """Return the content of the tag with the specified ID in the passed HTML document"""
339 return get_element_by_attribute('id', id, html
)
342 def get_element_by_class(class_name
, html
):
343 """Return the content of the first tag with the specified class in the passed HTML document"""
344 retval
= get_elements_by_class(class_name
, html
)
345 return retval
[0] if retval
else None
348 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
349 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
350 return retval
[0] if retval
else None
353 def get_elements_by_class(class_name
, html
):
354 """Return the content of all tags with the specified class in the passed HTML document as a list"""
355 return get_elements_by_attribute(
356 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
357 html, escape_value=False)
360 def get_elements_by_attribute(attribute, value, html, escape_value=True):
361 """Return the content of the tag with the specified attribute in the passed HTML document"""
363 value = re.escape(value) if escape_value else value
366 for m in re.finditer(r'''(?xs)
368 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^
']*'))*?
370 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^
"]*"|
='[^']*'))*?
374 ''' % (re.escape(attribute), value), html):
375 res = m.group('content
')
377 if res.startswith('"') or res.startswith("'"):
380 retlist.append(unescapeHTML(res))
385 class HTMLAttributeParser(compat_HTMLParser):
386 """Trivial HTML parser to gather the attributes for a single element"""
389 compat_HTMLParser.__init__(self)
391 def handle_starttag(self, tag, attrs):
392 self.attrs = dict(attrs)
395 def extract_attributes(html_element):
396 """Given a string for an HTML element such as
398 a="foo" B="bar" c="&98;az" d=boz
399 empty= noval entity="&"
402 Decode and return a dictionary of attributes.
404 'a
': 'foo
', 'b
': 'bar
', c: 'baz
', d: 'boz
',
405 'empty
': '', 'noval
': None, 'entity
': '&',
406 'sq
': '"', 'dq': '\''
408 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
409 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
411 parser = HTMLAttributeParser()
412 parser.feed(html_element)
417 def clean_html(html):
418 """Clean an HTML snippet into a readable string"""
420 if html is None: # Convenience for sanitizing descriptions etc.
424 html = html.replace('\n', ' ')
425 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
426 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
428 html = re.sub('<.*?>', '', html)
429 # Replace html entities
430 html = unescapeHTML(html)
434 def sanitize_open(filename, open_mode):
435 """Try to open the given filename, and slightly tweak it if this fails.
437 Attempts to open the given filename. If this fails, it tries to change
438 the filename slightly, step by step, until it's either able to open it
439 or it fails and raises a final exception, like the standard open()
442 It returns the tuple (stream, definitive_file_name).
446 if sys.platform == 'win32':
448 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
449 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
450 stream = open(encodeFilename(filename), open_mode)
451 return (stream, filename)
452 except (IOError, OSError) as err:
453 if err.errno in (errno.EACCES,):
456 # In case of error, try to remove win32 forbidden chars
457 alt_filename = sanitize_path(filename)
458 if alt_filename == filename:
461 # An exception here should be caught in the caller
462 stream = open(encodeFilename(alt_filename), open_mode)
463 return (stream, alt_filename)
466 def timeconvert(timestr):
467 """Convert RFC 2822 defined time string into system timestamp"""
469 timetuple = email.utils.parsedate_tz(timestr)
470 if timetuple is not None:
471 timestamp = email.utils.mktime_tz(timetuple)
475 def sanitize_filename(s, restricted=False, is_id=False):
476 """Sanitizes a string so it could be used as part of a filename.
477 If restricted is set, use a stricter subset of allowed characters.
478 Set is_id if this is not an arbitrary string, but an ID that should be kept
481 def replace_insane(char):
482 if restricted and char in ACCENT_CHARS:
483 return ACCENT_CHARS[char]
484 if char == '?' or ord(char) < 32 or ord(char) == 127:
487 return '' if restricted else '\''
489 return '_
-' if restricted else ' -'
490 elif char in '\\/|
*<>':
492 if restricted and (char in '!&\'()[]{}$
;`^
,#' or char.isspace()):
494 if restricted
and ord(char
) > 127:
499 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
)
500 result
= ''.join(map(replace_insane
, s
))
502 while '__' in result
:
503 result
= result
.replace('__', '_')
504 result
= result
.strip('_')
505 # Common case of "Foreign band name - English song title"
506 if restricted
and result
.startswith('-_'):
508 if result
.startswith('-'):
509 result
= '_' + result
[len('-'):]
510 result
= result
.lstrip('.')
516 def sanitize_path(s
):
517 """Sanitizes and normalizes path on Windows"""
518 if sys
.platform
!= 'win32':
520 drive_or_unc
, _
= os
.path
.splitdrive(s
)
521 if sys
.version_info
< (2, 7) and not drive_or_unc
:
522 drive_or_unc
, _
= os
.path
.splitunc(s
)
523 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
527 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
528 for path_part
in norm_path
]
530 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
531 return os
.path
.join(*sanitized_path
)
534 # Prepend protocol-less URLs with `http:` scheme in order to mitigate the number of
535 # unwanted failures due to missing protocol
536 def sanitize_url(url
):
537 return 'http:%s' % url
if url
.startswith('//') else url
540 def sanitized_Request(url
, *args
, **kwargs
):
541 return compat_urllib_request
.Request(sanitize_url(url
), *args
, **kwargs
)
545 """Expand shell variables and ~"""
546 return os
.path
.expandvars(compat_expanduser(s
))
549 def orderedSet(iterable
):
550 """ Remove all duplicates from the input iterable """
558 def _htmlentity_transform(entity_with_semicolon
):
559 """Transforms an HTML entity to a character."""
560 entity
= entity_with_semicolon
[:-1]
562 # Known non-numeric HTML entity
563 if entity
in compat_html_entities
.name2codepoint
:
564 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
566 # TODO: HTML5 allows entities without a semicolon. For example,
567 # 'Éric' should be decoded as 'Éric'.
568 if entity_with_semicolon
in compat_html_entities_html5
:
569 return compat_html_entities_html5
[entity_with_semicolon
]
571 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
573 numstr
= mobj
.group(1)
574 if numstr
.startswith('x'):
576 numstr
= '0%s' % numstr
579 # See https://github.com/rg3/youtube-dl/issues/7518
581 return compat_chr(int(numstr
, base
))
585 # Unknown entity in name, return its literal representation
586 return '&%s;' % entity
592 assert type(s
) == compat_str
595 r
'&([^;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
598 def get_subprocess_encoding():
599 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
600 # For subprocess calls, encode with locale encoding
601 # Refer to http://stackoverflow.com/a/9951851/35070
602 encoding
= preferredencoding()
604 encoding
= sys
.getfilesystemencoding()
610 def encodeFilename(s
, for_subprocess
=False):
612 @param s The name of the file
615 assert type(s
) == compat_str
617 # Python 3 has a Unicode API
618 if sys
.version_info
>= (3, 0):
621 # Pass '' directly to use Unicode APIs on Windows 2000 and up
622 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
623 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
624 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
627 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
628 if sys
.platform
.startswith('java'):
631 return s
.encode(get_subprocess_encoding(), 'ignore')
634 def decodeFilename(b
, for_subprocess
=False):
636 if sys
.version_info
>= (3, 0):
639 if not isinstance(b
, bytes):
642 return b
.decode(get_subprocess_encoding(), 'ignore')
645 def encodeArgument(s
):
646 if not isinstance(s
, compat_str
):
647 # Legacy code that uses byte strings
648 # Uncomment the following line after fixing all post processors
649 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
650 s
= s
.decode('ascii')
651 return encodeFilename(s
, True)
654 def decodeArgument(b
):
655 return decodeFilename(b
, True)
658 def decodeOption(optval
):
661 if isinstance(optval
, bytes):
662 optval
= optval
.decode(preferredencoding())
664 assert isinstance(optval
, compat_str
)
668 def formatSeconds(secs
):
670 return '%d:%02d:%02d' % (secs
// 3600, (secs
% 3600) // 60, secs
% 60)
672 return '%d:%02d' % (secs
// 60, secs
% 60)
677 def make_HTTPS_handler(params
, **kwargs
):
678 opts_no_check_certificate
= params
.get('nocheckcertificate', False)
679 if hasattr(ssl
, 'create_default_context'): # Python >= 3.4 or 2.7.9
680 context
= ssl
.create_default_context(ssl
.Purpose
.SERVER_AUTH
)
681 if opts_no_check_certificate
:
682 context
.check_hostname
= False
683 context
.verify_mode
= ssl
.CERT_NONE
685 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
688 # (create_default_context present but HTTPSHandler has no context=)
691 if sys
.version_info
< (3, 2):
692 return YoutubeDLHTTPSHandler(params
, **kwargs
)
694 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLSv1
)
695 context
.verify_mode
= (ssl
.CERT_NONE
696 if opts_no_check_certificate
697 else ssl
.CERT_REQUIRED
)
698 context
.set_default_verify_paths()
699 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
702 def bug_reports_message():
703 if ytdl_is_updateable():
704 update_cmd
= 'type youtube-dl -U to update'
706 update_cmd
= 'see https://yt-dl.org/update on how to update'
707 msg
= '; please report this issue on https://yt-dl.org/bug .'
708 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
709 msg
+= ' Be sure to call youtube-dl with the --verbose flag and include its complete output.'
713 class YoutubeDLError(Exception):
714 """Base exception for YoutubeDL errors."""
718 class ExtractorError(YoutubeDLError
):
719 """Error during info extraction."""
721 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None):
722 """ tb, if given, is the original traceback (so that it can be printed out).
723 If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
726 if sys
.exc_info()[0] in (compat_urllib_error
.URLError
, socket
.timeout
, UnavailableVideoError
):
728 if video_id
is not None:
729 msg
= video_id
+ ': ' + msg
731 msg
+= ' (caused by %r)' % cause
733 msg
+= bug_reports_message()
734 super(ExtractorError
, self
).__init
__(msg
)
737 self
.exc_info
= sys
.exc_info() # preserve original exception
739 self
.video_id
= video_id
741 def format_traceback(self
):
742 if self
.traceback
is None:
744 return ''.join(traceback
.format_tb(self
.traceback
))
747 class UnsupportedError(ExtractorError
):
748 def __init__(self
, url
):
749 super(UnsupportedError
, self
).__init
__(
750 'Unsupported URL: %s' % url
, expected
=True)
754 class RegexNotFoundError(ExtractorError
):
755 """Error when a regex didn't match"""
759 class GeoRestrictedError(ExtractorError
):
760 """Geographic restriction Error exception.
