2 # -*- coding: utf-8 -*-
28 import xml.etree.ElementTree
32 import urllib.request as compat_urllib_request
33 except ImportError: # Python 2
34 import urllib2 as compat_urllib_request
37 import urllib.error as compat_urllib_error
38 except ImportError: # Python 2
39 import urllib2 as compat_urllib_error
42 import urllib.parse as compat_urllib_parse
43 except ImportError: # Python 2
44 import urllib as compat_urllib_parse
47 from urllib.parse import urlparse as compat_urllib_parse_urlparse
48 except ImportError: # Python 2
49 from urlparse import urlparse as compat_urllib_parse_urlparse
52 import urllib.parse as compat_urlparse
53 except ImportError: # Python 2
54 import urlparse as compat_urlparse
57 import http.cookiejar as compat_cookiejar
58 except ImportError: # Python 2
59 import cookielib as compat_cookiejar
62 import html.entities as compat_html_entities
63 except ImportError: # Python 2
64 import htmlentitydefs as compat_html_entities
67 import html.parser as compat_html_parser
68 except ImportError: # Python 2
69 import HTMLParser as compat_html_parser
72 import http.client as compat_http_client
73 except ImportError: # Python 2
74 import httplib as compat_http_client
77 from urllib.error import HTTPError as compat_HTTPError
78 except ImportError: # Python 2
79 from urllib2 import HTTPError as compat_HTTPError
82 from urllib.request import urlretrieve as compat_urlretrieve
83 except ImportError: # Python 2
84 from urllib import urlretrieve as compat_urlretrieve
88 from subprocess import DEVNULL
89 compat_subprocess_get_DEVNULL = lambda: DEVNULL
91 compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w')
94 from urllib.parse import parse_qs as compat_parse_qs
95 except ImportError: # Python 2
96 # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib.
97 # Python 2's version is apparently totally broken
98 def _unquote(string, encoding='utf-8', errors='replace'):
101 res = string.split('%')
108 # pct_sequence: contiguous sequence of percent-encoded bytes, decoded
115 pct_sequence += item[:2].decode('hex')
118 # This segment was just a single percent-encoded character.
119 # May be part of a sequence of code units, so delay decoding.
120 # (Stored in pct_sequence).
124 # Encountered non-percent-encoded characters. Flush the current
126 string += pct_sequence.decode(encoding, errors) + rest
129 # Flush the final pct_sequence
130 string += pct_sequence.decode(encoding, errors)
133 def _parse_qsl(qs, keep_blank_values=False, strict_parsing=False,
134 encoding='utf-8', errors='replace'):
135 qs, _coerce_result = qs, unicode
136 pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]
138 for name_value in pairs:
139 if not name_value and not strict_parsing:
141 nv = name_value.split('=', 1)
144 raise ValueError("bad query field: %r" % (name_value,))
145 # Handle case of a control-name with no equal sign
146 if keep_blank_values:
150 if len(nv[1]) or keep_blank_values:
151 name = nv[0].replace('+', ' ')
152 name = _unquote(name, encoding=encoding, errors=errors)
153 name = _coerce_result(name)
154 value = nv[1].replace('+', ' ')
155 value = _unquote(value, encoding=encoding, errors=errors)
156 value = _coerce_result(value)
157 r.append((name, value))
160 def compat_parse_qs(qs, keep_blank_values=False, strict_parsing=False,
161 encoding='utf-8', errors='replace'):
163 pairs = _parse_qsl(qs, keep_blank_values, strict_parsing,
164 encoding=encoding, errors=errors)
165 for name, value in pairs:
166 if name in parsed_result:
167 parsed_result[name].append(value)
169 parsed_result[name] = [value]
173 compat_str = unicode # Python 2
178 compat_chr = unichr # Python 2
183 from xml.etree.ElementTree import ParseError as compat_xml_parse_error
184 except ImportError: # Python 2.6
185 from xml.parsers.expat import ExpatError as compat_xml_parse_error
188 if type(c) is int: return c
191 # This is not clearly defined otherwise
192 compiled_regex_type = type(re.compile(''))
195 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20100101 Firefox/10.0 (Chrome)',
196 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
197 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
198 'Accept-Encoding': 'gzip, deflate',
199 'Accept-Language': 'en-us,en;q=0.5',
202 def preferredencoding():
203 """Get preferred encoding.
205 Returns the best encoding scheme for the system, based on
206 locale.getpreferredencoding() and some further tweaks.
