2 # -*- coding: utf-8 -*-
4 from __future__ import unicode_literals
32 import xml.etree.ElementTree
42 compat_socket_create_connection,
46 compat_urllib_parse_urlparse,
47 compat_urllib_request,
53 # This is not clearly defined otherwise
54 compiled_regex_type = type(re.compile(''))
57 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)',
58 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
59 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
60 'Accept-Encoding': 'gzip, deflate',
61 'Accept-Language': 'en-us,en;q=0.5',
67 ENGLISH_MONTH_NAMES = [
68 'January', 'February', 'March', 'April', 'May', 'June',
69 'July', 'August', 'September', 'October', 'November', 'December']
72 def preferredencoding():
73 """Get preferred encoding.
75 Returns the best encoding scheme for the system, based on
76 locale.getpreferredencoding() and some further tweaks.
79 pref = locale.getpreferredencoding()
87 def write_json_file(obj, fn):
88 """ Encode obj as JSON and write it to fn, atomically if possible """
90 fn = encodeFilename(fn)
91 if sys.version_info < (3, 0) and sys.platform != 'win32':
92 encoding = get_filesystem_encoding()
93 # os.path.basename returns a bytes object, but NamedTemporaryFile
94 # will fail if the filename contains non ascii characters unless we
95 # use a unicode object
96 path_basename = lambda f: os.path.basename(fn).decode(encoding)
97 # the same for os.path.dirname
98 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
100 path_basename = os.path.basename
101 path_dirname = os.path.dirname
105 'prefix': path_basename(fn) + '.',
106 'dir': path_dirname(fn),
110 # In Python 2.x, json.dump expects a bytestream.
111 # In Python 3.x, it writes to a character stream
112 if sys.version_info < (3, 0):
120 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
125 if sys.platform == 'win32':
126 # Need to remove existing file on Windows, else os.rename raises
127 # WindowsError or FileExistsError.
132 os.rename(tf.name, fn)
141 if sys.version_info >= (2, 7):
142 def find_xpath_attr(node, xpath, key, val=None):
143 """ Find the xpath xpath[@key=val] """
144 assert re.match(r'^[a-zA-Z_-]+$', key)
146 assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val)
147 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
148 return node.find(expr)
150 def find_xpath_attr(node, xpath, key, val=None):
151 # Here comes the crazy part: In 2.6, if the xpath is a unicode,
152 # .//node does not match if a node is a direct child of . !
153 if isinstance(xpath, compat_str):
154 xpath = xpath.encode('ascii')
156 for f in node.findall(xpath):
157 if key not in f.attrib:
159 if val is None or f.attrib.get(key) == val:
163 # On python2.6 the xml.etree.ElementTree.Element methods don't support
164 # the namespace parameter
167 def xpath_with_ns(path, ns_map):
168 components = [c.split(':') for c in path.split('/')]
172 replaced.append(c[0])
175 replaced.append('{%s}%s' % (ns_map[ns], tag))
176 return '/'.join(replaced)
179 def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
180 if sys.version_info < (2, 7): # Crazy 2.6
181 xpath = xpath.encode('ascii')
185 if default is not NO_DEFAULT:
188 name = xpath if name is None else name
189 raise ExtractorError('Could not find XML element %s' % name)
195 def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
196 n = xpath_element(node, xpath, name, fatal=fatal, default=default)
197 if n is None or n == default:
200 if default is not NO_DEFAULT:
203 name = xpath if name is None else name
204 raise ExtractorError('Could not find XML element\'s text %s' % name)
210 def xpath_attr(node, xpath, key, name=None, fatal=False, default=NO_DEFAULT):
211 n = find_xpath_attr(node, xpath, key)
213 if default is not NO_DEFAULT:
216 name = '%s[@%s]' % (xpath, key) if name is None else name
217 raise ExtractorError('Could not find XML attribute %s' % name)
223 def get_element_by_id(id, html):
224 """Return the content of the tag with the specified ID in the passed HTML document"""
225 return get_element_by_attribute("id", id, html)
228 def get_element_by_attribute(attribute, value, html):
229 """Return the content of the tag with the specified attribute in the passed HTML document"""
231 m = re.search(r'''(?xs)
233 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
235 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
239 ''' % (re.escape(attribute), re.escape(value)), html)
243 res = m.group('content')
245 if res.startswith('"') or res.startswith("'"):
248 return unescapeHTML(res)
251 def clean_html(html):
252 """Clean an HTML snippet into a readable string"""
254 if html is None: # Convenience for sanitizing descriptions etc.
258 html = html.replace('\n', ' ')
259 html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html)
260 html = re.sub(r'<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
262 html = re.sub('<.*?>', '', html)
263 # Replace html entities
264 html = unescapeHTML(html)
268 def sanitize_open(filename, open_mode):
269 """Try to open the given filename, and slightly tweak it if this fails.
271 Attempts to open the given filename. If this fails, it tries to change
272 the filename slightly, step by step, until it's either able to open it
273 or it fails and raises a final exception, like the standard open()
276 It returns the tuple (stream, definitive_file_name).
280 if sys.platform == 'win32':
282 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
283 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
284 stream = open(encodeFilename(filename), open_mode)
285 return (stream, filename)
286 except (IOError, OSError) as err:
287 if err.errno in (errno.EACCES,):
290 # In case of error, try to remove win32 forbidden chars
291 alt_filename = sanitize_path(filename)
292 if alt_filename == filename:
295 # An exception here should be caught in the caller
296 stream = open(encodeFilename(alt_filename), open_mode)
297 return (stream, alt_filename)
300 def timeconvert(timestr):
301 """Convert RFC 2822 defined time string into system timestamp"""
303 timetuple = email.utils.parsedate_tz(timestr)
304 if timetuple is not None:
305 timestamp = email.utils.mktime_tz(timetuple)
309 def sanitize_filename(s, restricted=False, is_id=False):
310 """Sanitizes a string so it could be used as part of a filename.
311 If restricted is set, use a stricter subset of allowed characters.
312 Set is_id if this is not an arbitrary string, but an ID that should be kept if possible
314 def replace_insane(char):
315 if char == '?' or ord(char) < 32 or ord(char) == 127:
318 return '' if restricted else '\''
320 return '_-' if restricted else ' -'
321 elif char in '\\/|*<>':
323 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
325 if restricted and ord(char) > 127:
330 s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
331 result = ''.join(map(replace_insane, s))
333 while '__' in result:
334 result = result.replace('__', '_')
335 result = result.strip('_')
336 # Common case of "Foreign band name - English song title"
337 if restricted and result.startswith('-_'):
339 if result.startswith('-'):
340 result = '_' + result[len('-'):]
341 result = result.lstrip('.')