762 This exception may be thrown when a video is not available from your
763 geographic location due to geographic restrictions imposed by a website.
765 def __init__(self
, msg
, countries
=None):
766 super(GeoRestrictedError
, self
).__init
__(msg
, expected
=True)
768 self
.countries
= countries
771 class DownloadError(YoutubeDLError
):
772 """Download Error exception.
774 This exception may be thrown by FileDownloader objects if they are not
775 configured to continue on errors. They will contain the appropriate
779 def __init__(self
, msg
, exc_info
=None):
780 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
781 super(DownloadError
, self
).__init
__(msg
)
782 self
.exc_info
= exc_info
785 class SameFileError(YoutubeDLError
):
786 """Same File exception.
788 This exception will be thrown by FileDownloader objects if they detect
789 multiple files would have to be downloaded to the same file on disk.
794 class PostProcessingError(YoutubeDLError
):
795 """Post Processing exception.
797 This exception may be raised by PostProcessor's .run() method to
798 indicate an error in the postprocessing task.
801 def __init__(self
, msg
):
802 super(PostProcessingError
, self
).__init
__(msg
)
806 class MaxDownloadsReached(YoutubeDLError
):
807 """ --max-downloads limit has been reached. """
811 class UnavailableVideoError(YoutubeDLError
):
812 """Unavailable Format exception.
814 This exception will be thrown when a video is requested
815 in a format that is not available for that video.
820 class ContentTooShortError(YoutubeDLError
):
821 """Content Too Short exception.
823 This exception may be raised by FileDownloader objects when a file they
824 download is too small for what the server announced first, indicating
825 the connection was probably interrupted.
828 def __init__(self
, downloaded
, expected
):
829 super(ContentTooShortError
, self
).__init
__(
830 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
833 self
.downloaded
= downloaded
834 self
.expected
= expected
837 class XAttrMetadataError(YoutubeDLError
):
838 def __init__(self
, code
=None, msg
='Unknown error'):
839 super(XAttrMetadataError
, self
).__init
__(msg
)
843 # Parsing code and msg
844 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
) or
845 'No space left' in self
.msg
or 'Disk quota excedded' in self
.msg
):
846 self
.reason
= 'NO_SPACE'
847 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
848 self
.reason
= 'VALUE_TOO_LONG'
850 self
.reason
= 'NOT_SUPPORTED'
853 class XAttrUnavailableError(YoutubeDLError
):
857 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
858 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
859 # expected HTTP responses to meet HTTP/1.0 or later (see also
860 # https://github.com/rg3/youtube-dl/issues/6727)
861 if sys
.version_info
< (3, 0):
862 kwargs
[b
'strict'] = True
863 hc
= http_class(*args
, **kwargs
)
864 source_address
= ydl_handler
._params
.get('source_address')
865 if source_address
is not None:
866 sa
= (source_address
, 0)
867 if hasattr(hc
, 'source_address'): # Python 2.7+
868 hc
.source_address
= sa
870 def _hc_connect(self
, *args
, **kwargs
):
871 sock
= compat_socket_create_connection(
872 (self
.host
, self
.port
), self
.timeout
, sa
)
874 self
.sock
= ssl
.wrap_socket(
875 sock
, self
.key_file
, self
.cert_file
,
876 ssl_version
=ssl
.PROTOCOL_TLSv1
)
879 hc
.connect
= functools
.partial(_hc_connect
, hc
)
884 def handle_youtubedl_headers(headers
):
885 filtered_headers
= headers
887 if 'Youtubedl-no-compression' in filtered_headers
:
888 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
889 del filtered_headers
['Youtubedl-no-compression']
891 return filtered_headers
894 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
895 """Handler for HTTP requests and responses.
897 This class, when installed with an OpenerDirector, automatically adds
898 the standard headers to every HTTP request and handles gzipped and
899 deflated responses from web servers. If compression is to be avoided in
900 a particular request, the original request in the program code only has
901 to include the HTTP header "Youtubedl-no-compression", which will be
902 removed before making the real request.
904 Part of this code was copied from:
906 http://techknack.net/python-urllib2-handlers/
908 Andrew Rowls, the author of that code, agreed to release it to the
912 def __init__(self
, params
, *args
, **kwargs
):
913 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
914 self
._params
= params
916 def http_open(self
, req
):
917 conn_class
= compat_http_client
.HTTPConnection
919 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
921 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
922 del req
.headers
['Ytdl-socks-proxy']
924 return self
.do_open(functools
.partial(
925 _create_http_connection
, self
, conn_class
, False),
931 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
933 return zlib
.decompress(data
)
936 def addinfourl_wrapper(stream
, headers
, url
, code
):
937 if hasattr(compat_urllib_request
.addinfourl
, 'getcode'):
938 return compat_urllib_request
.addinfourl(stream
, headers
, url
, code
)
939 ret
= compat_urllib_request
.addinfourl(stream
, headers
, url
)
943 def http_request(self
, req
):
944 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
945 # always respected by websites, some tend to give out URLs with non percent-encoded
946 # non-ASCII characters (see telemb.py, ard.py [#3412])
947 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
948 # To work around aforementioned issue we will replace request's original URL with
949 # percent-encoded one
950 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
951 # the code of this workaround has been moved here from YoutubeDL.urlopen()
952 url
= req
.get_full_url()
953 url_escaped
= escape_url(url
)
955 # Substitute URL if any change after escaping
956 if url
!= url_escaped
:
957 req
= update_Request(req
, url
=url_escaped
)
959 for h
, v
in std_headers
.items():
960 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
961 # The dict keys are capitalized because of this bug by urllib
962 if h
.capitalize() not in req
.headers
:
965 req
.headers
= handle_youtubedl_headers(req
.headers
)
967 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
968 # Python 2.6 is brain-dead when it comes to fragments
969 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
970 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
974 def http_response(self
, req
, resp
):
977 if resp
.headers
.get('Content-encoding', '') == 'gzip':
978 content
= resp
.read()
979 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
981 uncompressed
= io
.BytesIO(gz
.read())
982 except IOError as original_ioerror
:
983 # There may be junk add the end of the file
984 # See http://stackoverflow.com/q/4928560/35070 for details
985 for i
in range(1, 1024):
987 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
988 uncompressed
= io
.BytesIO(gz
.read())
993 raise original_ioerror
994 resp
= self
.addinfourl_wrapper(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
995 resp
.msg
= old_resp
.msg
996 del resp
.headers
['Content-encoding']
998 if resp
.headers
.get('Content-encoding', '') == 'deflate':
999 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1000 resp
= self
.addinfourl_wrapper(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1001 resp
.msg
= old_resp
.msg
1002 del resp
.headers
['Content-encoding']
1003 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1004 # https://github.com/rg3/youtube-dl/issues/6457).
1005 if 300 <= resp
.code
< 400:
1006 location
= resp
.headers
.get('Location')
1008 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1009 if sys
.version_info
>= (3, 0):
1010 location
= location
.encode('iso-8859-1').decode('utf-8')
1012 location
= location
.decode('utf-8')
1013 location_escaped
= escape_url(location
)
1014 if location
!= location_escaped
:
1015 del resp
.headers
['Location']
1016 if sys
.version_info
< (3, 0):
1017 location_escaped
= location_escaped
.encode('utf-8')
1018 resp
.headers
['Location'] = location_escaped
1021 https_request
= http_request
1022 https_response
= http_response
1025 def make_socks_conn_class(base_class
, socks_proxy
):
1026 assert issubclass(base_class
, (
1027 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1029 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1030 if url_components
.scheme
.lower() == 'socks5':
1031 socks_type
= ProxyType
.SOCKS5
1032 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1033 socks_type
= ProxyType
.SOCKS4
1034 elif url_components
.scheme
.lower() == 'socks4a':
1035 socks_type
= ProxyType
.SOCKS4A
1037 def unquote_if_non_empty(s
):
1040 return compat_urllib_parse_unquote_plus(s
)
1044 url_components
.hostname
, url_components
.port
or 1080,
1046 unquote_if_non_empty(url_components
.username
),
1047 unquote_if_non_empty(url_components
.password
),
1050 class SocksConnection(base_class
):
1052 self
.sock
= sockssocket()
1053 self
.sock
.setproxy(*proxy_args
)
1054 if type(self
.timeout
) in (int, float):
1055 self
.sock
.settimeout(self
.timeout
)
1056 self
.sock
.connect((self
.host
, self
.port
))
1058 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1059 if hasattr(self
, '_context'): # Python > 2.6
1060 self
.sock
= self
._context
.wrap_socket(
1061 self
.sock
, server_hostname
=self
.host
)
1063 self
.sock
= ssl
.wrap_socket(self
.sock
)
1065 return SocksConnection
1068 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1069 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1070 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1071 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1072 self
._params
= params
1074 def https_open(self
, req
):
1076 conn_class
= self
._https
_conn
_class
1078 if hasattr(self
, '_context'): # python > 2.6
1079 kwargs
['context'] = self
._context
1080 if hasattr(self
, '_check_hostname'): # python 3.x
1081 kwargs
['check_hostname'] = self
._check
_hostname
1083 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1085 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1086 del req
.headers
['Ytdl-socks-proxy']
1088 return self
.do_open(functools
.partial(
1089 _create_http_connection
, self
, conn_class
, True),
1093 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1094 def __init__(self
, cookiejar
=None):
1095 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1097 def http_response(self
, request
, response
):
1098 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1099 # characters in Set-Cookie HTTP header of last response (see
1100 # https://github.com/rg3/youtube-dl/issues/6769).
1101 # In order to at least prevent crashing we will percent encode Set-Cookie
1102 # header before HTTPCookieProcessor starts processing it.