209 pref = locale.getpreferredencoding()
216 if sys.version_info < (3,0):
218 print(s.encode(preferredencoding(), 'xmlcharrefreplace'))
221 assert type(s) == type(u'')
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):
227 def write_json_file(obj, fn):
228 with open(fn, 'wb') as f:
231 def write_json_file(obj, fn):
232 with open(fn, 'w', encoding='utf-8') as f:
235 if sys.version_info >= (2,7):
236 def find_xpath_attr(node, xpath, key, val):
237 """ Find the xpath xpath[@key=val] """
238 assert re.match(r'^[a-zA-Z]+$', key)
239 assert re.match(r'^[a-zA-Z0-9@\s:._]*$', val)
240 expr = xpath + u"[@%s='%s']" % (key, val)
241 return node.find(expr)
243 def find_xpath_attr(node, xpath, key, val):
244 for f in node.findall(xpath):
245 if f.attrib.get(key) == val:
249 # On python2.6 the xml.etree.ElementTree.Element methods don't support
250 # the namespace parameter
251 def xpath_with_ns(path, ns_map):
252 components = [c.split(':') for c in path.split('/')]
256 replaced.append(c[0])
259 replaced.append('{%s}%s' % (ns_map[ns], tag))
260 return '/'.join(replaced)
262 def htmlentity_transform(matchobj):
263 """Transforms an HTML entity to a character.
265 This function receives a match object and is intended to be used with
266 the re.sub() function.
268 entity = matchobj.group(1)
270 # Known non-numeric HTML entity
271 if entity in compat_html_entities.name2codepoint:
272 return compat_chr(compat_html_entities.name2codepoint[entity])
274 mobj = re.match(u'(?u)#(x?\\d+)', entity)
276 numstr = mobj.group(1)
277 if numstr.startswith(u'x'):
279 numstr = u'0%s' % numstr
282 return compat_chr(int(numstr, base))
284 # Unknown entity in name, return its literal representation
285 return (u'&%s;' % entity)
287 compat_html_parser.locatestarttagend = re.compile(r"""<[a-zA-Z][-.a-zA-Z0-9:_]*(?:\s+(?:(?<=['"\s])[^\s/>][^\s/=>]*(?:\s*=+\s*(?:'[^']*'|"[^"]*"|(?!['"])[^>\s]*))?\s*)*)?\s*""", re.VERBOSE) # backport bugfix
288 class BaseHTMLParser(compat_html_parser.HTMLParser):
290 compat_html_parser.HTMLParser.__init__(self)
293 def loads(self, html):
298 class AttrParser(BaseHTMLParser):
299 """Modified HTMLParser that isolates a tag with the specified attribute"""
300 def __init__(self, attribute, value):
301 self.attribute = attribute
306 self.watch_startpos = False
308 BaseHTMLParser.__init__(self)
310 def error(self, message):
311 if self.error_count > 10 or self.started:
312 raise compat_html_parser.HTMLParseError(message, self.getpos())
313 self.rawdata = '\n'.join(self.html.split('\n')[self.getpos()[0]:]) # skip one line
314 self.error_count += 1
317 def handle_starttag(self, tag, attrs):
320 self.find_startpos(None)
321 if self.attribute in attrs and attrs[self.attribute] == self.value:
324 self.watch_startpos = True
326 if not tag in self.depth: self.depth[tag] = 0
329 def handle_endtag(self, tag):
331 if tag in self.depth: self.depth[tag] -= 1
332 if self.depth[self.result[0]] == 0:
334 self.result.append(self.getpos())
336 def find_startpos(self, x):
337 """Needed to put the start position of the result (self.result[1])
338 after the opening tag with the requested id"""
339 if self.watch_startpos:
340 self.watch_startpos = False
341 self.result.append(self.getpos())
342 handle_entityref = handle_charref = handle_data = handle_comment = \
343 handle_decl = handle_pi = unknown_decl = find_startpos
345 def get_result(self):
346 if self.result is None:
348 if len(self.result) != 3:
350 lines = self.html.split('\n')
351 lines = lines[self.result[1][0]-1:self.result[2][0]]
352 lines[0] = lines[0][self.result[1][1]:]
354 lines[-1] = lines[-1][:self.result[2][1]-self.result[1][1]]
355 lines[-1] = lines[-1][:self.result[2][1]]
356 return '\n'.join(lines).strip()
357 # Hack for https://github.com/rg3/youtube-dl/issues/662
358 if sys.version_info < (2, 7, 3):
359 AttrParser.parse_endtag = (lambda self, i:
360 i + len("</scr'+'ipt>")
361 if self.rawdata[i:].startswith("</scr'+'ipt>")
362 else compat_html_parser.HTMLParser.parse_endtag(self, i))
364 def get_element_by_id(id, html):
365 """Return the content of the tag with the specified ID in the passed HTML document"""
366 return get_element_by_attribute("id", id, html)
368 def get_element_by_attribute(attribute, value, html):
369 """Return the content of the tag with the specified attribute in the passed HTML document"""
370 parser = AttrParser(attribute, value)
373 except compat_html_parser.HTMLParseError:
375 return parser.get_result()
377 class MetaParser(BaseHTMLParser):
379 Modified HTMLParser that isolates a meta tag with the specified name
382 def __init__(self, name):
383 BaseHTMLParser.__init__(self)
388 def handle_starttag(self, tag, attrs):
392 if attrs.get('name') == self.name:
393 self.result = attrs.get('content')
395 def get_result(self):
398 def get_meta_content(name, html):