347 def sanitize_path(s):
348 """Sanitizes and normalizes path on Windows"""
349 if sys.platform != 'win32':
351 drive_or_unc, _ = os.path.splitdrive(s)
352 if sys.version_info < (2, 7) and not drive_or_unc:
353 drive_or_unc, _ = os.path.splitunc(s)
354 norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
358 path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|\.$)', '#', path_part)
359 for path_part in norm_path]
361 sanitized_path.insert(0, drive_or_unc + os.path.sep)
362 return os.path.join(*sanitized_path)
365 def orderedSet(iterable):
366 """ Remove all duplicates from the input iterable """
374 def _htmlentity_transform(entity):
375 """Transforms an HTML entity to a character."""
376 # Known non-numeric HTML entity
377 if entity in compat_html_entities.name2codepoint:
378 return compat_chr(compat_html_entities.name2codepoint[entity])
380 mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
382 numstr = mobj.group(1)
383 if numstr.startswith('x'):
385 numstr = '0%s' % numstr
388 return compat_chr(int(numstr, base))
390 # Unknown entity in name, return its literal representation
391 return ('&%s;' % entity)
397 assert type(s) == compat_str
400 r'&([^;]+);', lambda m: _htmlentity_transform(m.group(1)), s)
403 def get_subprocess_encoding():
404 if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
405 # For subprocess calls, encode with locale encoding
406 # Refer to http://stackoverflow.com/a/9951851/35070
407 encoding = preferredencoding()
409 encoding = sys.getfilesystemencoding()
415 def encodeFilename(s, for_subprocess=False):
417 @param s The name of the file
420 assert type(s) == compat_str
422 # Python 3 has a Unicode API
423 if sys.version_info >= (3, 0):
426 # Pass '' directly to use Unicode APIs on Windows 2000 and up
427 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
428 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
429 if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
432 return s.encode(get_subprocess_encoding(), 'ignore')
435 def decodeFilename(b, for_subprocess=False):
437 if sys.version_info >= (3, 0):
440 if not isinstance(b, bytes):
443 return b.decode(get_subprocess_encoding(), 'ignore')
446 def encodeArgument(s):
447 if not isinstance(s, compat_str):
448 # Legacy code that uses byte strings
449 # Uncomment the following line after fixing all post processors
450 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
451 s = s.decode('ascii')
452 return encodeFilename(s, True)
455 def decodeArgument(b):
456 return decodeFilename(b, True)
459 def decodeOption(optval):
462 if isinstance(optval, bytes):
463 optval = optval.decode(preferredencoding())
465 assert isinstance(optval, compat_str)
469 def formatSeconds(secs):
471 return '%d:%02d:%02d' % (secs // 3600, (secs % 3600) // 60, secs % 60)
473 return '%d:%02d' % (secs // 60, secs % 60)
478 def make_HTTPS_handler(params, **kwargs):
479 opts_no_check_certificate = params.get('nocheckcertificate', False)
480 if hasattr(ssl, 'create_default_context'): # Python >= 3.4 or 2.7.9
481 context = ssl.create_default_context(ssl.Purpose.SERVER_AUTH)
482 if opts_no_check_certificate:
483 context.check_hostname = False
484 context.verify_mode = ssl.CERT_NONE
486 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
489 # (create_default_context present but HTTPSHandler has no context=)
492 if sys.version_info < (3, 2):
493 return YoutubeDLHTTPSHandler(params, **kwargs)
495 context = ssl.SSLContext(ssl.PROTOCOL_TLSv1)
496 context.verify_mode = (ssl.CERT_NONE
497 if opts_no_check_certificate
498 else ssl.CERT_REQUIRED)
499 context.set_default_verify_paths()
500 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
503 def bug_reports_message():
504 if ytdl_is_updateable():
505 update_cmd = 'type youtube-dl -U to update'
507 update_cmd = 'see https://yt-dl.org/update on how to update'
508 msg = '; please report this issue on https://yt-dl.org/bug .'
509 msg += ' Make sure you are using the latest version; %s.' % update_cmd
510 msg += ' Be sure to call youtube-dl with the --verbose flag and include its complete output.'
514 class ExtractorError(Exception):
515 """Error during info extraction."""
517 def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):
518 """ tb, if given, is the original traceback (so that it can be printed out).
519 If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
522 if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError):
524 if video_id is not None:
525 msg = video_id + ': ' + msg
527 msg += ' (caused by %r)' % cause
529 msg += bug_reports_message()
530 super(ExtractorError, self).__init__(msg)
533 self.exc_info = sys.exc_info() # preserve original exception
535 self.video_id = video_id
537 def format_traceback(self):
538 if self.traceback is None:
540 return ''.join(traceback.format_tb(self.traceback))
543 class UnsupportedError(ExtractorError):
544 def __init__(self, url):
545 super(UnsupportedError, self).__init__(
546 'Unsupported URL: %s' % url, expected=True)
550 class RegexNotFoundError(ExtractorError):
551 """Error when a regex didn't match"""
555 class DownloadError(Exception):
556 """Download Error exception.
558 This exception may be thrown by FileDownloader objects if they are not
559 configured to continue on errors. They will contain the appropriate
563 def __init__(self, msg, exc_info=None):
564 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
565 super(DownloadError, self).__init__(msg)
566 self.exc_info = exc_info
569 class SameFileError(Exception):
570 """Same File exception.
572 This exception will be thrown by FileDownloader objects if they detect
573 multiple files would have to be downloaded to the same file on disk.
578 class PostProcessingError(Exception):
579 """Post Processing exception.
581 This exception may be raised by PostProcessor's .run() method to
582 indicate an error in the postprocessing task.
585 def __init__(self, msg):
589 class MaxDownloadsReached(Exception):
590 """ --max-downloads limit has been reached. """
594 class UnavailableVideoError(Exception):
595 """Unavailable Format exception.
597 This exception will be thrown when a video is requested
598 in a format that is not available for that video.
603 class ContentTooShortError(Exception):
604 """Content Too Short exception.
606 This exception may be raised by FileDownloader objects when a file they
607 download is too small for what the server announced first, indicating
608 the connection was probably interrupted.