1103 # if sys.version_info < (3, 0) and response.headers:
1104 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1105 # set_cookie = response.headers.get(set_cookie_header)
1107 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1108 # if set_cookie != set_cookie_escaped:
1109 # del response.headers[set_cookie_header]
1110 # response.headers[set_cookie_header] = set_cookie_escaped
1111 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1113 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1114 https_response
= http_response
1117 def extract_timezone(date_str
):
1119 r
'^.{8,}?(?P<tz>Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
1122 timezone
= datetime
.timedelta()
1124 date_str
= date_str
[:-len(m
.group('tz'))]
1125 if not m
.group('sign'):
1126 timezone
= datetime
.timedelta()
1128 sign
= 1 if m
.group('sign') == '+' else -1
1129 timezone
= datetime
.timedelta(
1130 hours
=sign
* int(m
.group('hours')),
1131 minutes
=sign
* int(m
.group('minutes')))
1132 return timezone
, date_str
1135 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1136 """ Return a UNIX timestamp from the given date """
1138 if date_str
is None:
1141 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1143 if timezone
is None:
1144 timezone
, date_str
= extract_timezone(date_str
)
1147 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
1148 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1149 return calendar
.timegm(dt
.timetuple())
1154 def date_formats(day_first
=True):
1155 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1158 def unified_strdate(date_str
, day_first
=True):
1159 """Return a string with the date in the format YYYYMMDD"""
1161 if date_str
is None:
1165 date_str
= date_str
.replace(',', ' ')
1166 # Remove AM/PM + timezone
1167 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1168 _
, date_str
= extract_timezone(date_str
)
1170 for expression
in date_formats(day_first
):
1172 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1175 if upload_date
is None:
1176 timetuple
= email
.utils
.parsedate_tz(date_str
)
1179 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1182 if upload_date
is not None:
1183 return compat_str(upload_date
)
1186 def unified_timestamp(date_str
, day_first
=True):
1187 if date_str
is None:
1190 date_str
= date_str
.replace(',', ' ')
1192 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1193 timezone
, date_str
= extract_timezone(date_str
)
1195 # Remove AM/PM + timezone
1196 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1198 # Remove unrecognized timezones from ISO 8601 alike timestamps
1199 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1201 date_str
= date_str
[:-len(m
.group('tz'))]
1203 for expression
in date_formats(day_first
):
1205 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1206 return calendar
.timegm(dt
.timetuple())
1209 timetuple
= email
.utils
.parsedate_tz(date_str
)
1211 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1214 def determine_ext(url
, default_ext
='unknown_video'):
1217 guess
= url
.partition('?')[0].rpartition('.')[2]
1218 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1220 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1221 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1222 return guess
.rstrip('/')
1227 def subtitles_filename(filename
, sub_lang
, sub_format
):
1228 return filename
.rsplit('.', 1)[0] + '.' + sub_lang
+ '.' + sub_format
1231 def date_from_str(date_str
):
1233 Return a datetime object from a string in the format YYYYMMDD or
1234 (now|today)[+-][0-9](day|week|month|year)(s)?"""
1235 today
= datetime
.date
.today()
1236 if date_str
in ('now', 'today'):
1238 if date_str
== 'yesterday':
1239 return today
- datetime
.timedelta(days
=1)
1240 match
= re
.match(r
'(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str
)
1241 if match
is not None:
1242 sign
= match
.group('sign')
1243 time
= int(match
.group('time'))
1246 unit
= match
.group('unit')
1247 # A bad approximation?
1251 elif unit
== 'year':
1255 delta
= datetime
.timedelta(**{unit: time}
)
1256 return today
+ delta
1257 return datetime
.datetime
.strptime(date_str
, '%Y%m%d').date()
1260 def hyphenate_date(date_str
):
1262 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1263 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1264 if match
is not None:
1265 return '-'.join(match
.groups())
1270 class DateRange(object):
1271 """Represents a time interval between two dates"""
1273 def __init__(self
, start
=None, end
=None):
1274 """start and end must be strings in the format accepted by date"""
1275 if start
is not None:
1276 self
.start
= date_from_str(start
)
1278 self
.start
= datetime
.datetime
.min.date()
1280 self
.end
= date_from_str(end
)
1282 self
.end
= datetime
.datetime
.max.date()
1283 if self
.start
> self
.end
:
1284 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1288 """Returns a range that only contains the given day"""
1289 return cls(day
, day
)
1291 def __contains__(self
, date
):
1292 """Check if the date is in the range"""
1293 if not isinstance(date
, datetime
.date
):
1294 date
= date_from_str(date
)
1295 return self
.start
<= date
<= self
.end
1298 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
1301 def platform_name():
1302 """ Returns the platform name as a compat_str """
1303 res
= platform
.platform()
1304 if isinstance(res
, bytes):
1305 res
= res
.decode(preferredencoding())
1307 assert isinstance(res
, compat_str
)
1311 def _windows_write_string(s
, out
):
1312 """ Returns True if the string was written using special methods,
1313 False if it has yet to be written out."""
1314 # Adapted from http://stackoverflow.com/a/3259271/35070
1317 import ctypes
.wintypes
1325 fileno
= out
.fileno()
1326 except AttributeError:
1327 # If the output stream doesn't have a fileno, it's virtual
1329 except io
.UnsupportedOperation
:
1330 # Some strange Windows pseudo files?
1332 if fileno
not in WIN_OUTPUT_IDS
:
1335 GetStdHandle
= ctypes
.WINFUNCTYPE(
1336 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
1337 (b
'GetStdHandle', ctypes
.windll
.kernel32
))
1338 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
1340 WriteConsoleW
= ctypes
.WINFUNCTYPE(
1341 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
1342 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
1343 ctypes
.wintypes
.LPVOID
)((b
'WriteConsoleW', ctypes
.windll
.kernel32
))
1344 written
= ctypes
.wintypes
.DWORD(0)
1346 GetFileType
= ctypes
.WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)((b
'GetFileType', ctypes
.windll
.kernel32
))
1347 FILE_TYPE_CHAR
= 0x0002
1348 FILE_TYPE_REMOTE
= 0x8000
1349 GetConsoleMode
= ctypes
.WINFUNCTYPE(
1350 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
1351 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
1352 (b
'GetConsoleMode', ctypes
.windll
.kernel32
))
1353 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
1355 def not_a_console(handle
):
1356 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
1358 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
or
1359 GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
1361 if not_a_console(h
):
1364 def next_nonbmp_pos(s
):
1366 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
1367 except StopIteration:
1371 count
= min(next_nonbmp_pos(s
), 1024)
1373 ret
= WriteConsoleW(
1374 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
1376 raise OSError('Failed to write string')
1377 if not count
: # We just wrote a non-BMP character
1378 assert written
.value
== 2
1381 assert written
.value
> 0
1382 s
= s
[written
.value
:]
1386 def write_string(s
, out
=None, encoding
=None):
1389 assert type(s
) == compat_str
1391 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
1392 if _windows_write_string(s
, out
):
1395 if ('b' in getattr(out
, 'mode', '') or
1396 sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
1397 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
1399 elif hasattr(out
, 'buffer'):
1400 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1401 byt
= s
.encode(enc
, 'ignore')
1402 out
.buffer.write(byt
)
1408 def bytes_to_intlist(bs
):
1411 if isinstance(bs
[0], int): # Python 3
1414 return [ord(c
) for c
in bs
]
1417 def intlist_to_bytes(xs
):
1420 return compat_struct_pack('%dB' % len(xs
), *xs
)
1423 # Cross-platform file locking
1424 if sys
.platform
== 'win32':
1425 import ctypes
.wintypes
1428 class OVERLAPPED(ctypes
.Structure
):
1430 ('Internal', ctypes
.wintypes
.LPVOID
),
1431 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1432 ('Offset', ctypes
.wintypes
.DWORD
),
1433 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1434 ('hEvent', ctypes
.wintypes
.HANDLE
),
1437 kernel32
= ctypes
.windll
.kernel32
1438 LockFileEx
= kernel32
.LockFileEx
1439 LockFileEx
.argtypes
= [
1440 ctypes
.wintypes
.HANDLE
, # hFile
1441 ctypes
.wintypes
.DWORD
, # dwFlags
1442 ctypes
.wintypes
.DWORD
, # dwReserved
1443 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1444 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1445 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1447 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1448 UnlockFileEx
= kernel32
.UnlockFileEx
1449 UnlockFileEx
.argtypes
= [
1450 ctypes
.wintypes
.HANDLE
, # hFile
1451 ctypes
.wintypes
.DWORD
, # dwReserved
1452 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1453 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1454 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1456 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1457 whole_low
= 0xffffffff
1458 whole_high
= 0x7fffffff
1460 def _lock_file(f
, exclusive
):
1461 overlapped
= OVERLAPPED()
1462 overlapped
.Offset
= 0
1463 overlapped
.OffsetHigh
= 0
1464 overlapped
.hEvent
= 0
1465 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1466 handle
= msvcrt
.get_osfhandle(f
.fileno())
1467 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
1468 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1469 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
1471 def _unlock_file(f
):
1472 assert f
._lock
_file
_overlapped
_p
1473 handle
= msvcrt
.get_osfhandle(f
.fileno())
1474 if not UnlockFileEx(handle
, 0,
1475 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1476 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
1479 # Some platforms, such as Jython, is missing fcntl
1483 def _lock_file(f
, exclusive
):
1484 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
1486 def _unlock_file(f
):
1487 fcntl
.flock(f
, fcntl
.LOCK_UN
)
1489 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
1491 def _lock_file(f
, exclusive
):
1492 raise IOError(UNSUPPORTED_MSG
)
1494 def _unlock_file(f
):
1495 raise IOError(UNSUPPORTED_MSG
)
1498 class locked_file(object):
1499 def __init__(self
, filename
, mode
, encoding
=None):
1500 assert mode
in ['r', 'a', 'w']
1501 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
1504 def __enter__(self
):
1505 exclusive
= self
.mode
!= 'r'
1507 _lock_file(self
.f
, exclusive
)
1513 def __exit__(self
, etype
, value
, traceback
):
1515 _unlock_file(self
.f
)
1522 def write(self
, *args
):
1523 return self
.f
.write(*args
)
1525 def read(self
, *args
):
1526 return self
.f
.read(*args
)
1529 def get_filesystem_encoding():
1530 encoding
= sys
.getfilesystemencoding()
1531 return encoding
if encoding
is not None else 'utf-8'
1534 def shell_quote(args
):
1536 encoding
= get_filesystem_encoding()
1538 if isinstance(a
, bytes):
1539 # We may get a filename encoded with 'encodeFilename'
1540 a
= a
.decode(encoding
)
1541 quoted_args
.append(pipes
.quote(a
))
1542 return ' '.join(quoted_args
)
1545 def smuggle_url(url
, data
):
1546 """ Pass additional data in a URL for internal use. """
1548 url
, idata
= unsmuggle_url(url
, {})
1550 sdata
= compat_urllib_parse_urlencode(
1551 {'__youtubedl_smuggle': json.dumps(data)}
)
1552 return url
+ '#' + sdata
1555 def unsmuggle_url(smug_url
, default
=None):
1556 if '#__youtubedl_smuggle' not in smug_url
:
1557 return smug_url
, default
1558 url
, _
, sdata
= smug_url
.rpartition('#')
1559 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
1560 data
= json
.loads(jsond
)
1564 def format_bytes(bytes):
1567 if type(bytes) is str:
1568 bytes = float(bytes)
1572 exponent
= int(math
.log(bytes, 1024.0))
1573 suffix
= ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent
]
1574 converted
= float(bytes) / float(1024 ** exponent
)
1575 return '%.2f%s' % (converted
, suffix
)
1578 def lookup_unit_table(unit_table
, s
):
1579 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
1581 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
1584 num_str
= m
.group('num').replace(',', '.')