400 Return the content attribute from the meta tag with the given name attribute.
402 parser = MetaParser(name)
405 except compat_html_parser.HTMLParseError:
407 return parser.get_result()
410 def clean_html(html):
411 """Clean an HTML snippet into a readable string"""
413 html = html.replace('\n', ' ')
414 html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html)
415 html = re.sub(r'<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
417 html = re.sub('<.*?>', '', html)
418 # Replace html entities
419 html = unescapeHTML(html)
423 def sanitize_open(filename, open_mode):
424 """Try to open the given filename, and slightly tweak it if this fails.
426 Attempts to open the given filename. If this fails, it tries to change
427 the filename slightly, step by step, until it's either able to open it
428 or it fails and raises a final exception, like the standard open()
431 It returns the tuple (stream, definitive_file_name).
435 if sys.platform == 'win32':
437 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
438 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
439 stream = open(encodeFilename(filename), open_mode)
440 return (stream, filename)
441 except (IOError, OSError) as err:
442 if err.errno in (errno.EACCES,):
445 # In case of error, try to remove win32 forbidden chars
446 alt_filename = os.path.join(
447 re.sub(u'[/<>:"\\|\\\\?\\*]', u'#', path_part)
448 for path_part in os.path.split(filename)
450 if alt_filename == filename:
453 # An exception here should be caught in the caller
454 stream = open(encodeFilename(filename), open_mode)
455 return (stream, alt_filename)
458 def timeconvert(timestr):
459 """Convert RFC 2822 defined time string into system timestamp"""
461 timetuple = email.utils.parsedate_tz(timestr)
462 if timetuple is not None:
463 timestamp = email.utils.mktime_tz(timetuple)
466 def sanitize_filename(s, restricted=False, is_id=False):
467 """Sanitizes a string so it could be used as part of a filename.
468 If restricted is set, use a stricter subset of allowed characters.
469 Set is_id if this is not an arbitrary string, but an ID that should be kept if possible
471 def replace_insane(char):
472 if char == '?' or ord(char) < 32 or ord(char) == 127:
475 return '' if restricted else '\''
477 return '_-' if restricted else ' -'
478 elif char in '\\/|*<>':
480 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
482 if restricted and ord(char) > 127:
486 result = u''.join(map(replace_insane, s))
488 while '__' in result:
489 result = result.replace('__', '_')
490 result = result.strip('_')
491 # Common case of "Foreign band name - English song title"
492 if restricted and result.startswith('-_'):
498 def orderedSet(iterable):
499 """ Remove all duplicates from the input iterable """
510 assert type(s) == compat_str
512 result = re.sub(r'(?u)&(.+?);', htmlentity_transform, s)
516 def encodeFilename(s, for_subprocess=False):
518 @param s The name of the file
521 assert type(s) == compat_str
523 # Python 3 has a Unicode API
524 if sys.version_info >= (3, 0):
527 if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
528 # Pass u'' directly to use Unicode APIs on Windows 2000 and up
529 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
530 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
531 if not for_subprocess:
534 # For subprocess calls, encode with locale encoding
535 # Refer to http://stackoverflow.com/a/9951851/35070
536 encoding = preferredencoding()
538 encoding = sys.getfilesystemencoding()
541 return s.encode(encoding, 'ignore')
544 def encodeArgument(s):
545 if not isinstance(s, compat_str):
546 # Legacy code that uses byte strings
547 # Uncomment the following line after fixing all post processors
548 #assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
549 s = s.decode('ascii')
550 return encodeFilename(s, True)
553 def decodeOption(optval):
556 if isinstance(optval, bytes):
557 optval = optval.decode(preferredencoding())
559 assert isinstance(optval, compat_str)
562 def formatSeconds(secs):
564 return '%d:%02d:%02d' % (secs // 3600, (secs % 3600) // 60, secs % 60)
566 return '%d:%02d' % (secs // 60, secs % 60)
571 def make_HTTPS_handler(opts_no_check_certificate, **kwargs):
572 if sys.version_info < (3, 2):
575 class HTTPSConnectionV3(httplib.HTTPSConnection):
576 def __init__(self, *args, **kwargs):
577 httplib.HTTPSConnection.__init__(self, *args, **kwargs)
580 sock = socket.create_connection((self.host, self.port), self.timeout)
581 if getattr(self, '_tunnel_host', False):
585 self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_SSLv3)
587 self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_SSLv23)
589 class HTTPSHandlerV3(compat_urllib_request.HTTPSHandler):
590 def https_open(self, req):
591 return self.do_open(HTTPSConnectionV3, req)
592 return HTTPSHandlerV3(**kwargs)
594 context = ssl.SSLContext(ssl.PROTOCOL_SSLv3)
595 context.verify_mode = (ssl.CERT_NONE
596 if opts_no_check_certificate
597 else ssl.CERT_REQUIRED)
598 context.set_default_verify_paths()
600 context.load_default_certs()
601 except AttributeError:
603 return compat_urllib_request.HTTPSHandler(context=context, **kwargs)
605 class ExtractorError(Exception):
606 """Error during info extraction."""