611 def __init__(self, downloaded, expected):
613 self.downloaded = downloaded
614 self.expected = expected
617 def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
618 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
619 # expected HTTP responses to meet HTTP/1.0 or later (see also
620 # https://github.com/rg3/youtube-dl/issues/6727)
621 if sys.version_info < (3, 0):
622 kwargs['strict'] = True
623 hc = http_class(*args, **kwargs)
624 source_address = ydl_handler._params.get('source_address')
625 if source_address is not None:
626 sa = (source_address, 0)
627 if hasattr(hc, 'source_address'): # Python 2.7+
628 hc.source_address = sa
630 def _hc_connect(self, *args, **kwargs):
631 sock = compat_socket_create_connection(
632 (self.host, self.port), self.timeout, sa)
634 self.sock = ssl.wrap_socket(
635 sock, self.key_file, self.cert_file,
636 ssl_version=ssl.PROTOCOL_TLSv1)
639 hc.connect = functools.partial(_hc_connect, hc)
644 class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
645 """Handler for HTTP requests and responses.
647 This class, when installed with an OpenerDirector, automatically adds
648 the standard headers to every HTTP request and handles gzipped and
649 deflated responses from web servers. If compression is to be avoided in
650 a particular request, the original request in the program code only has
651 to include the HTTP header "Youtubedl-No-Compression", which will be
652 removed before making the real request.
654 Part of this code was copied from:
656 http://techknack.net/python-urllib2-handlers/
658 Andrew Rowls, the author of that code, agreed to release it to the
662 def __init__(self, params, *args, **kwargs):
663 compat_urllib_request.HTTPHandler.__init__(self, *args, **kwargs)
664 self._params = params
666 def http_open(self, req):
667 return self.do_open(functools.partial(
668 _create_http_connection, self, compat_http_client.HTTPConnection, False),
674 return zlib.decompress(data, -zlib.MAX_WBITS)
676 return zlib.decompress(data)
679 def addinfourl_wrapper(stream, headers, url, code):
680 if hasattr(compat_urllib_request.addinfourl, 'getcode'):
681 return compat_urllib_request.addinfourl(stream, headers, url, code)
682 ret = compat_urllib_request.addinfourl(stream, headers, url)
686 def http_request(self, req):
687 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
688 # always respected by websites, some tend to give out URLs with non percent-encoded
689 # non-ASCII characters (see telemb.py, ard.py [#3412])
690 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
691 # To work around aforementioned issue we will replace request's original URL with
692 # percent-encoded one
693 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
694 # the code of this workaround has been moved here from YoutubeDL.urlopen()
695 url = req.get_full_url()
696 url_escaped = escape_url(url)
698 # Substitute URL if any change after escaping
699 if url != url_escaped:
700 req_type = HEADRequest if req.get_method() == 'HEAD' else compat_urllib_request.Request
702 url_escaped, data=req.data, headers=req.headers,
703 origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
704 new_req.timeout = req.timeout
707 for h, v in std_headers.items():
708 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
709 # The dict keys are capitalized because of this bug by urllib
710 if h.capitalize() not in req.headers:
712 if 'Youtubedl-no-compression' in req.headers:
713 if 'Accept-encoding' in req.headers:
714 del req.headers['Accept-encoding']
715 del req.headers['Youtubedl-no-compression']
717 if sys.version_info < (2, 7) and '#' in req.get_full_url():
718 # Python 2.6 is brain-dead when it comes to fragments
719 req._Request__original = req._Request__original.partition('#')[0]
720 req._Request__r_type = req._Request__r_type.partition('#')[0]
724 def http_response(self, req, resp):
727 if resp.headers.get('Content-encoding', '') == 'gzip':
728 content = resp.read()
729 gz = gzip.GzipFile(fileobj=io.BytesIO(content), mode='rb')
731 uncompressed = io.BytesIO(gz.read())
732 except IOError as original_ioerror:
733 # There may be junk add the end of the file
734 # See http://stackoverflow.com/q/4928560/35070 for details
735 for i in range(1, 1024):
737 gz = gzip.GzipFile(fileobj=io.BytesIO(content[:-i]), mode='rb')
738 uncompressed = io.BytesIO(gz.read())
743 raise original_ioerror
744 resp = self.addinfourl_wrapper(uncompressed, old_resp.headers, old_resp.url, old_resp.code)
745 resp.msg = old_resp.msg
747 if resp.headers.get('Content-encoding', '') == 'deflate':
748 gz = io.BytesIO(self.deflate(resp.read()))
749 resp = self.addinfourl_wrapper(gz, old_resp.headers, old_resp.url, old_resp.code)
750 resp.msg = old_resp.msg
751 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986
752 if 300 <= resp.code < 400:
753 location = resp.headers.get('Location')
755 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
756 if sys.version_info >= (3, 0):
757 location = location.encode('iso-8859-1').decode('utf-8')
758 location_escaped = escape_url(location)
759 if location != location_escaped:
760 del resp.headers['Location']
761 resp.headers['Location'] = location_escaped
764 https_request = http_request
765 https_response = http_response
768 class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
769 def __init__(self, params, https_conn_class=None, *args, **kwargs):
770 compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)
771 self._https_conn_class = https_conn_class or compat_http_client.HTTPSConnection
772 self._params = params
774 def https_open(self, req):
776 if hasattr(self, '_context'): # python > 2.6
777 kwargs['context'] = self._context
778 if hasattr(self, '_check_hostname'): # python 3.x
779 kwargs['check_hostname'] = self._check_hostname
780 return self.do_open(functools.partial(
781 _create_http_connection, self, self._https_conn_class, True),
785 def parse_iso8601(date_str, delimiter='T', timezone=None):
786 """ Return a UNIX timestamp from the given date """
793 r'(\.[0-9]+)?(?:Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
796 timezone = datetime.timedelta()
798 date_str = date_str[:-len(m.group(0))]
799 if not m.group('sign'):
800 timezone = datetime.timedelta()
802 sign = 1 if m.group('sign') == '+' else -1
803 timezone = datetime.timedelta(
804 hours=sign * int(m.group('hours')),
805 minutes=sign * int(m.group('minutes')))
806 date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
807 dt = datetime.datetime.strptime(date_str, date_format) - timezone
808 return calendar.timegm(dt.timetuple())
811 def unified_strdate(date_str, day_first=True):
812 """Return a string with the date in the format YYYYMMDD"""
818 date_str = date_str.replace(',', ' ')
819 # %z (UTC offset) is only supported in python>=3.2
820 if not re.match(r'^[0-9]{1,2}-[0-9]{1,2}-[0-9]{4}$', date_str):
821 date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
822 # Remove AM/PM + timezone
823 date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
825 format_expressions = [
830 '%b %dst %Y %I:%M%p',
831 '%b %dnd %Y %I:%M%p',
832 '%b %dth %Y %I:%M%p',
838 '%Y-%m-%d %H:%M:%S.%f',
841 '%Y-%m-%dT%H:%M:%SZ',
842 '%Y-%m-%dT%H:%M:%S.%fZ',
843 '%Y-%m-%dT%H:%M:%S.%f0Z',
845 '%Y-%m-%dT%H:%M:%S.%f',
849 format_expressions.extend([
857 format_expressions.extend([
864 for expression in format_expressions:
866 upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
869 if upload_date is None:
870 timetuple = email.utils.parsedate_tz(date_str)
872 upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
876 def determine_ext(url, default_ext='unknown_video'):
879 guess = url.partition('?')[0].rpartition('.')[2]
880 if re.match(r'^[A-Za-z0-9]+$', guess):
886 def subtitles_filename(filename, sub_lang, sub_format):
887 return filename.rsplit('.', 1)[0] + '.' + sub_lang + '.' + sub_format
890 def date_from_str(date_str):
892 Return a datetime object from a string in the format YYYYMMDD or
893 (now|today)[+-][0-9](day|week|month|year)(s)?"""