1585 mult
= unit_table
[m
.group('unit')]
1586 return int(float(num_str
) * mult
)
1589 def parse_filesize(s
):
1593 # The lower-case forms are of course incorrect and unofficial,
1594 # but we support those too
1611 'megabytes': 1000 ** 2,
1612 'mebibytes': 1024 ** 2,
1618 'gigabytes': 1000 ** 3,
1619 'gibibytes': 1024 ** 3,
1625 'terabytes': 1000 ** 4,
1626 'tebibytes': 1024 ** 4,
1632 'petabytes': 1000 ** 5,
1633 'pebibytes': 1024 ** 5,
1639 'exabytes': 1000 ** 6,
1640 'exbibytes': 1024 ** 6,
1646 'zettabytes': 1000 ** 7,
1647 'zebibytes': 1024 ** 7,
1653 'yottabytes': 1000 ** 8,
1654 'yobibytes': 1024 ** 8,
1657 return lookup_unit_table(_UNIT_TABLE
, s
)
1666 if re
.match(r
'^[\d,.]+$', s
):
1667 return str_to_int(s
)
1678 return lookup_unit_table(_UNIT_TABLE
, s
)
1681 def month_by_name(name
, lang
='en'):
1682 """ Return the number of a month by (locale-independently) English name """
1684 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
1687 return month_names
.index(name
) + 1
1692 def month_by_abbreviation(abbrev
):
1693 """ Return the number of a month by (locale-independently) English
1697 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
1702 def fix_xml_ampersands(xml_str
):
1703 """Replace all the '&' by '&' in XML"""
1705 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1710 def setproctitle(title
):
1711 assert isinstance(title
, compat_str
)
1713 # ctypes in Jython is not complete
1714 # http://bugs.jython.org/issue2148
1715 if sys
.platform
.startswith('java'):
1719 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
1723 # LoadLibrary in Windows Python 2.7.13 only expects
1724 # a bytestring, but since unicode_literals turns
1725 # every string into a unicode string, it fails.
1727 title_bytes
= title
.encode('utf-8')
1728 buf
= ctypes
.create_string_buffer(len(title_bytes
))
1729 buf
.value
= title_bytes
1731 libc
.prctl(15, buf
, 0, 0, 0)
1732 except AttributeError:
1733 return # Strange libc, just skip this
1736 def remove_start(s
, start
):
1737 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
1740 def remove_end(s
, end
):
1741 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
1744 def remove_quotes(s
):
1745 if s
is None or len(s
) < 2:
1747 for quote
in ('"', "'", ):
1748 if s
[0] == quote
and s
[-1] == quote
:
1753 def url_basename(url
):
1754 path
= compat_urlparse
.urlparse(url
).path
1755 return path
.strip('/').split('/')[-1]
1759 return re
.match(r
'https?://[^?#&]+/', url
).group()
1762 def urljoin(base
, path
):
1763 if isinstance(path
, bytes):
1764 path
= path
.decode('utf-8')
1765 if not isinstance(path
, compat_str
) or not path
:
1767 if re
.match(r
'^(?:https?:)?//', path
):
1769 if isinstance(base
, bytes):
1770 base
= base
.decode('utf-8')
1771 if not isinstance(base
, compat_str
) or not re
.match(
1772 r
'^(?:https?:)?//', base
):
1774 return compat_urlparse
.urljoin(base
, path
)
1777 class HEADRequest(compat_urllib_request
.Request
):
1778 def get_method(self
):
1782 class PUTRequest(compat_urllib_request
.Request
):
1783 def get_method(self
):
1787 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
1790 v
= getattr(v
, get_attr
, None)
1796 return int(v
) * invscale
// scale
1801 def str_or_none(v
, default
=None):
1802 return default
if v
is None else compat_str(v
)
1805 def str_to_int(int_str
):
1806 """ A more relaxed version of int_or_none """
1809 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
1813 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
1817 return float(v
) * invscale
/ scale
1822 def strip_or_none(v
):
1823 return None if v
is None else v
.strip()
1826 def parse_duration(s
):
1827 if not isinstance(s
, compat_basestring
):
1832 days
, hours
, mins
, secs
, ms
= [None] * 5
1833 m
= re
.match(r
'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?Z?$', s
)
1835 days
, hours
, mins
, secs
, ms
= m
.groups()
1840 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
1843 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
1846 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
1849 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
1852 days
, hours
, mins
, secs
, ms
= m
.groups()
1854 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
1856 hours
, mins
= m
.groups()
1862 duration
+= float(secs
)
1864 duration
+= float(mins
) * 60
1866 duration
+= float(hours
) * 60 * 60
1868 duration
+= float(days
) * 24 * 60 * 60
1870 duration
+= float(ms
)
1874 def prepend_extension(filename
, ext
, expected_real_ext
=None):
1875 name
, real_ext
= os
.path
.splitext(filename
)
1877 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
1878 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
1879 else '{0}.{1}'.format(filename
, ext
))
1882 def replace_extension(filename
, ext
, expected_real_ext
=None):
1883 name
, real_ext
= os
.path
.splitext(filename
)
1884 return '{0}.{1}'.format(
1885 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
1889 def check_executable(exe
, args
=[]):
1890 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
1891 args can be a list of arguments for a short output (like -version) """
1893 subprocess
.Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate()
1899 def get_exe_version(exe
, args
=['--version'],
1900 version_re
=None, unrecognized
='present'):
1901 """ Returns the version of the specified executable,
1902 or False if the executable is not present """
1904 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
1905 # SIGTTOU if youtube-dl is run in the background.
1906 # See https://github.com/rg3/youtube-dl/issues/955#issuecomment-209789656
1907 out
, _
= subprocess
.Popen(
1908 [encodeArgument(exe
)] + args
,
1909 stdin
=subprocess
.PIPE
,
1910 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate()
1913 if isinstance(out
, bytes): # Python 2.x
1914 out
= out
.decode('ascii', 'ignore')
1915 return detect_exe_version(out
, version_re
, unrecognized
)
1918 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
1919 assert isinstance(output
, compat_str
)
1920 if version_re
is None:
1921 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
1922 m
= re
.search(version_re
, output
)
1929 class PagedList(object):
1931 # This is only useful for tests
1932 return len(self
.getslice())
1935 class OnDemandPagedList(PagedList
):
1936 def __init__(self
, pagefunc
, pagesize
, use_cache
=False):
1937 self
._pagefunc
= pagefunc
1938 self
._pagesize
= pagesize
1939 self
._use
_cache
= use_cache
1943 def getslice(self
, start
=0, end
=None):
1945 for pagenum
in itertools
.count(start
// self
._pagesize
):
1946 firstid
= pagenum
* self
._pagesize
1947 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
1948 if start
>= nextfirstid
:
1953 page_results
= self
._cache
.get(pagenum
)
1954 if page_results
is None:
1955 page_results
= list(self
._pagefunc
(pagenum
))
1957 self
._cache
[pagenum
] = page_results
1960 start
% self
._pagesize
1961 if firstid
<= start
< nextfirstid
1965 ((end
- 1) % self
._pagesize
) + 1
1966 if (end
is not None and firstid
<= end
<= nextfirstid
)
1969 if startv
!= 0 or endv
is not None:
1970 page_results
= page_results
[startv
:endv
]
1971 res
.extend(page_results
)
1973 # A little optimization - if current page is not "full", ie. does
1974 # not contain page_size videos then we can assume that this page
1975 # is the last one - there are no more ids on further pages -
1976 # i.e. no need to query again.