607 def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):
608 """ tb, if given, is the original traceback (so that it can be printed out).
609 If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
612 if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError):
614 if video_id is not None:
615 msg = video_id + ': ' + msg
617 msg = msg + u'; please report this issue on https://yt-dl.org/bug . Be sure to call youtube-dl with the --verbose flag and include its complete output. Make sure you are using the latest version; type youtube-dl -U to update.'
618 super(ExtractorError, self).__init__(msg)
621 self.exc_info = sys.exc_info() # preserve original exception
623 self.video_id = video_id
625 def format_traceback(self):
626 if self.traceback is None:
628 return u''.join(traceback.format_tb(self.traceback))
631 class RegexNotFoundError(ExtractorError):
632 """Error when a regex didn't match"""
636 class DownloadError(Exception):
637 """Download Error exception.
639 This exception may be thrown by FileDownloader objects if they are not
640 configured to continue on errors. They will contain the appropriate
643 def __init__(self, msg, exc_info=None):
644 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
645 super(DownloadError, self).__init__(msg)
646 self.exc_info = exc_info
649 class SameFileError(Exception):
650 """Same File exception.
652 This exception will be thrown by FileDownloader objects if they detect
653 multiple files would have to be downloaded to the same file on disk.
658 class PostProcessingError(Exception):
659 """Post Processing exception.
661 This exception may be raised by PostProcessor's .run() method to
662 indicate an error in the postprocessing task.
664 def __init__(self, msg):
667 class MaxDownloadsReached(Exception):
668 """ --max-downloads limit has been reached. """
672 class UnavailableVideoError(Exception):
673 """Unavailable Format exception.
675 This exception will be thrown when a video is requested
676 in a format that is not available for that video.
681 class ContentTooShortError(Exception):
682 """Content Too Short exception.
684 This exception may be raised by FileDownloader objects when a file they
685 download is too small for what the server announced first, indicating
686 the connection was probably interrupted.
692 def __init__(self, downloaded, expected):
693 self.downloaded = downloaded
694 self.expected = expected
696 class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
697 """Handler for HTTP requests and responses.
699 This class, when installed with an OpenerDirector, automatically adds
700 the standard headers to every HTTP request and handles gzipped and
701 deflated responses from web servers. If compression is to be avoided in
702 a particular request, the original request in the program code only has
703 to include the HTTP header "Youtubedl-No-Compression", which will be
704 removed before making the real request.