894 today = datetime.date.today()
895 if date_str in ('now', 'today'):
897 if date_str == 'yesterday':
898 return today - datetime.timedelta(days=1)
899 match = re.match('(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
900 if match is not None:
901 sign = match.group('sign')
902 time = int(match.group('time'))
905 unit = match.group('unit')
906 # A bad aproximation?
914 delta = datetime.timedelta(**{unit: time})
916 return datetime.datetime.strptime(date_str, "%Y%m%d").date()
919 def hyphenate_date(date_str):
921 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
922 match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str)
923 if match is not None:
924 return '-'.join(match.groups())
929 class DateRange(object):
930 """Represents a time interval between two dates"""
932 def __init__(self, start=None, end=None):
933 """start and end must be strings in the format accepted by date"""
934 if start is not None:
935 self.start = date_from_str(start)
937 self.start = datetime.datetime.min.date()
939 self.end = date_from_str(end)
941 self.end = datetime.datetime.max.date()
942 if self.start > self.end:
943 raise ValueError('Date range: "%s" , the start date must be before the end date' % self)
947 """Returns a range that only contains the given day"""
950 def __contains__(self, date):
951 """Check if the date is in the range"""
952 if not isinstance(date, datetime.date):
953 date = date_from_str(date)
954 return self.start <= date <= self.end
957 return '%s - %s' % (self.start.isoformat(), self.end.isoformat())
961 """ Returns the platform name as a compat_str """
962 res = platform.platform()
963 if isinstance(res, bytes):
964 res = res.decode(preferredencoding())
966 assert isinstance(res, compat_str)
970 def _windows_write_string(s, out):
971 """ Returns True if the string was written using special methods,
972 False if it has yet to be written out."""
973 # Adapted from http://stackoverflow.com/a/3259271/35070
976 import ctypes.wintypes
984 fileno = out.fileno()
985 except AttributeError:
986 # If the output stream doesn't have a fileno, it's virtual
988 except io.UnsupportedOperation:
989 # Some strange Windows pseudo files?
991 if fileno not in WIN_OUTPUT_IDS:
994 GetStdHandle = ctypes.WINFUNCTYPE(
995 ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
996 (b"GetStdHandle", ctypes.windll.kernel32))
997 h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
999 WriteConsoleW = ctypes.WINFUNCTYPE(
1000 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
1001 ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
1002 ctypes.wintypes.LPVOID)((b"WriteConsoleW", ctypes.windll.kernel32))
1003 written = ctypes.wintypes.DWORD(0)
1005 GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b"GetFileType", ctypes.windll.kernel32))
1006 FILE_TYPE_CHAR = 0x0002
1007 FILE_TYPE_REMOTE = 0x8000
1008 GetConsoleMode = ctypes.WINFUNCTYPE(
1009 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
1010 ctypes.POINTER(ctypes.wintypes.DWORD))(
1011 (b"GetConsoleMode", ctypes.windll.kernel32))
1012 INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
1014 def not_a_console(handle):
1015 if handle == INVALID_HANDLE_VALUE or handle is None:
1017 return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR or
1018 GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0)
1020 if not_a_console(h):
1023 def next_nonbmp_pos(s):
1025 return next(i for i, c in enumerate(s) if ord(c) > 0xffff)
1026 except StopIteration:
1030 count = min(next_nonbmp_pos(s), 1024)
1032 ret = WriteConsoleW(
1033 h, s, count if count else 2, ctypes.byref(written), None)
1035 raise OSError('Failed to write string')
1036 if not count: # We just wrote a non-BMP character
1037 assert written.value == 2
1040 assert written.value > 0
1041 s = s[written.value:]
1045 def write_string(s, out=None, encoding=None):
1048 assert type(s) == compat_str
1050 if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'):
1051 if _windows_write_string(s, out):
1054 if ('b' in getattr(out, 'mode', '') or
1055 sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr
1056 byt = s.encode(encoding or preferredencoding(), 'ignore')
1058 elif hasattr(out, 'buffer'):
1059 enc = encoding or getattr(out, 'encoding', None) or preferredencoding()
1060 byt = s.encode(enc, 'ignore')
1061 out.buffer.write(byt)
1067 def bytes_to_intlist(bs):
1070 if isinstance(bs[0], int): # Python 3
1073 return [ord(c) for c in bs]
1076 def intlist_to_bytes(xs):
1079 return struct_pack('%dB' % len(xs), *xs)
1082 # Cross-platform file locking
1083 if sys.platform == 'win32':
1084 import ctypes.wintypes
1087 class OVERLAPPED(ctypes.Structure):
1089 ('Internal', ctypes.wintypes.LPVOID),
1090 ('InternalHigh', ctypes.wintypes.LPVOID),
1091 ('Offset', ctypes.wintypes.DWORD),
1092 ('OffsetHigh', ctypes.wintypes.DWORD),
1093 ('hEvent', ctypes.wintypes.HANDLE),
1096 kernel32 = ctypes.windll.kernel32
1097 LockFileEx = kernel32.LockFileEx
1098 LockFileEx.argtypes = [
1099 ctypes.wintypes.HANDLE, # hFile
1100 ctypes.wintypes.DWORD, # dwFlags
1101 ctypes.wintypes.DWORD, # dwReserved
1102 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1103 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1104 ctypes.POINTER(OVERLAPPED) # Overlapped
1106 LockFileEx.restype = ctypes.wintypes.BOOL
1107 UnlockFileEx = kernel32.UnlockFileEx
1108 UnlockFileEx.argtypes = [
1109 ctypes.wintypes.HANDLE, # hFile
1110 ctypes.wintypes.DWORD, # dwReserved