1977 if len(page_results
) + startv
< self
._pagesize
:
1980 # If we got the whole page, but the next page is not interesting,
1981 # break out early as well
1982 if end
== nextfirstid
:
1987 class InAdvancePagedList(PagedList
):
1988 def __init__(self
, pagefunc
, pagecount
, pagesize
):
1989 self
._pagefunc
= pagefunc
1990 self
._pagecount
= pagecount
1991 self
._pagesize
= pagesize
1993 def getslice(self
, start
=0, end
=None):
1995 start_page
= start
// self
._pagesize
1997 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
1998 skip_elems
= start
- start_page
* self
._pagesize
1999 only_more
= None if end
is None else end
- start
2000 for pagenum
in range(start_page
, end_page
):
2001 page
= list(self
._pagefunc
(pagenum
))
2003 page
= page
[skip_elems
:]
2005 if only_more
is not None:
2006 if len(page
) < only_more
:
2007 only_more
-= len(page
)
2009 page
= page
[:only_more
]
2016 def uppercase_escape(s
):
2017 unicode_escape
= codecs
.getdecoder('unicode_escape')
2019 r
'\\U[0-9a-fA-F]{8}',
2020 lambda m
: unicode_escape(m
.group(0))[0],
2024 def lowercase_escape(s
):
2025 unicode_escape
= codecs
.getdecoder('unicode_escape')
2027 r
'\\u[0-9a-fA-F]{4}',
2028 lambda m
: unicode_escape(m
.group(0))[0],
2032 def escape_rfc3986(s
):
2033 """Escape non-ASCII characters as suggested by RFC 3986"""
2034 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
2035 s
= s
.encode('utf-8')
2036 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2039 def escape_url(url
):
2040 """Escape URL as suggested by RFC 3986"""
2041 url_parsed
= compat_urllib_parse_urlparse(url
)
2042 return url_parsed
._replace
(
2043 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2044 path
=escape_rfc3986(url_parsed
.path
),
2045 params
=escape_rfc3986(url_parsed
.params
),
2046 query
=escape_rfc3986(url_parsed
.query
),
2047 fragment
=escape_rfc3986(url_parsed
.fragment
)
2051 def read_batch_urls(batch_fd
):
2053 if not isinstance(url
, compat_str
):
2054 url
= url
.decode('utf-8', 'replace')
2055 BOM_UTF8
= '\xef\xbb\xbf'
2056 if url
.startswith(BOM_UTF8
):
2057 url
= url
[len(BOM_UTF8
):]
2059 if url
.startswith(('#', ';', ']')):
2063 with contextlib
.closing(batch_fd
) as fd
:
2064 return [url
for url
in map(fixup
, fd
) if url
]
2067 def urlencode_postdata(*args
, **kargs
):
2068 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2071 def update_url_query(url
, query
):
2074 parsed_url
= compat_urlparse
.urlparse(url
)
2075 qs
= compat_parse_qs(parsed_url
.query
)
2077 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2078 query
=compat_urllib_parse_urlencode(qs
, True)))
2081 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2082 req_headers
= req
.headers
.copy()
2083 req_headers
.update(headers
)
2084 req_data
= data
or req
.data
2085 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2086 req_get_method
= req
.get_method()
2087 if req_get_method
== 'HEAD':
2088 req_type
= HEADRequest
2089 elif req_get_method
== 'PUT':
2090 req_type
= PUTRequest
2092 req_type
= compat_urllib_request
.Request
2094 req_url
, data
=req_data
, headers
=req_headers
,
2095 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2096 if hasattr(req
, 'timeout'):
2097 new_req
.timeout
= req
.timeout
2101 def try_multipart_encode(data
, boundary
):
2102 content_type
= 'multipart/form-data; boundary=%s' % boundary
2105 for k
, v
in data
.items():
2106 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2107 if isinstance(k
, compat_str
):
2108 k
= k
.encode('utf-8')
2109 if isinstance(v
, compat_str
):
2110 v
= v
.encode('utf-8')
2111 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2112 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2113 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2114 if boundary
.encode('ascii') in content
:
2115 raise ValueError('Boundary overlaps with data')
2118 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2120 return out
, content_type
2123 def multipart_encode(data
, boundary
=None):
2125 Encode a dict to RFC 7578-compliant form-data
2128 A dict where keys and values can be either Unicode or bytes-like
2131 If specified a Unicode object, it's used as the boundary. Otherwise
2132 a random boundary is generated.
2134 Reference: https://tools.ietf.org/html/rfc7578
2136 has_specified_boundary
= boundary
is not None
2139 if boundary
is None:
2140 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2143 out
, content_type
= try_multipart_encode(data
, boundary
)
2146 if has_specified_boundary
:
2150 return out
, content_type
2153 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
2154 if isinstance(key_or_keys
, (list, tuple)):
2155 for key
in key_or_keys
:
2156 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
2160 return d
.get(key_or_keys
, default
)
2163 def try_get(src
, getter
, expected_type
=None):
2164 if not isinstance(getter
, (list, tuple)):
2169 except (AttributeError, KeyError, TypeError, IndexError):
2172 if expected_type
is None or isinstance(v
, expected_type
):
2176 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2177 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
2189 TV_PARENTAL_GUIDELINES
= {
2199 def parse_age_limit(s
):
2201 return s
if 0 <= s
<= 21 else None
2202 if not isinstance(s
, compat_basestring
):
2204 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2206 return int(m
.group('age'))
2208 return US_RATINGS
[s
]
2209 return TV_PARENTAL_GUIDELINES
.get(s
)
2212 def strip_jsonp(code
):
2214 r
'(?s)^[a-zA-Z0-9_.$]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r
'\1', code
)
2217 def js_to_json(code
):
2218 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*'
2219 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
2221 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
2222 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
2227 if v
in ('true', 'false', 'null'):
2229 elif v
.startswith('/*') or v
.startswith('//') or v
== ',':
2232 if v
[0] in ("'", '"'):
2233 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
2238 }.get(m
.group(0), m
.group(0)), v
[1:-1])
2240 for regex
, base
in INTEGER_TABLE
:
2241 im
= re
.match(regex
, v
)
2243 i
= int(im
.group(1), base
)
2244 return '"%d":' % i
if v
.endswith(':') else '%d' % i
2248 return re
.sub(r
'''(?sx)
2249 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
2250 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
2251 {comment}|,(?={skip}[\]}}])|
2252 [a-zA-Z_][.a-zA-Z_0-9]*|
2253 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
2255 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
2258 def qualities(quality_ids
):
2259 """ Get a numeric quality value out of a list of possible values """
2262 return quality_ids
.index(qid
)
2268 DEFAULT_OUTTMPL
= '%(title)s-%(id)s.%(ext)s'
2271 def limit_length(s
, length
):
2272 """ Add ellipses to overly long strings """
2277 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
2281 def version_tuple(v
):
2282 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
2285 def is_outdated_version(version
, limit
, assume_new
=True):
2287 return not assume_new
2289 return version_tuple(version
) < version_tuple(limit
)
2291 return not assume_new
2294 def ytdl_is_updateable():
2295 """ Returns if youtube-dl can be updated with -U """
2296 from zipimport
import zipimporter
2298 return isinstance(globals().get('__loader__'), zipimporter
) or hasattr(sys
, 'frozen')
2301 def args_to_str(args
):
2302 # Get a short string representation for a subprocess command
2303 return ' '.join(compat_shlex_quote(a
) for a
in args
)
2306 def error_to_compat_str(err
):
2308 # On python 2 error byte string must be decoded with proper
2309 # encoding rather than ascii
2310 if sys
.version_info
[0] < 3:
2311 err_str
= err_str
.decode(preferredencoding())
2315 def mimetype2ext(mt
):
2321 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
2322 # it's the most popular one
2323 'audio/mpeg': 'mp3',
2328 _
, _
, res
= mt
.rpartition('/')
2329 res
= res
.split(';')[0].strip().lower()
2333 'smptett+xml': 'tt',
2337 'x-mp4-fragmented': 'mp4',
2340 'x-mpegurl': 'm3u8',
2341 'vnd.apple.mpegurl': 'm3u8',
2345 'vnd.ms-sstr+xml': 'ism',
2351 def parse_codecs(codecs_str
):
2352 # http://tools.ietf.org/html/rfc6381
2355 splited_codecs
= list(filter(None, map(
2356 lambda str: str.strip(), codecs_str
.strip().strip(',').split(','))))
2357 vcodec
, acodec
= None, None
2358 for full_codec
in splited_codecs
:
2359 codec
= full_codec
.split('.')[0]
2360 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v'):
2363 elif codec
in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3'):
2367 write_string('WARNING: Unknown codec %s' % full_codec
, sys
.stderr
)
2368 if not vcodec
and not acodec
:
2369 if len(splited_codecs
) == 2:
2374 elif len(splited_codecs
) == 1:
2381 'vcodec': vcodec
or 'none',
2382 'acodec': acodec
or 'none',
2387 def urlhandle_detect_ext(url_handle
):
2388 getheader
= url_handle
.headers
.get
2390 cd
= getheader('Content-Disposition')
2392 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
2394 e
= determine_ext(m
.group('filename'), default_ext
=None)
2398 return mimetype2ext(getheader('Content-Type'))
2401 def encode_data_uri(data
, mime_type
):
2402 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
2405 def age_restricted(content_limit
, age_limit
):
2406 """ Returns True iff the content should be blocked """
2408 if age_limit
is None: # No limit set
2410 if content_limit
is None:
2411 return False # Content available for everyone
2412 return age_limit
< content_limit
2415 def is_html(first_bytes
):
2416 """ Detect whether a file contains HTML by examining its first bytes. """
2419 (b
'\xef\xbb\xbf', 'utf-8'),
2420 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
2421 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
2422 (b
'\xff\xfe', 'utf-16-le'),
2423 (b
'\xfe\xff', 'utf-16-be'),
2425 for bom
, enc
in BOMS
:
2426 if first_bytes
.startswith(bom
):
2427 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
2430 s
= first_bytes
.decode('utf-8', 'replace')
2432 return re
.match(r
'^\s*<', s
)
2435 def determine_protocol(info_dict
):
2436 protocol
= info_dict
.get('protocol')
2437 if protocol
is not None:
2440 url
= info_dict
['url']
2441 if url
.startswith('rtmp'):
2443 elif url
.startswith('mms'):
2445 elif url
.startswith('rtsp'):
2448 ext
= determine_ext(url
)
2454 return compat_urllib_parse_urlparse(url
).scheme
2457 def render_table(header_row
, data
):
2458 """ Render a list of rows, each as a list of values """
2459 table
= [header_row
] + data
2460 max_lens
= [max(len(compat_str(v
)) for v
in col
) for col
in zip(*table
)]
2461 format_str
= ' '.join('%-' + compat_str(ml
+ 1) + 's' for ml
in max_lens
[:-1]) + '%s'
2462 return '\n'.join(format_str
% tuple(row
) for row
in table
)
2465 def _match_one(filter_part
, dct
):
2466 COMPARISON_OPERATORS
= {
2474 operator_rex
= re
.compile(r
'''(?x)\s*
2476 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
2478 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
2479 (?P<quote>["\'])(?P
<quotedstrval
>(?
:\\.|
(?
!(?P
=quote
)|
\\).)+?
)(?P
=quote
)|
2480 (?P
<strval
>(?
![0-9.])[a
-z0
-9A
-Z
]*)
2483 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
2484 m = operator_rex.search(filter_part)
2486 op = COMPARISON_OPERATORS[m.group('op')]
2487 actual_value = dct.get(m.group('key'))
2488 if (m.group('quotedstrval') is not None or
2489 m.group('strval') is not None or
2490 # If the original field is a string and matching comparisonvalue is
2491 # a number we should respect the origin of the original field
2492 # and process comparison value as a string (see
2493 # https://github.com/rg3/youtube-dl/issues/11082).