706 Part of this code was copied from:
708 http://techknack.net/python-urllib2-handlers/
710 Andrew Rowls, the author of that code, agreed to release it to the
717 return zlib.decompress(data, -zlib.MAX_WBITS)
719 return zlib.decompress(data)
722 def addinfourl_wrapper(stream, headers, url, code):
723 if hasattr(compat_urllib_request.addinfourl, 'getcode'):
724 return compat_urllib_request.addinfourl(stream, headers, url, code)
725 ret = compat_urllib_request.addinfourl(stream, headers, url)
729 def http_request(self, req):
730 for h,v in std_headers.items():
734 if 'Youtubedl-no-compression' in req.headers:
735 if 'Accept-encoding' in req.headers:
736 del req.headers['Accept-encoding']
737 del req.headers['Youtubedl-no-compression']
738 if 'Youtubedl-user-agent' in req.headers:
739 if 'User-agent' in req.headers:
740 del req.headers['User-agent']
741 req.headers['User-agent'] = req.headers['Youtubedl-user-agent']
742 del req.headers['Youtubedl-user-agent']
745 def http_response(self, req, resp):
748 if resp.headers.get('Content-encoding', '') == 'gzip':
749 content = resp.read()
750 gz = gzip.GzipFile(fileobj=io.BytesIO(content), mode='rb')
752 uncompressed = io.BytesIO(gz.read())
753 except IOError as original_ioerror:
754 # There may be junk add the end of the file
755 # See http://stackoverflow.com/q/4928560/35070 for details
756 for i in range(1, 1024):
758 gz = gzip.GzipFile(fileobj=io.BytesIO(content[:-i]), mode='rb')
759 uncompressed = io.BytesIO(gz.read())
764 raise original_ioerror
765 resp = self.addinfourl_wrapper(uncompressed, old_resp.headers, old_resp.url, old_resp.code)
766 resp.msg = old_resp.msg
768 if resp.headers.get('Content-encoding', '') == 'deflate':
769 gz = io.BytesIO(self.deflate(resp.read()))
770 resp = self.addinfourl_wrapper(gz, old_resp.headers, old_resp.url, old_resp.code)
771 resp.msg = old_resp.msg
774 https_request = http_request
775 https_response = http_response
778 def parse_iso8601(date_str):
779 """ Return a UNIX timestamp from the given date """
785 r'Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$',
788 timezone = datetime.timedelta()
790 date_str = date_str[:-len(m.group(0))]
791 if not m.group('sign'):
792 timezone = datetime.timedelta()
794 sign = 1 if m.group('sign') == '+' else -1
795 timezone = datetime.timedelta(
796 hours=sign * int(m.group('hours')),
797 minutes=sign * int(m.group('minutes')))
799 dt = datetime.datetime.strptime(date_str, '%Y-%m-%dT%H:%M:%S') - timezone
800 return calendar.timegm(dt.timetuple())
803 def unified_strdate(date_str):
804 """Return a string with the date in the format YYYYMMDD"""
811 date_str = date_str.replace(',', ' ')
812 # %z (UTC offset) is only supported in python>=3.2
813 date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
814 format_expressions = [
826 '%Y-%m-%dT%H:%M:%SZ',
827 '%Y-%m-%dT%H:%M:%S.%fZ',
828 '%Y-%m-%dT%H:%M:%S.%f0Z',
830 '%Y-%m-%dT%H:%M:%S.%f',
833 for expression in format_expressions:
835 upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
838 if upload_date is None:
839 timetuple = email.utils.parsedate_tz(date_str)
841 upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
844 def determine_ext(url, default_ext=u'unknown_video'):
845 guess = url.partition(u'?')[0].rpartition(u'.')[2]
846 if re.match(r'^[A-Za-z0-9]+$', guess):
851 def subtitles_filename(filename, sub_lang, sub_format):
852 return filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
854 def date_from_str(date_str):
856 Return a datetime object from a string in the format YYYYMMDD or
857 (now|today)[+-][0-9](day|week|month|year)(s)?"""
858 today = datetime.date.today()
859 if date_str == 'now'or date_str == 'today':
861 match = re.match('(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
862 if match is not None:
863 sign = match.group('sign')
864 time = int(match.group('time'))
867 unit = match.group('unit')
876 delta = datetime.timedelta(**{unit: time})
878 return datetime.datetime.strptime(date_str, "%Y%m%d").date()
880 def hyphenate_date(date_str):
882 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
883 match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str)
884 if match is not None:
885 return '-'.join(match.groups())
889 class DateRange(object):
890 """Represents a time interval between two dates"""
891 def __init__(self, start=None, end=None):
892 """start and end must be strings in the format accepted by date"""
893 if start is not None:
894 self.start = date_from_str(start)
896 self.start = datetime.datetime.min.date()
898 self.end = date_from_str(end)
900 self.end = datetime.datetime.max.date()
901 if self.start > self.end:
902 raise ValueError('Date range: "%s" , the start date must be before the end date' % self)
905 """Returns a range that only contains the given day"""
907 def __contains__(self, date):
908 """Check if the date is in the range"""
909 if not isinstance(date, datetime.date):
910 date = date_from_str(date)
911 return self.start <= date <= self.end
913 return '%s - %s' % ( self.start.isoformat(), self.end.isoformat())
917 """ Returns the platform name as a compat_str """
918 res = platform.platform()
919 if isinstance(res, bytes):
920 res = res.decode(preferredencoding())
922 assert isinstance(res, compat_str)
926 def _windows_write_string(s, out):