1111 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1112 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1113 ctypes.POINTER(OVERLAPPED) # Overlapped
1115 UnlockFileEx.restype = ctypes.wintypes.BOOL
1116 whole_low = 0xffffffff
1117 whole_high = 0x7fffffff
1119 def _lock_file(f, exclusive):
1120 overlapped = OVERLAPPED()
1121 overlapped.Offset = 0
1122 overlapped.OffsetHigh = 0
1123 overlapped.hEvent = 0
1124 f._lock_file_overlapped_p = ctypes.pointer(overlapped)
1125 handle = msvcrt.get_osfhandle(f.fileno())
1126 if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0,
1127 whole_low, whole_high, f._lock_file_overlapped_p):
1128 raise OSError('Locking file failed: %r' % ctypes.FormatError())
1130 def _unlock_file(f):
1131 assert f._lock_file_overlapped_p
1132 handle = msvcrt.get_osfhandle(f.fileno())
1133 if not UnlockFileEx(handle, 0,
1134 whole_low, whole_high, f._lock_file_overlapped_p):
1135 raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
1140 def _lock_file(f, exclusive):
1141 fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
1143 def _unlock_file(f):
1144 fcntl.flock(f, fcntl.LOCK_UN)
1147 class locked_file(object):
1148 def __init__(self, filename, mode, encoding=None):
1149 assert mode in ['r', 'a', 'w']
1150 self.f = io.open(filename, mode, encoding=encoding)
1153 def __enter__(self):
1154 exclusive = self.mode != 'r'
1156 _lock_file(self.f, exclusive)
1162 def __exit__(self, etype, value, traceback):
1164 _unlock_file(self.f)
1171 def write(self, *args):
1172 return self.f.write(*args)
1174 def read(self, *args):
1175 return self.f.read(*args)
1178 def get_filesystem_encoding():
1179 encoding = sys.getfilesystemencoding()
1180 return encoding if encoding is not None else 'utf-8'
1183 def shell_quote(args):
1185 encoding = get_filesystem_encoding()
1187 if isinstance(a, bytes):
1188 # We may get a filename encoded with 'encodeFilename'
1189 a = a.decode(encoding)
1190 quoted_args.append(pipes.quote(a))
1191 return ' '.join(quoted_args)
1194 def smuggle_url(url, data):
1195 """ Pass additional data in a URL for internal use. """
1197 sdata = compat_urllib_parse.urlencode(
1198 {'__youtubedl_smuggle': json.dumps(data)})
1199 return url + '#' + sdata
1202 def unsmuggle_url(smug_url, default=None):
1203 if '#__youtubedl_smuggle' not in smug_url:
1204 return smug_url, default
1205 url, _, sdata = smug_url.rpartition('#')
1206 jsond = compat_parse_qs(sdata)['__youtubedl_smuggle'][0]
1207 data = json.loads(jsond)
1211 def format_bytes(bytes):
1214 if type(bytes) is str:
1215 bytes = float(bytes)
1219 exponent = int(math.log(bytes, 1024.0))
1220 suffix = ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent]
1221 converted = float(bytes) / float(1024 ** exponent)
1222 return '%.2f%s' % (converted, suffix)
1225 def parse_filesize(s):
1229 # The lower-case forms are of course incorrect and inofficial,
1230 # but we support those too
1268 units_re = '|'.join(re.escape(u) for u in _UNIT_TABLE)
1270 r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)' % units_re, s)
1274 num_str = m.group('num').replace(',', '.')
1275 mult = _UNIT_TABLE[m.group('unit')]
1276 return int(float(num_str) * mult)
1279 def month_by_name(name):
1280 """ Return the number of a month by (locale-independently) English name """
1283 return ENGLISH_MONTH_NAMES.index(name) + 1
1288 def month_by_abbreviation(abbrev):
1289 """ Return the number of a month by (locale-independently) English
1293 return [s[:3] for s in ENGLISH_MONTH_NAMES].index(abbrev) + 1
1298 def fix_xml_ampersands(xml_str):
1299 """Replace all the '&' by '&' in XML"""
1301 r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1306 def setproctitle(title):
1307 assert isinstance(title, compat_str)
1309 libc = ctypes.cdll.LoadLibrary("libc.so.6")
1312 title_bytes = title.encode('utf-8')
1313 buf = ctypes.create_string_buffer(len(title_bytes))
1314 buf.value = title_bytes
1316 libc.prctl(15, buf, 0, 0, 0)
1317 except AttributeError:
1318 return # Strange libc, just skip this
1321 def remove_start(s, start):
1322 if s.startswith(start):
1323 return s[len(start):]
1327 def remove_end(s, end):
1329 return s[:-len(end)]
1333 def url_basename(url):
1334 path = compat_urlparse.urlparse(url).path
1335 return path.strip('/').split('/')[-1]
1338 class HEADRequest(compat_urllib_request.Request):
1339 def get_method(self):
1343 def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
1346 v = getattr(v, get_attr, None)
1349 return default if v is None else (int(v) * invscale // scale)
1352 def str_or_none(v, default=None):
1353 return default if v is None else compat_str(v)
1356 def str_to_int(int_str):
1357 """ A more relaxed version of int_or_none """
1360 int_str = re.sub(r'[,\.\+]', '', int_str)
1364 def float_or_none(v, scale=1, invscale=1, default=None):
1365 return default if v is None else (float(v) * invscale / scale)
1368 def parse_duration(s):
1369 if not isinstance(s, compat_basestring):
1377 (?P<only_mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*|
1378 (?P<only_hours>[0-9.]+)\s*(?:hours?)|
1380 \s*(?P<hours_reversed>[0-9]+)\s*(?:[:h]|hours?)\s*(?P<mins_reversed>[0-9]+)\s*(?:[:m]|mins?\.?|minutes?)\s*|
1383 (?:(?P<days>[0-9]+)\s*(?:[:d]|days?)\s*)?
1384 (?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*
1386 (?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*
1388 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?