2494 actual_value is not None and m.group('intval') is not None and
2495 isinstance(actual_value, compat_str)):
2496 if m.group('op') not in ('=', '!='):
2498 'Operator %s does not support string values!' % m.group('op'))
2499 comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval')
2500 quote = m.group('quote')
2501 if quote is not None:
2502 comparison_value = comparison_value.replace(r'\%s' % quote, quote)
2505 comparison_value = int(m.group('intval'))
2507 comparison_value = parse_filesize(m.group('intval'))
2508 if comparison_value is None:
2509 comparison_value = parse_filesize(m.group('intval') + 'B')
2510 if comparison_value is None:
2512 'Invalid integer value %r in filter part %r' % (
2513 m.group('intval'), filter_part))
2514 if actual_value is None:
2515 return m.group('none_inclusive')
2516 return op(actual_value, comparison_value)
2519 '': lambda v: v is not None,
2520 '!': lambda v: v is None,
2522 operator_rex = re.compile(r'''(?x
)\s
*
2523 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
2525 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
2526 m = operator_rex.search(filter_part)
2528 op = UNARY_OPERATORS[m.group('op')]
2529 actual_value = dct.get(m.group('key'))
2530 return op(actual_value)
2532 raise ValueError('Invalid filter part %r' % filter_part)
2535 def match_str(filter_str, dct):
2536 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """
2539 _match_one(filter_part, dct) for filter_part in filter_str.split('&'))
2542 def match_filter_func(filter_str):
2543 def _match_func(info_dict):
2544 if match_str(filter_str, info_dict):
2547 video_title = info_dict.get('title', info_dict.get('id', 'video'))
2548 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
2552 def parse_dfxp_time_expr(time_expr):
2556 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
2558 return float(mobj.group('time_offset'))
2560 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
2562 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
2565 def srt_subtitles_timecode(seconds):
2566 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
2569 def dfxp2srt(dfxp_data):
2570 LEGACY_NAMESPACES = (
2571 ('http://www.w3.org/ns/ttml', [
2572 'http://www.w3.org/2004/11/ttaf1',
2573 'http://www.w3.org/2006/04/ttaf1',
2574 'http://www.w3.org/2006/10/ttaf1',
2576 ('http://www.w3.org/ns/ttml#styling', [
2577 'http://www.w3.org/ns/ttml#style',
2581 SUPPORTED_STYLING = [
2590 _x = functools.partial(xpath_with_ns, ns_map={
2591 'ttml': 'http://www.w3.org/ns/ttml',
2592 'tts': 'http://www.w3.org/ns/ttml#styling',
2598 class TTMLPElementParser(object):
2600 _unclosed_elements = []
2601 _applied_styles = []
2603 def start(self, tag, attrib):
2604 if tag in (_x('ttml:br'), 'br'):
2607 unclosed_elements = []
2609 element_style_id = attrib.get('style')
2611 style.update(default_style)
2612 if element_style_id:
2613 style.update(styles.get(element_style_id, {}))
2614 for prop in SUPPORTED_STYLING:
2615 prop_val = attrib.get(_x('tts:' + prop))
2617 style[prop] = prop_val
2620 for k, v in sorted(style.items()):
2621 if self._applied_styles and self._applied_styles[-1].get(k) == v:
2624 font += ' color="%s"' % v
2625 elif k == 'fontSize':
2626 font += ' size="%s"' % v
2627 elif k == 'fontFamily':
2628 font += ' face="%s"' % v
2629 elif k == 'fontWeight' and v == 'bold':
2631 unclosed_elements.append('b')
2632 elif k == 'fontStyle' and v == 'italic':
2634 unclosed_elements.append('i')
2635 elif k == 'textDecoration' and v == 'underline':
2637 unclosed_elements.append('u')
2639 self._out += '<font' + font + '>'
2640 unclosed_elements.append('font')
2642 if self._applied_styles:
2643 applied_style.update(self._applied_styles[-1])
2644 applied_style.update(style)
2645 self._applied_styles.append(applied_style)
2646 self._unclosed_elements.append(unclosed_elements)
2649 if tag not in (_x('ttml:br'), 'br'):
2650 unclosed_elements = self._unclosed_elements.pop()
2651 for element in reversed(unclosed_elements):
2652 self._out += '</%s>' % element
2653 if unclosed_elements and self._applied_styles:
2654 self._applied_styles.pop()
2656 def data(self, data):
2660 return self._out.strip()
2662 def parse_node(node):
2663 target = TTMLPElementParser()
2664 parser = xml.etree.ElementTree.XMLParser(target=target)
2665 parser.feed(xml.etree.ElementTree.tostring(node))
2666 return parser.close()
2668 for k, v in LEGACY_NAMESPACES:
2670 dfxp_data = dfxp_data.replace(ns, k)
2672 dfxp = compat_etree_fromstring(dfxp_data.encode('utf-8'))
2674 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
2677 raise ValueError('Invalid dfxp/TTML subtitle')
2681 for style in dfxp.findall(_x('.//ttml:style')):
2682 style_id = style.get('id')
2683 parent_style_id = style.get('style')
2685 if parent_style_id not in styles:
2688 styles[style_id] = styles[parent_style_id].copy()
2689 for prop in SUPPORTED_STYLING:
2690 prop_val = style.get(_x('tts:' + prop))
2692 styles.setdefault(style_id, {})[prop] = prop_val
2698 for p in ('body', 'div'):
2699 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
2702 style = styles.get(ele.get('style'))
2705 default_style.update(style)
2707 for para, index in zip(paras, itertools.count(1)):
2708 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
2709 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
2710 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
2711 if begin_time is None:
2716 end_time = begin_time + dur
2717 out.append('%d\n%s --> %s\n%s\n\n' % (
2719 srt_subtitles_timecode(begin_time),
2720 srt_subtitles_timecode(end_time),
2726 def cli_option(params, command_option, param):
2727 param = params.get(param)
2729 param = compat_str(param)
2730 return [command_option, param] if param is not None else []
2733 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
2734 param = params.get(param)
2735 assert isinstance(param, bool)
2737 return [command_option + separator + (true_value if param else false_value)]
2738 return [command_option, true_value if param else false_value]
2741 def cli_valueless_option(params, command_option, param, expected_value=True):
2742 param = params.get(param)
2743 return [command_option] if param == expected_value else []
2746 def cli_configuration_args(params, param, default=[]):
2747 ex_args = params.get(param)
2750 assert isinstance(ex_args, list)
2754 class ISO639Utils(object):
2755 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
2944 def short2long(cls, code):
2945 """Convert language code from ISO 639-1 to ISO 639-2/T"""
2946 return cls._lang_map.get(code[:2])
2949 def long2short(cls, code):
2950 """Convert language code from ISO 639-2/T to ISO 639-1"""
2951 for short_name, long_name in cls._lang_map.items():
2952 if long_name == code:
2956 class ISO3166Utils(object):
2957 # From http://data.okfn.org/data/core/country-list
2959 'AF': 'Afghanistan',
2960 'AX': 'Åland Islands',
2963 'AS': 'American Samoa',
2968 'AG': 'Antigua and Barbuda',
2985 'BO': 'Bolivia, Plurinational State of',
2986 'BQ': 'Bonaire, Sint Eustatius and Saba',
2987 'BA': 'Bosnia and Herzegovina',
2989 'BV': 'Bouvet Island',
2991 'IO': 'British Indian Ocean Territory',
2992 'BN': 'Brunei Darussalam',
2994 'BF': 'Burkina Faso',
3000 'KY': 'Cayman Islands',
3001 'CF': 'Central African Republic',
3005 'CX': 'Christmas Island',
3006 'CC': 'Cocos (Keeling) Islands',
3010 'CD': 'Congo, the Democratic Republic of the',
3011 'CK': 'Cook Islands',
3013 'CI': 'Côte d\'Ivoire',
3018 'CZ': 'Czech Republic',
3022 'DO': 'Dominican Republic',
3025 'SV': 'El Salvador',
3026 'GQ': 'Equatorial Guinea',
3030 'FK': 'Falkland Islands (Malvinas)',
3031 'FO': 'Faroe Islands',
3035 'GF': 'French Guiana',
3036 'PF': 'French Polynesia',
3037 'TF': 'French Southern Territories',
3052 'GW': 'Guinea-Bissau',
3055 'HM': 'Heard Island and McDonald Islands',
3056 'VA': 'Holy See (Vatican City State)',
3063 'IR': 'Iran, Islamic Republic of',
3066 'IM': 'Isle of Man',
3076 'KP': 'Korea, Democratic People\'s Republic of',
3077 'KR': 'Korea, Republic of',
3080 'LA': 'Lao People\'s Democratic Republic',
3086 'LI': 'Liechtenstein',
3090 'MK': 'Macedonia, the Former Yugoslav Republic of',
3097 'MH': 'Marshall Islands',
3103 'FM': 'Micronesia, Federated States of',
3104 'MD': 'Moldova, Republic of',
3115 'NL': 'Netherlands',
3116 'NC': 'New Caledonia',
3117 'NZ': 'New Zealand',
3122 'NF': 'Norfolk Island',
3123 'MP': 'Northern Mariana Islands',
3128 'PS': 'Palestine, State of',
3130 'PG': 'Papua New Guinea',
3133 'PH': 'Philippines',
3137 'PR': 'Puerto Rico',
3141 'RU': 'Russian Federation',
3143 'BL': 'Saint Barthélemy',
3144 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
3145 'KN': 'Saint Kitts and Nevis',
3146 'LC': 'Saint Lucia',
3147 'MF': 'Saint Martin (French part)',
3148 'PM': 'Saint Pierre and Miquelon',
3149 'VC': 'Saint Vincent and the Grenadines',
3152 'ST': 'Sao Tome and Principe',
3153 'SA': 'Saudi Arabia',
3157 'SL': 'Sierra Leone',
3159 'SX': 'Sint Maarten (Dutch part)',
3162 'SB': 'Solomon Islands',