927 """ Returns True if the string was written using special methods,
928 False if it has yet to be written out."""
929 # Adapted from http://stackoverflow.com/a/3259271/35070
932 import ctypes.wintypes
940 fileno = out.fileno()
941 except AttributeError:
942 # If the output stream doesn't have a fileno, it's virtual
944 if fileno not in WIN_OUTPUT_IDS:
947 GetStdHandle = ctypes.WINFUNCTYPE(
948 ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
949 ("GetStdHandle", ctypes.windll.kernel32))
950 h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
952 WriteConsoleW = ctypes.WINFUNCTYPE(
953 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
954 ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
955 ctypes.wintypes.LPVOID)(("WriteConsoleW", ctypes.windll.kernel32))
956 written = ctypes.wintypes.DWORD(0)
958 GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)(("GetFileType", ctypes.windll.kernel32))
959 FILE_TYPE_CHAR = 0x0002
960 FILE_TYPE_REMOTE = 0x8000
961 GetConsoleMode = ctypes.WINFUNCTYPE(
962 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
963 ctypes.POINTER(ctypes.wintypes.DWORD))(
964 ("GetConsoleMode", ctypes.windll.kernel32))
965 INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
967 def not_a_console(handle):
968 if handle == INVALID_HANDLE_VALUE or handle is None:
970 return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR
971 or GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0)
976 def next_nonbmp_pos(s):
978 return next(i for i, c in enumerate(s) if ord(c) > 0xffff)
979 except StopIteration:
983 count = min(next_nonbmp_pos(s), 1024)
986 h, s, count if count else 2, ctypes.byref(written), None)
988 raise OSError('Failed to write string')
989 if not count: # We just wrote a non-BMP character
990 assert written.value == 2
993 assert written.value > 0
994 s = s[written.value:]
998 def write_string(s, out=None, encoding=None):
1001 assert type(s) == compat_str
1003 if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'):
1004 if _windows_write_string(s, out):
1007 if ('b' in getattr(out, 'mode', '') or
1008 sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr
1009 byt = s.encode(encoding or preferredencoding(), 'ignore')
1011 elif hasattr(out, 'buffer'):
1012 enc = encoding or getattr(out, 'encoding', None) or preferredencoding()
1013 byt = s.encode(enc, 'ignore')
1014 out.buffer.write(byt)
1020 def bytes_to_intlist(bs):
1023 if isinstance(bs[0], int): # Python 3
1026 return [ord(c) for c in bs]
1029 def intlist_to_bytes(xs):
1032 if isinstance(chr(0), bytes): # Python 2
1033 return ''.join([chr(x) for x in xs])
1038 def get_cachedir(params={}):
1039 cache_root = os.environ.get('XDG_CACHE_HOME',
1040 os.path.expanduser('~/.cache'))
1041 return params.get('cachedir', os.path.join(cache_root, 'youtube-dl'))
1044 # Cross-platform file locking
1045 if sys.platform == 'win32':
1046 import ctypes.wintypes
1049 class OVERLAPPED(ctypes.Structure):
1051 ('Internal', ctypes.wintypes.LPVOID),
1052 ('InternalHigh', ctypes.wintypes.LPVOID),
1053 ('Offset', ctypes.wintypes.DWORD),
1054 ('OffsetHigh', ctypes.wintypes.DWORD),
1055 ('hEvent', ctypes.wintypes.HANDLE),
1058 kernel32 = ctypes.windll.kernel32
1059 LockFileEx = kernel32.LockFileEx
1060 LockFileEx.argtypes = [
1061 ctypes.wintypes.HANDLE, # hFile
1062 ctypes.wintypes.DWORD, # dwFlags
1063 ctypes.wintypes.DWORD, # dwReserved
1064 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1065 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1066 ctypes.POINTER(OVERLAPPED) # Overlapped
1068 LockFileEx.restype = ctypes.wintypes.BOOL
1069 UnlockFileEx = kernel32.UnlockFileEx
1070 UnlockFileEx.argtypes = [
1071 ctypes.wintypes.HANDLE, # hFile
1072 ctypes.wintypes.DWORD, # dwReserved
1073 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1074 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1075 ctypes.POINTER(OVERLAPPED) # Overlapped
1077 UnlockFileEx.restype = ctypes.wintypes.BOOL
1078 whole_low = 0xffffffff
1079 whole_high = 0x7fffffff
1081 def _lock_file(f, exclusive):
1082 overlapped = OVERLAPPED()
1083 overlapped.Offset = 0
1084 overlapped.OffsetHigh = 0
1085 overlapped.hEvent = 0