1393 if m.group('only_mins'):
1394 return float_or_none(m.group('only_mins'), invscale=60)
1395 if m.group('only_hours'):
1396 return float_or_none(m.group('only_hours'), invscale=60 * 60)
1398 res += int(m.group('secs'))
1399 if m.group('mins_reversed'):
1400 res += int(m.group('mins_reversed')) * 60
1402 res += int(m.group('mins')) * 60
1403 if m.group('hours'):
1404 res += int(m.group('hours')) * 60 * 60
1405 if m.group('hours_reversed'):
1406 res += int(m.group('hours_reversed')) * 60 * 60
1408 res += int(m.group('days')) * 24 * 60 * 60
1410 res += float(m.group('ms'))
1414 def prepend_extension(filename, ext, expected_real_ext=None):
1415 name, real_ext = os.path.splitext(filename)
1417 '{0}.{1}{2}'.format(name, ext, real_ext)
1418 if not expected_real_ext or real_ext[1:] == expected_real_ext
1419 else '{0}.{1}'.format(filename, ext))
1422 def replace_extension(filename, ext, expected_real_ext=None):
1423 name, real_ext = os.path.splitext(filename)
1424 return '{0}.{1}'.format(
1425 name if not expected_real_ext or real_ext[1:] == expected_real_ext else filename,
1429 def check_executable(exe, args=[]):
1430 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
1431 args can be a list of arguments for a short output (like -version) """
1433 subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
1439 def get_exe_version(exe, args=['--version'],
1440 version_re=None, unrecognized='present'):
1441 """ Returns the version of the specified executable,
1442 or False if the executable is not present """
1444 out, _ = subprocess.Popen(
1445 [encodeArgument(exe)] + args,
1446 stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
1449 if isinstance(out, bytes): # Python 2.x
1450 out = out.decode('ascii', 'ignore')
1451 return detect_exe_version(out, version_re, unrecognized)
1454 def detect_exe_version(output, version_re=None, unrecognized='present'):
1455 assert isinstance(output, compat_str)
1456 if version_re is None:
1457 version_re = r'version\s+([-0-9._a-zA-Z]+)'
1458 m = re.search(version_re, output)
1465 class PagedList(object):
1467 # This is only useful for tests
1468 return len(self.getslice())
1471 class OnDemandPagedList(PagedList):
1472 def __init__(self, pagefunc, pagesize):
1473 self._pagefunc = pagefunc
1474 self._pagesize = pagesize
1476 def getslice(self, start=0, end=None):
1478 for pagenum in itertools.count(start // self._pagesize):
1479 firstid = pagenum * self._pagesize
1480 nextfirstid = pagenum * self._pagesize + self._pagesize
1481 if start >= nextfirstid:
1484 page_results = list(self._pagefunc(pagenum))
1487 start % self._pagesize
1488 if firstid <= start < nextfirstid
1492 ((end - 1) % self._pagesize) + 1
1493 if (end is not None and firstid <= end <= nextfirstid)
1496 if startv != 0 or endv is not None:
1497 page_results = page_results[startv:endv]
1498 res.extend(page_results)
1500 # A little optimization - if current page is not "full", ie. does
1501 # not contain page_size videos then we can assume that this page
1502 # is the last one - there are no more ids on further pages -
1503 # i.e. no need to query again.
1504 if len(page_results) + startv < self._pagesize:
1507 # If we got the whole page, but the next page is not interesting,
1508 # break out early as well
1509 if end == nextfirstid:
1514 class InAdvancePagedList(PagedList):
1515 def __init__(self, pagefunc, pagecount, pagesize):
1516 self._pagefunc = pagefunc
1517 self._pagecount = pagecount
1518 self._pagesize = pagesize
1520 def getslice(self, start=0, end=None):
1522 start_page = start // self._pagesize
1524 self._pagecount if end is None else (end // self._pagesize + 1))
1525 skip_elems = start - start_page * self._pagesize
1526 only_more = None if end is None else end - start
1527 for pagenum in range(start_page, end_page):
1528 page = list(self._pagefunc(pagenum))
1530 page = page[skip_elems:]
1532 if only_more is not None:
1533 if len(page) < only_more:
1534 only_more -= len(page)
1536 page = page[:only_more]
1543 def uppercase_escape(s):
1544 unicode_escape = codecs.getdecoder('unicode_escape')
1546 r'\\U[0-9a-fA-F]{8}',
1547 lambda m: unicode_escape(m.group(0))[0],
1551 def lowercase_escape(s):
1552 unicode_escape = codecs.getdecoder('unicode_escape')
1554 r'\\u[0-9a-fA-F]{4}',
1555 lambda m: unicode_escape(m.group(0))[0],
1559 def escape_rfc3986(s):
1560 """Escape non-ASCII characters as suggested by RFC 3986"""
1561 if sys.version_info < (3, 0) and isinstance(s, compat_str):
1562 s = s.encode('utf-8')
1563 return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]")
1566 def escape_url(url):
1567 """Escape URL as suggested by RFC 3986"""
1568 url_parsed = compat_urllib_parse_urlparse(url)
1569 return url_parsed._replace(
1570 path=escape_rfc3986(url_parsed.path),
1571 params=escape_rfc3986(url_parsed.params),
1572 query=escape_rfc3986(url_parsed.query),
1573 fragment=escape_rfc3986(url_parsed.fragment)
1577 struct.pack('!I', 0)
1579 # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
1580 def struct_pack(spec, *args):
1581 if isinstance(spec, compat_str):
1582 spec = spec.encode('ascii')
1583 return struct.pack(spec, *args)
1585 def struct_unpack(spec, *args):
1586 if isinstance(spec, compat_str):
1587 spec = spec.encode('ascii')
1588 return struct.unpack(spec, *args)
1590 struct_pack = struct.pack
1591 struct_unpack = struct.unpack
1594 def read_batch_urls(batch_fd):
1596 if not isinstance(url, compat_str):
1597 url = url.decode('utf-8', 'replace')
1598 BOM_UTF8 = '\xef\xbb\xbf'
1599 if url.startswith(BOM_UTF8):
1600 url = url[len(BOM_UTF8):]
1602 if url.startswith(('#', ';', ']')):
1606 with contextlib.closing(batch_fd) as fd:
1607 return [url for url in map(fixup, fd) if url]
1610 def urlencode_postdata(*args, **kargs):
1611 return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
1615 etree_iter = xml.etree.ElementTree.Element.iter
1616 except AttributeError: # Python <=2.6
1617 etree_iter = lambda n: n.findall('.//*')
1621 class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
1622 def doctype(self, name, pubid, system):
1623 pass # Ignore doctypes
1625 parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