3164 'ZA': 'South Africa',
3165 'GS': 'South Georgia and the South Sandwich Islands',
3166 'SS': 'South Sudan',
3171 'SJ': 'Svalbard and Jan Mayen',
3174 'CH': 'Switzerland',
3175 'SY': 'Syrian Arab Republic',
3176 'TW': 'Taiwan, Province of China',
3178 'TZ': 'Tanzania, United Republic of',
3180 'TL': 'Timor-Leste',
3184 'TT': 'Trinidad and Tobago',
3187 'TM': 'Turkmenistan',
3188 'TC': 'Turks and Caicos Islands',
3192 'AE': 'United Arab Emirates',
3193 'GB': 'United Kingdom',
3194 'US': 'United States',
3195 'UM': 'United States Minor Outlying Islands',
3199 'VE': 'Venezuela, Bolivarian Republic of',
3201 'VG': 'Virgin Islands, British',
3202 'VI': 'Virgin Islands, U.S.',
3203 'WF': 'Wallis and Futuna',
3204 'EH': 'Western Sahara',
3211 def short2full(cls, code):
3212 """Convert an ISO 3166-2 country code to the corresponding full name"""
3213 return cls._country_map.get(code.upper())
3216 class GeoUtils(object):
3217 # Major IPv4 address blocks per country
3219 'AD': '85.94.160.0/19',
3220 'AE': '94.200.0.0/13',
3221 'AF': '149.54.0.0/17',
3222 'AG': '209.59.64.0/18',
3223 'AI': '204.14.248.0/21',
3224 'AL': '46.99.0.0/16',
3225 'AM': '46.70.0.0/15',
3226 'AO': '105.168.0.0/13',
3227 'AP': '159.117.192.0/21',
3228 'AR': '181.0.0.0/12',
3229 'AS': '202.70.112.0/20',
3230 'AT': '84.112.0.0/13',
3231 'AU': '1.128.0.0/11',
3232 'AW': '181.41.0.0/18',
3233 'AZ': '5.191.0.0/16',
3234 'BA': '31.176.128.0/17',
3235 'BB': '65.48.128.0/17',
3236 'BD': '114.130.0.0/16',
3238 'BF': '129.45.128.0/17',
3239 'BG': '95.42.0.0/15',
3240 'BH': '37.131.0.0/17',
3241 'BI': '154.117.192.0/18',
3242 'BJ': '137.255.0.0/16',
3243 'BL': '192.131.134.0/24',
3244 'BM': '196.12.64.0/18',
3245 'BN': '156.31.0.0/16',
3246 'BO': '161.56.0.0/16',
3247 'BQ': '161.0.80.0/20',
3248 'BR': '152.240.0.0/12',
3249 'BS': '24.51.64.0/18',
3250 'BT': '119.2.96.0/19',
3251 'BW': '168.167.0.0/16',
3252 'BY': '178.120.0.0/13',
3253 'BZ': '179.42.192.0/18',
3254 'CA': '99.224.0.0/11',
3255 'CD': '41.243.0.0/16',
3256 'CF': '196.32.200.0/21',
3257 'CG': '197.214.128.0/17',
3258 'CH': '85.0.0.0/13',
3259 'CI': '154.232.0.0/14',
3260 'CK': '202.65.32.0/19',
3261 'CL': '152.172.0.0/14',
3262 'CM': '165.210.0.0/15',
3263 'CN': '36.128.0.0/10',
3264 'CO': '181.240.0.0/12',
3265 'CR': '201.192.0.0/12',
3266 'CU': '152.206.0.0/15',
3267 'CV': '165.90.96.0/19',
3268 'CW': '190.88.128.0/17',
3269 'CY': '46.198.0.0/15',
3270 'CZ': '88.100.0.0/14',
3272 'DJ': '197.241.0.0/17',
3273 'DK': '87.48.0.0/12',
3274 'DM': '192.243.48.0/20',
3275 'DO': '152.166.0.0/15',
3276 'DZ': '41.96.0.0/12',
3277 'EC': '186.68.0.0/15',
3278 'EE': '90.190.0.0/15',
3279 'EG': '156.160.0.0/11',
3280 'ER': '196.200.96.0/20',
3281 'ES': '88.0.0.0/11',
3282 'ET': '196.188.0.0/14',
3283 'EU': '2.16.0.0/13',
3284 'FI': '91.152.0.0/13',
3285 'FJ': '144.120.0.0/16',
3286 'FM': '119.252.112.0/20',
3287 'FO': '88.85.32.0/19',
3289 'GA': '41.158.0.0/15',
3291 'GD': '74.122.88.0/21',
3292 'GE': '31.146.0.0/16',
3293 'GF': '161.22.64.0/18',
3294 'GG': '62.68.160.0/19',
3295 'GH': '45.208.0.0/14',
3296 'GI': '85.115.128.0/19',
3297 'GL': '88.83.0.0/19',
3298 'GM': '160.182.0.0/15',
3299 'GN': '197.149.192.0/18',
3300 'GP': '104.250.0.0/19',
3301 'GQ': '105.235.224.0/20',
3302 'GR': '94.64.0.0/13',
3303 'GT': '168.234.0.0/16',
3304 'GU': '168.123.0.0/16',
3305 'GW': '197.214.80.0/20',
3306 'GY': '181.41.64.0/18',
3307 'HK': '113.252.0.0/14',
3308 'HN': '181.210.0.0/16',
3309 'HR': '93.136.0.0/13',
3310 'HT': '148.102.128.0/17',
3311 'HU': '84.0.0.0/14',
3312 'ID': '39.192.0.0/10',
3313 'IE': '87.32.0.0/12',
3314 'IL': '79.176.0.0/13',
3315 'IM': '5.62.80.0/20',
3316 'IN': '117.192.0.0/10',
3317 'IO': '203.83.48.0/21',
3318 'IQ': '37.236.0.0/14',
3319 'IR': '2.176.0.0/12',
3320 'IS': '82.221.0.0/16',
3321 'IT': '79.0.0.0/10',
3322 'JE': '87.244.64.0/18',
3323 'JM': '72.27.0.0/17',
3324 'JO': '176.29.0.0/16',
3325 'JP': '126.0.0.0/8',
3326 'KE': '105.48.0.0/12',
3327 'KG': '158.181.128.0/17',
3328 'KH': '36.37.128.0/17',
3329 'KI': '103.25.140.0/22',
3330 'KM': '197.255.224.0/20',
3331 'KN': '198.32.32.0/19',
3332 'KP': '175.45.176.0/22',
3333 'KR': '175.192.0.0/10',
3334 'KW': '37.36.0.0/14',
3335 'KY': '64.96.0.0/15',
3336 'KZ': '2.72.0.0/13',
3337 'LA': '115.84.64.0/18',
3338 'LB': '178.135.0.0/16',
3339 'LC': '192.147.231.0/24',
3340 'LI': '82.117.0.0/19',
3341 'LK': '112.134.0.0/15',
3342 'LR': '41.86.0.0/19',
3343 'LS': '129.232.0.0/17',
3344 'LT': '78.56.0.0/13',
3345 'LU': '188.42.0.0/16',
3346 'LV': '46.109.0.0/16',
3347 'LY': '41.252.0.0/14',
3348 'MA': '105.128.0.0/11',
3349 'MC': '88.209.64.0/18',
3350 'MD': '37.246.0.0/16',
3351 'ME': '178.175.0.0/17',
3352 'MF': '74.112.232.0/21',
3353 'MG': '154.126.0.0/17',
3354 'MH': '117.103.88.0/21',
3355 'MK': '77.28.0.0/15',
3356 'ML': '154.118.128.0/18',
3357 'MM': '37.111.0.0/17',
3358 'MN': '49.0.128.0/17',
3359 'MO': '60.246.0.0/16',
3360 'MP': '202.88.64.0/20',
3361 'MQ': '109.203.224.0/19',
3362 'MR': '41.188.64.0/18',
3363 'MS': '208.90.112.0/22',
3364 'MT': '46.11.0.0/16',
3365 'MU': '105.16.0.0/12',
3366 'MV': '27.114.128.0/18',
3367 'MW': '105.234.0.0/16',
3368 'MX': '187.192.0.0/11',
3369 'MY': '175.136.0.0/13',
3370 'MZ': '197.218.0.0/15',
3371 'NA': '41.182.0.0/16',
3372 'NC': '101.101.0.0/18',
3373 'NE': '197.214.0.0/18',
3374 'NF': '203.17.240.0/22',
3375 'NG': '105.112.0.0/12',
3376 'NI': '186.76.0.0/15',
3377 'NL': '145.96.0.0/11',
3378 'NO': '84.208.0.0/13',
3379 'NP': '36.252.0.0/15',
3380 'NR': '203.98.224.0/19',
3381 'NU': '49.156.48.0/22',
3382 'NZ': '49.224.0.0/14',
3383 'OM': '5.36.0.0/15',
3384 'PA': '186.72.0.0/15',
3385 'PE': '186.160.0.0/14',
3386 'PF': '123.50.64.0/18',
3387 'PG': '124.240.192.0/19',
3388 'PH': '49.144.0.0/13',
3389 'PK': '39.32.0.0/11',
3390 'PL': '83.0.0.0/11',
3391 'PM': '70.36.0.0/20',
3392 'PR': '66.50.0.0/16',
3393 'PS': '188.161.0.0/16',
3394 'PT': '85.240.0.0/13',
3395 'PW': '202.124.224.0/20',
3396 'PY': '181.120.0.0/14',
3397 'QA': '37.210.0.0/15',
3398 'RE': '139.26.0.0/16',
3399 'RO': '79.112.0.0/13',
3400 'RS': '178.220.0.0/14',
3401 'RU': '5.136.0.0/13',
3402 'RW': '105.178.0.0/15',
3403 'SA': '188.48.0.0/13',
3404 'SB': '202.1.160.0/19',
3405 'SC': '154.192.0.0/11',
3406 'SD': '154.96.0.0/13',
3407 'SE': '78.64.0.0/12',
3408 'SG': '152.56.0.0/14',
3409 'SI': '188.196.0.0/14',
3410 'SK': '78.98.0.0/15',
3411 'SL': '197.215.0.0/17',
3412 'SM': '89.186.32.0/19',
3413 'SN': '41.82.0.0/15',
3414 'SO': '197.220.64.0/19',
3415 'SR': '186.179.128.0/17',
3416 'SS': '105.235.208.0/21',
3417 'ST': '197.159.160.0/19',
3418 'SV': '168.243.0.0/16',
3419 'SX': '190.102.0.0/20',
3421 'SZ': '41.84.224.0/19',
3422 'TC': '65.255.48.0/20',
3423 'TD': '154.68.128.0/19',
3424 'TG': '196.168.0.0/14',
3425 'TH': '171.96.0.0/13',
3426 'TJ': '85.9.128.0/18',
3427 'TK': '27.96.24.0/21',
3428 'TL': '180.189.160.0/20',
3429 'TM': '95.85.96.0/19',
3430 'TN': '197.0.0.0/11',
3431 'TO': '175.176.144.0/21',
3432 'TR': '78.160.0.0/11',
3433 'TT': '186.44.0.0/15',
3434 'TV': '202.2.96.0/19',
3435 'TW': '120.96.0.0/11',
3436 'TZ': '156.156.0.0/14',
3437 'UA': '93.72.0.0/13',
3438 'UG': '154.224.0.0/13',
3440 'UY': '167.56.0.0/13',
3441 'UZ': '82.215.64.0/18',
3442 'VA': '212.77.0.0/19',
3443 'VC': '24.92.144.0/20',
3444 'VE': '186.88.0.0/13',
3445 'VG': '172.103.64.0/18',
3446 'VI': '146.226.0.0/16',
3447 'VN': '14.160.0.0/11',
3448 'VU': '202.80.32.0/20',
3449 'WF': '117.20.32.0/21',
3450 'WS': '202.4.32.0/19',
3451 'YE': '134.35.0.0/16',
3452 'YT': '41.242.116.0/22',
3453 'ZA': '41.0.0.0/11',
3454 'ZM': '165.56.0.0/13',
3455 'ZW': '41.85.192.0/19',
3459 def random_ipv4(cls, code):
3460 block = cls._country_ip_map.get(code.upper())
3463 addr, preflen = block.split('/')
3464 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
3465 addr_max = addr_min | (0xffffffff >> int(preflen))
3466 return compat_str(socket.inet_ntoa(
3467 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
3470 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
3471 def __init__(self, proxies=None):
3472 # Set default handlers
3473 for type in ('http', 'https'):
3474 setattr(self, '%s_open' % type,
3475 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
3476 meth(r, proxy, type))
3477 return compat_urllib_request.ProxyHandler.__init__(self, proxies)
3479 def proxy_open(self, req, proxy, type):
3480 req_proxy = req.headers.get('Ytdl-request-proxy')
3481 if req_proxy is not None:
3483 del req.headers['Ytdl-request-proxy']
3485 if proxy == '__noproxy__':
3486 return None # No Proxy
3487 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
3488 req.add_header('Ytdl-socks-proxy', proxy)
3489 # youtube-dl's http/https handlers do wrapping the socket with socks
3491 return compat_urllib_request.ProxyHandler.proxy_open(
3492 self, req, proxy, type)
3495 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
3496 # released into Public Domain
3497 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
3499 def long_to_bytes(n, blocksize=0):