1086 f._lock_file_overlapped_p = ctypes.pointer(overlapped)
1087 handle = msvcrt.get_osfhandle(f.fileno())
1088 if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0,
1089 whole_low, whole_high, f._lock_file_overlapped_p):
1090 raise OSError('Locking file failed: %r' % ctypes.FormatError())
1092 def _unlock_file(f):
1093 assert f._lock_file_overlapped_p
1094 handle = msvcrt.get_osfhandle(f.fileno())
1095 if not UnlockFileEx(handle, 0,
1096 whole_low, whole_high, f._lock_file_overlapped_p):
1097 raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
1102 def _lock_file(f, exclusive):
1103 fcntl.lockf(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
1105 def _unlock_file(f):
1106 fcntl.lockf(f, fcntl.LOCK_UN)
1109 class locked_file(object):
1110 def __init__(self, filename, mode, encoding=None):
1111 assert mode in ['r', 'a', 'w']
1112 self.f = io.open(filename, mode, encoding=encoding)
1115 def __enter__(self):
1116 exclusive = self.mode != 'r'
1118 _lock_file(self.f, exclusive)
1124 def __exit__(self, etype, value, traceback):
1126 _unlock_file(self.f)
1133 def write(self, *args):
1134 return self.f.write(*args)
1136 def read(self, *args):
1137 return self.f.read(*args)
1140 def shell_quote(args):
1142 encoding = sys.getfilesystemencoding()
1143 if encoding is None:
1146 if isinstance(a, bytes):
1147 # We may get a filename encoded with 'encodeFilename'
1148 a = a.decode(encoding)
1149 quoted_args.append(pipes.quote(a))
1150 return u' '.join(quoted_args)
1153 def takewhile_inclusive(pred, seq):
1154 """ Like itertools.takewhile, but include the latest evaluated element
1155 (the first element so that Not pred(e)) """
1162 def smuggle_url(url, data):
1163 """ Pass additional data in a URL for internal use. """
1165 sdata = compat_urllib_parse.urlencode(
1166 {u'__youtubedl_smuggle': json.dumps(data)})
1167 return url + u'#' + sdata
1170 def unsmuggle_url(smug_url, default=None):
1171 if not '#__youtubedl_smuggle' in smug_url:
1172 return smug_url, default
1173 url, _, sdata = smug_url.rpartition(u'#')
1174 jsond = compat_parse_qs(sdata)[u'__youtubedl_smuggle'][0]
1175 data = json.loads(jsond)
1179 def format_bytes(bytes):
1182 if type(bytes) is str:
1183 bytes = float(bytes)
1187 exponent = int(math.log(bytes, 1024.0))
1188 suffix = [u'B', u'KiB', u'MiB', u'GiB', u'TiB', u'PiB', u'EiB', u'ZiB', u'YiB'][exponent]
1189 converted = float(bytes) / float(1024 ** exponent)
1190 return u'%.2f%s' % (converted, suffix)
1193 def str_to_int(int_str):
1194 int_str = re.sub(r'[,\.]', u'', int_str)
1198 def get_term_width():
1199 columns = os.environ.get('COLUMNS', None)
1204 sp = subprocess.Popen(
1206 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
1207 out, err = sp.communicate()
1208 return int(out.split()[1])
1214 def month_by_name(name):
1215 """ Return the number of a month by (locale-independently) English name """
1218 u'January', u'February', u'March', u'April', u'May', u'June',
1219 u'July', u'August', u'September', u'October', u'November', u'December']
1221 return ENGLISH_NAMES.index(name) + 1
1226 def fix_xml_ampersands(xml_str):
1227 """Replace all the '&' by '&' in XML"""
1229 r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1234 def setproctitle(title):
1235 assert isinstance(title, compat_str)
1237 libc = ctypes.cdll.LoadLibrary("libc.so.6")
1240 title_bytes = title.encode('utf-8')
1241 buf = ctypes.create_string_buffer(len(title_bytes))
1242 buf.value = title_bytes
1244 libc.prctl(15, buf, 0, 0, 0)
1245 except AttributeError:
1246 return # Strange libc, just skip this
1249 def remove_start(s, start):
1250 if s.startswith(start):
1251 return s[len(start):]
1255 def url_basename(url):
1256 path = compat_urlparse.urlparse(url).path
1257 return path.strip(u'/').split(u'/')[-1]
1260 class HEADRequest(compat_urllib_request.Request):
1261 def get_method(self):
1265 def int_or_none(v, scale=1, default=None, get_attr=None):
1268 v = getattr(v, get_attr, None)
1269 return default if v is None else (int(v) // scale)
1272 def float_or_none(v, scale=1, default=None):
1273 return default if v is None else (float(v) / scale)
1276 def parse_duration(s):