1626 kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
1627 tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
1628 # Fix up XML parser in Python 2.x
1629 if sys.version_info < (3, 0):
1630 for n in etree_iter(tree):
1631 if n.text is not None:
1632 if not isinstance(n.text, compat_str):
1633 n.text = n.text.decode('utf-8')
1646 def parse_age_limit(s):
1649 m = re.match(r'^(?P<age>\d{1,2})\+?$', s)
1650 return int(m.group('age')) if m else US_RATINGS.get(s, None)
1653 def strip_jsonp(code):
1655 r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code)
1658 def js_to_json(code):
1661 if v in ('true', 'false', 'null'):
1663 if v.startswith('"'):
1665 if v.startswith("'"):
1667 v = re.sub(r"\\\\|\\'|\"", lambda m: {
1674 res = re.sub(r'''(?x)
1675 "(?:[^"\\]*(?:\\\\|\\['"nu]))*[^"\\]*"|
1676 '(?:[^'\\]*(?:\\\\|\\['"nu]))*[^'\\]*'|
1677 [a-zA-Z_][.a-zA-Z_0-9]*
1679 res = re.sub(r',(\s*[\]}])', lambda m: m.group(1), res)
1683 def qualities(quality_ids):
1684 """ Get a numeric quality value out of a list of possible values """
1687 return quality_ids.index(qid)
1693 DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s'
1696 def limit_length(s, length):
1697 """ Add ellipses to overly long strings """
1702 return s[:length - len(ELLIPSES)] + ELLIPSES
1706 def version_tuple(v):
1707 return tuple(int(e) for e in re.split(r'[-.]', v))
1710 def is_outdated_version(version, limit, assume_new=True):
1712 return not assume_new
1714 return version_tuple(version) < version_tuple(limit)
1716 return not assume_new
1719 def ytdl_is_updateable():
1720 """ Returns if youtube-dl can be updated with -U """
1721 from zipimport import zipimporter
1723 return isinstance(globals().get('__loader__'), zipimporter) or hasattr(sys, 'frozen')
1726 def args_to_str(args):
1727 # Get a short string representation for a subprocess command
1728 return ' '.join(shlex_quote(a) for a in args)
1731 def mimetype2ext(mt):
1732 _, _, res = mt.rpartition('/')
1736 'x-mp4-fragmented': 'mp4',
1741 def urlhandle_detect_ext(url_handle):
1744 getheader = lambda h: url_handle.headers[h]
1745 except AttributeError: # Python < 3
1746 getheader = url_handle.info().getheader
1748 cd = getheader('Content-Disposition')
1750 m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd)
1752 e = determine_ext(m.group('filename'), default_ext=None)
1756 return mimetype2ext(getheader('Content-Type'))
1759 def age_restricted(content_limit, age_limit):
1760 """ Returns True iff the content should be blocked """
1762 if age_limit is None: # No limit set
1764 if content_limit is None:
1765 return False # Content available for everyone
1766 return age_limit < content_limit
1769 def is_html(first_bytes):
1770 """ Detect whether a file contains HTML by examining its first bytes. """
1773 (b'\xef\xbb\xbf', 'utf-8'),
1774 (b'\x00\x00\xfe\xff', 'utf-32-be'),
1775 (b'\xff\xfe\x00\x00', 'utf-32-le'),
1776 (b'\xff\xfe', 'utf-16-le'),
1777 (b'\xfe\xff', 'utf-16-be'),
1779 for bom, enc in BOMS:
1780 if first_bytes.startswith(bom):
1781 s = first_bytes[len(bom):].decode(enc, 'replace')
1784 s = first_bytes.decode('utf-8', 'replace')
1786 return re.match(r'^\s*<', s)
1789 def determine_protocol(info_dict):
1790 protocol = info_dict.get('protocol')
1791 if protocol is not None:
1794 url = info_dict['url']
1795 if url.startswith('rtmp'):
1797 elif url.startswith('mms'):
1799 elif url.startswith('rtsp'):
1802 ext = determine_ext(url)
1808 return compat_urllib_parse_urlparse(url).scheme
1811 def render_table(header_row, data):
1812 """ Render a list of rows, each as a list of values """
1813 table = [header_row] + data
1814 max_lens = [max(len(compat_str(v)) for v in col) for col in zip(*table)]
1815 format_str = ' '.join('%-' + compat_str(ml + 1) + 's' for ml in max_lens[:-1]) + '%s'
1816 return '\n'.join(format_str % tuple(row) for row in table)
1819 def _match_one(filter_part, dct):
1820 COMPARISON_OPERATORS = {
1828 operator_rex = re.compile(r'''(?x)\s*
1830 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1832 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
1833 (?P<strval>(?![0-9.])[a-z0-9A-Z]*)
1836 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
1837 m = operator_rex.search(filter_part)
1839 op = COMPARISON_OPERATORS[m.group('op')]
1840 if m.group('strval') is not None:
1841 if m.group('op') not in ('=', '!='):
1843 'Operator %s does not support string values!' % m.group('op'))
1844 comparison_value = m.group('strval')
1847 comparison_value = int(m.group('intval'))
1849 comparison_value = parse_filesize(m.group('intval'))
1850 if comparison_value is None:
1851 comparison_value = parse_filesize(m.group('intval') + 'B')
1852 if comparison_value is None:
1854 'Invalid integer value %r in filter part %r' % (
1855 m.group('intval'), filter_part))
1856 actual_value = dct.get(m.group('key'))
1857 if actual_value is None:
1858 return m.group('none_inclusive')
1859 return op(actual_value, comparison_value)
1862 '': lambda v: v is not None,
1863 '!': lambda v: v is None,
1865 operator_rex = re.compile(r'''(?x)\s*
1866 (?P<op>%s)\s*(?P<key>[a-z_]+)
1868 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
1869 m = operator_rex.search(filter_part)
1871 op = UNARY_OPERATORS[m.group('op')]
1872 actual_value = dct.get(m.group('key'))
1873 return op(actual_value)
1875 raise ValueError('Invalid filter part %r' % filter_part)
1878 def match_str(filter_str, dct):
1879 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """
1882 _match_one(filter_part, dct) for filter_part in filter_str.split('&'))
1885 def match_filter_func(filter_str):
1886 def _match_func(info_dict):
1887 if match_str(filter_str, info_dict):
1890 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1891 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
1895 def parse_dfxp_time_expr(time_expr):
1899 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
1901 return float(mobj.group('time_offset'))
1903 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:\.\d+)?)$', time_expr)
1905 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3))
1908 def srt_subtitles_timecode(seconds):
1909 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