3500 """long_to_bytes(n:long, blocksize:int) : string
3501 Convert a long integer to a byte string.
3503 If optional blocksize is given and greater than zero, pad the front of the
3504 byte string with binary zeros so that the length is a multiple of
3507 # after much testing, this algorithm was deemed to be the fastest
3511 s = compat_struct_pack('>I', n & 0xffffffff) + s
3513 # strip off leading zeros
3514 for i in range(len(s)):
3515 if s[i] != b'\000'[0]:
3518 # only happens when n == 0
3522 # add back some pad bytes. this could be done more efficiently w.r.t. the
3523 # de-padding being done above, but sigh...
3524 if blocksize > 0 and len(s) % blocksize:
3525 s = (blocksize - len(s) % blocksize) * b'\000' + s
3529 def bytes_to_long(s):
3530 """bytes_to_long(string) : long
3531 Convert a byte string to a long integer.
3533 This is (essentially) the inverse of long_to_bytes().
3538 extra = (4 - length % 4)
3539 s = b'\000' * extra + s
3540 length = length + extra
3541 for i in range(0, length, 4):
3542 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
3546 def ohdave_rsa_encrypt(data, exponent, modulus):
3548 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
3551 data: data to encrypt, bytes-like object
3552 exponent, modulus: parameter e and N of RSA algorithm, both integer
3553 Output: hex string of encrypted data
3555 Limitation: supports one block encryption only
3558 payload = int(binascii.hexlify(data[::-1]), 16)
3559 encrypted = pow(payload, exponent, modulus)
3560 return '%x' % encrypted
3563 def pkcs1pad(data, length):
3565 Padding input data with PKCS#1 scheme
3567 @param {int[]} data input data
3568 @param {int} length target length
3569 @returns {int[]} padded data
3571 if len(data) > length - 11:
3572 raise ValueError('Input data too
long for PKCS
#1 padding')
3574 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
3575 return [0, 2] + pseudo_random
+ [0] + data
3578 def encode_base_n(num
, n
, table
=None):
3579 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
3581 table
= FULL_TABLE
[:n
]
3584 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
3591 ret
= table
[num
% n
] + ret
3596 def decode_packed_codes(code
):
3597 mobj
= re
.search(PACKED_CODES_RE
, code
)
3598 obfucasted_code
, base
, count
, symbols
= mobj
.groups()
3601 symbols
= symbols
.split('|')
3606 base_n_count
= encode_base_n(count
, base
)
3607 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
3610 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
3614 def parse_m3u8_attributes(attrib
):
3616 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
3617 if val
.startswith('"'):
3623 def urshift(val
, n
):
3624 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
3627 # Based on png2str() written by @gdkchan and improved by @yokrysty
3628 # Originally posted at https://github.com/rg3/youtube-dl/issues/9706
3629 def decode_png(png_data
):
3630 # Reference: https://www.w3.org/TR/PNG/
3631 header
= png_data
[8:]
3633 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
3634 raise IOError('Not a valid PNG file.')
3636 int_map
= {1: '>B', 2: '>H', 4: '>I'}
3637 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
3642 length
= unpack_integer(header
[:4])
3645 chunk_type
= header
[:4]
3648 chunk_data
= header
[:length
]
3649 header
= header
[length
:]
3651 header
= header
[4:] # Skip CRC
3659 ihdr
= chunks
[0]['data']
3661 width
= unpack_integer(ihdr
[:4])
3662 height
= unpack_integer(ihdr
[4:8])
3666 for chunk
in chunks
:
3667 if chunk
['type'] == b
'IDAT':
3668 idat
+= chunk
['data']
3671 raise IOError('Unable to read PNG data.')
3673 decompressed_data
= bytearray(zlib
.decompress(idat
))
3678 def _get_pixel(idx
):
3683 for y
in range(height
):
3684 basePos
= y
* (1 + stride
)
3685 filter_type
= decompressed_data
[basePos
]
3689 pixels
.append(current_row
)
3691 for x
in range(stride
):
3692 color
= decompressed_data
[1 + basePos
+ x
]
3693 basex
= y
* stride
+ x
3698 left
= _get_pixel(basex
- 3)
3700 up
= _get_pixel(basex
- stride
)
3702 if filter_type
== 1: # Sub
3703 color
= (color
+ left
) & 0xff
3704 elif filter_type
== 2: # Up
3705 color
= (color
+ up
) & 0xff
3706 elif filter_type
== 3: # Average
3707 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
3708 elif filter_type
== 4: # Paeth
3714 c
= _get_pixel(basex
- stride
- 3)
3722 if pa
<= pb
and pa
<= pc
:
3723 color
= (color
+ a
) & 0xff
3725 color
= (color
+ b
) & 0xff
3727 color
= (color
+ c
) & 0xff
3729 current_row
.append(color
)
3731 return width
, height
, pixels
3734 def write_xattr(path
, key
, value
):
3735 # This mess below finds the best xattr tool for the job
3737 # try the pyxattr module...
3740 if hasattr(xattr
, 'set'): # pyxattr
3741 # Unicode arguments are not supported in python-pyxattr until
3743 # See https://github.com/rg3/youtube-dl/issues/5498
3744 pyxattr_required_version
= '0.5.0'
3745 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
3746 # TODO: fallback to CLI tools
3747 raise XAttrUnavailableError(
3748 'python-pyxattr is detected but is too old. '
3749 'youtube-dl requires %s or above while your version is %s. '
3750 'Falling back to other xattr implementations' % (
3751 pyxattr_required_version
, xattr
.__version
__))
3753 setxattr
= xattr
.set
3755 setxattr
= xattr
.setxattr
3758 setxattr(path
, key
, value
)
3759 except EnvironmentError as e
:
3760 raise XAttrMetadataError(e
.errno
, e
.strerror
)
3763 if compat_os_name
== 'nt':
3764 # Write xattrs to NTFS Alternate Data Streams:
3765 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
3766 assert ':' not in key
3767 assert os
.path
.exists(path
)
3769 ads_fn
= path
+ ':' + key
3771 with open(ads_fn
, 'wb') as f
:
3773 except EnvironmentError as e
:
3774 raise XAttrMetadataError(e
.errno
, e
.strerror
)
3776 user_has_setfattr
= check_executable('setfattr', ['--version'])
3777 user_has_xattr
= check_executable('xattr', ['-h'])
3779 if user_has_setfattr
or user_has_xattr
:
3781 value
= value
.decode('utf-8')
3782 if user_has_setfattr
:
3783 executable
= 'setfattr'
3784 opts
= ['-n', key
, '-v', value
]
3785 elif user_has_xattr
:
3786 executable
= 'xattr'
3787 opts
= ['-w', key
, value
]
3789 cmd
= ([encodeFilename(executable
, True)] +
3790 [encodeArgument(o
) for o
in opts
] +
3791 [encodeFilename(path
, True)])
3794 p
= subprocess
.Popen(
3795 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
3796 except EnvironmentError as e
:
3797 raise XAttrMetadataError(e
.errno
, e
.strerror
)
3798 stdout
, stderr
= p
.communicate()
3799 stderr
= stderr
.decode('utf-8', 'replace')
3800 if p
.returncode
!= 0:
3801 raise XAttrMetadataError(p
.returncode
, stderr
)
3804 # On Unix, and can't find pyxattr, setfattr, or xattr.
3805 if sys
.platform
.startswith('linux'):
3806 raise XAttrUnavailableError(
3807 "Couldn't find a tool to set the xattrs. "
3808 "Install either the python 'pyxattr' or 'xattr' "
3809 "modules, or the GNU 'attr' package "
3810 "(which contains the 'setfattr' tool).")
3812 raise XAttrUnavailableError(
3813 "Couldn't find a tool to set the xattrs. "
3814 "Install either the python 'xattr' module, "
3815 "or the 'xattr' binary.")
3818 def random_birthday(year_field
, month_field
, day_field
):
3820 year_field
: str(random
.randint(1950, 1995)),
3821 month_field
: str(random
.randint(1, 12)),
3822 day_field
: str(random
.randint(1, 31)),