1281 r'(?:(?:(?P<hours>[0-9]+)[:h])?(?P<mins>[0-9]+)[:m])?(?P<secs>[0-9]+)s?(?::[0-9]+)?$', s)
1284 res = int(m.group('secs'))
1286 res += int(m.group('mins')) * 60
1287 if m.group('hours'):
1288 res += int(m.group('hours')) * 60 * 60
1292 def prepend_extension(filename, ext):
1293 name, real_ext = os.path.splitext(filename)
1294 return u'{0}.{1}{2}'.format(name, ext, real_ext)
1297 def check_executable(exe, args=[]):
1298 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
1299 args can be a list of arguments for a short output (like -version) """
1301 subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
1307 class PagedList(object):
1308 def __init__(self, pagefunc, pagesize):
1309 self._pagefunc = pagefunc
1310 self._pagesize = pagesize
1313 # This is only useful for tests
1314 return len(self.getslice())
1316 def getslice(self, start=0, end=None):
1318 for pagenum in itertools.count(start // self._pagesize):
1319 firstid = pagenum * self._pagesize
1320 nextfirstid = pagenum * self._pagesize + self._pagesize
1321 if start >= nextfirstid:
1324 page_results = list(self._pagefunc(pagenum))
1327 start % self._pagesize
1328 if firstid <= start < nextfirstid
1332 ((end - 1) % self._pagesize) + 1
1333 if (end is not None and firstid <= end <= nextfirstid)
1336 if startv != 0 or endv is not None:
1337 page_results = page_results[startv:endv]
1338 res.extend(page_results)
1340 # A little optimization - if current page is not "full", ie. does
1341 # not contain page_size videos then we can assume that this page
1342 # is the last one - there are no more ids on further pages -
1343 # i.e. no need to query again.
1344 if len(page_results) + startv < self._pagesize:
1347 # If we got the whole page, but the next page is not interesting,
1348 # break out early as well
1349 if end == nextfirstid:
1354 def uppercase_escape(s):
1355 unicode_escape = codecs.getdecoder('unicode_escape')
1357 r'\\U[0-9a-fA-F]{8}',
1358 lambda m: unicode_escape(m.group(0))[0],
1362 struct.pack(u'!I', 0)
1364 # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
1365 def struct_pack(spec, *args):
1366 if isinstance(spec, compat_str):
1367 spec = spec.encode('ascii')
1368 return struct.pack(spec, *args)
1370 def struct_unpack(spec, *args):
1371 if isinstance(spec, compat_str):
1372 spec = spec.encode('ascii')
1373 return struct.unpack(spec, *args)
1375 struct_pack = struct.pack
1376 struct_unpack = struct.unpack
1379 def read_batch_urls(batch_fd):
1381 if not isinstance(url, compat_str):
1382 url = url.decode('utf-8', 'replace')
1383 BOM_UTF8 = u'\xef\xbb\xbf'
1384 if url.startswith(BOM_UTF8):
1385 url = url[len(BOM_UTF8):]
1387 if url.startswith(('#', ';', ']')):
1391 with contextlib.closing(batch_fd) as fd:
1392 return [url for url in map(fixup, fd) if url]
1395 def urlencode_postdata(*args, **kargs):
1396 return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
1400 class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
1401 def doctype(self, name, pubid, system):
1402 pass # Ignore doctypes
1404 parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
1405 kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
1406 return xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
1409 if sys.version_info < (3, 0) and sys.platform == 'win32':
1410 def compat_getpass(prompt, *args, **kwargs):
1411 if isinstance(prompt, compat_str):
1412 prompt = prompt.encode(preferredencoding())
1413 return getpass.getpass(prompt, *args, **kwargs)
1415 compat_getpass = getpass.getpass
1427 def strip_jsonp(code):
1428 return re.sub(r'(?s)^[a-zA-Z_]+\s*\(\s*(.*)\);\s*?\s*$', r'\1', code)
1431 def qualities(quality_ids):
1432 """ Get a numeric quality value out of a list of possible values """
1435 return quality_ids.index(qid)
1441 DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s'
1444 subprocess_check_output = subprocess.check_output
1445 except AttributeError:
1446 def subprocess_check_output(*args, **kwargs):
1447 assert 'input' not in kwargs
1448 p = subprocess.Popen(*args, stdout=subprocess.PIPE, **kwargs)
1449 output, _ = p.communicate()
1452 raise subprocess.CalledProcessError(ret, p.args, output=output)