1912 def dfxp2srt(dfxp_data):
1913 _x = functools.partial(xpath_with_ns, ns_map={
1914 'ttml': 'http://www.w3.org/ns/ttml',
1915 'ttaf1': 'http://www.w3.org/2006/10/ttaf1',
1918 def parse_node(node):
1919 str_or_empty = functools.partial(str_or_none, default='')
1921 out = str_or_empty(node.text)
1924 if child.tag in (_x('ttml:br'), _x('ttaf1:br'), 'br'):
1925 out += '\n' + str_or_empty(child.tail)
1926 elif child.tag in (_x('ttml:span'), _x('ttaf1:span'), 'span'):
1927 out += str_or_empty(parse_node(child))
1929 out += str_or_empty(xml.etree.ElementTree.tostring(child))
1933 dfxp = xml.etree.ElementTree.fromstring(dfxp_data.encode('utf-8'))
1935 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall('.//p')
1938 raise ValueError('Invalid dfxp/TTML subtitle')
1940 for para, index in zip(paras, itertools.count(1)):
1941 begin_time = parse_dfxp_time_expr(para.attrib['begin'])
1942 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
1944 end_time = begin_time + parse_dfxp_time_expr(para.attrib['dur'])
1945 out.append('%d\n%s --> %s\n%s\n\n' % (
1947 srt_subtitles_timecode(begin_time),
1948 srt_subtitles_timecode(end_time),
1954 def cli_option(params, command_option, param):
1955 param = params.get(param)
1956 return [command_option, param] if param is not None else []
1959 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
1960 param = params.get(param)
1961 assert isinstance(param, bool)
1963 return [command_option + separator + (true_value if param else false_value)]
1964 return [command_option, true_value if param else false_value]
1967 def cli_valueless_option(params, command_option, param, expected_value=True):
1968 param = params.get(param)
1969 return [command_option] if param == expected_value else []
1972 def cli_configuration_args(params, param, default=[]):
1973 ex_args = params.get(param)
1976 assert isinstance(ex_args, list)
1980 class ISO639Utils(object):
1981 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
2170 def short2long(cls, code):
2171 """Convert language code from ISO 639-1 to ISO 639-2/T"""
2172 return cls._lang_map.get(code[:2])
2175 def long2short(cls, code):
2176 """Convert language code from ISO 639-2/T to ISO 639-1"""
2177 for short_name, long_name in cls._lang_map.items():
2178 if long_name == code:
2182 class ISO3166Utils(object):
2183 # From http://data.okfn.org/data/core/country-list
2185 'AF': 'Afghanistan',
2186 'AX': 'Ã…land Islands',
2189 'AS': 'American Samoa',
2194 'AG': 'Antigua and Barbuda',
2211 'BO': 'Bolivia, Plurinational State of',
2212 'BQ': 'Bonaire, Sint Eustatius and Saba',
2213 'BA': 'Bosnia and Herzegovina',
2215 'BV': 'Bouvet Island',
2217 'IO': 'British Indian Ocean Territory',
2218 'BN': 'Brunei Darussalam',
2220 'BF': 'Burkina Faso',
2226 'KY': 'Cayman Islands',
2227 'CF': 'Central African Republic',
2231 'CX': 'Christmas Island',
2232 'CC': 'Cocos (Keeling) Islands',
2236 'CD': 'Congo, the Democratic Republic of the',
2237 'CK': 'Cook Islands',
2239 'CI': 'Côte d\'Ivoire',
2244 'CZ': 'Czech Republic',
2248 'DO': 'Dominican Republic',
2251 'SV': 'El Salvador',
2252 'GQ': 'Equatorial Guinea',
2256 'FK': 'Falkland Islands (Malvinas)',
2257 'FO': 'Faroe Islands',
2261 'GF': 'French Guiana',
2262 'PF': 'French Polynesia',
2263 'TF': 'French Southern Territories',
2278 'GW': 'Guinea-Bissau',
2281 'HM': 'Heard Island and McDonald Islands',
2282 'VA': 'Holy See (Vatican City State)',
2289 'IR': 'Iran, Islamic Republic of',
2292 'IM': 'Isle of Man',
2302 'KP': 'Korea, Democratic People\'s Republic of',
2303 'KR': 'Korea, Republic of',
2306 'LA': 'Lao People\'s Democratic Republic',
2312 'LI': 'Liechtenstein',
2316 'MK': 'Macedonia, the Former Yugoslav Republic of',
2323 'MH': 'Marshall Islands',
2329 'FM': 'Micronesia, Federated States of',
2330 'MD': 'Moldova, Republic of',
2341 'NL': 'Netherlands',
2342 'NC': 'New Caledonia',
2343 'NZ': 'New Zealand',
2348 'NF': 'Norfolk Island',
2349 'MP': 'Northern Mariana Islands',
2354 'PS': 'Palestine, State of',
2356 'PG': 'Papua New Guinea',
2359 'PH': 'Philippines',
2363 'PR': 'Puerto Rico',
2367 'RU': 'Russian Federation',
2369 'BL': 'Saint Barthélemy',
2370 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
2371 'KN': 'Saint Kitts and Nevis',
2372 'LC': 'Saint Lucia',
2373 'MF': 'Saint Martin (French part)',
2374 'PM': 'Saint Pierre and Miquelon',
2375 'VC': 'Saint Vincent and the Grenadines',
2378 'ST': 'Sao Tome and Principe',
2379 'SA': 'Saudi Arabia',
2383 'SL': 'Sierra Leone',
2385 'SX': 'Sint Maarten (Dutch part)',
2388 'SB': 'Solomon Islands',
2390 'ZA': 'South Africa',
2391 'GS': 'South Georgia and the South Sandwich Islands',
2392 'SS': 'South Sudan',
2397 'SJ': 'Svalbard and Jan Mayen',
2400 'CH': 'Switzerland',
2401 'SY': 'Syrian Arab Republic',
2402 'TW': 'Taiwan, Province of China',
2404 'TZ': 'Tanzania, United Republic of',
2406 'TL': 'Timor-Leste',
2410 'TT': 'Trinidad and Tobago',
2413 'TM': 'Turkmenistan',
2414 'TC': 'Turks and Caicos Islands',
2418 'AE': 'United Arab Emirates',
2419 'GB': 'United Kingdom',
2420 'US': 'United States',
2421 'UM': 'United States Minor Outlying Islands',
2425 'VE': 'Venezuela, Bolivarian Republic of',
2427 'VG': 'Virgin Islands, British',
2428 'VI': 'Virgin Islands, U.S.',
2429 'WF': 'Wallis and Futuna',
2430 'EH': 'Western Sahara',
2437 def short2full(cls, code):
2438 """Convert an ISO 3166-2 country code to the corresponding full name"""
2439 return cls._country_map.get(code.upper())
2442 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
2443 def __init__(self, proxies=None):
2444 # Set default handlers
2445 for type in ('http', 'https'):
2446 setattr(self, '%s_open' % type,
2447 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
2448 meth(r, proxy, type))
2449 return compat_urllib_request.ProxyHandler.__init__(self, proxies)
2451 def proxy_open(self, req, proxy, type):
2452 req_proxy = req.headers.get('Ytdl-request-proxy')
2453 if req_proxy is not None:
2455 del req.headers['Ytdl-request-proxy']
2457 if proxy == '__noproxy__':
2458 return None # No Proxy
2459 return compat_urllib_request.ProxyHandler.proxy_open(
2460 self, req, proxy, type)