import datetime
import email.utils
import errno
+import functools
import gzip
import itertools
import io
import zlib
from .compat import (
+ compat_basestring,
compat_chr,
compat_getenv,
compat_html_entities,
+ compat_http_client,
compat_parse_qs,
+ compat_socket_create_connection,
compat_str,
compat_urllib_error,
compat_urllib_parse,
def find_xpath_attr(node, xpath, key, val):
# Here comes the crazy part: In 2.6, if the xpath is a unicode,
# .//node does not match if a node is a direct child of . !
- if isinstance(xpath, unicode):
+ if isinstance(xpath, compat_str):
xpath = xpath.encode('ascii')
for f in node.findall(xpath):
xpath = xpath.encode('ascii')
n = node.find(xpath)
- if n is None:
+ if n is None or n.text is None:
if fatal:
name = xpath if name is None else name
raise ExtractorError('Could not find XML element %s' % name)
def clean_html(html):
"""Clean an HTML snippet into a readable string"""
+
+ if html is None: # Convenience for sanitizing descriptions etc.
+ return html
+
# Newline vs <br />
html = html.replace('\n', ' ')
html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html)
return '_'
return char
+ # Handle timestamps
+ s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
result = ''.join(map(replace_insane, s))
if not is_id:
while '__' in result:
if not isinstance(s, compat_str):
# Legacy code that uses byte strings
# Uncomment the following line after fixing all post processors
- #assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
+ # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
s = s.decode('ascii')
return encodeFilename(s, True)
return '%d' % secs
-def make_HTTPS_handler(opts_no_check_certificate, **kwargs):
- if sys.version_info < (3, 2):
- import httplib
-
- class HTTPSConnectionV3(httplib.HTTPSConnection):
- def __init__(self, *args, **kwargs):
- httplib.HTTPSConnection.__init__(self, *args, **kwargs)
-
- def connect(self):
- sock = socket.create_connection((self.host, self.port), self.timeout)
- if getattr(self, '_tunnel_host', False):
- self.sock = sock
- self._tunnel()
- try:
- self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_TLSv1)
- except ssl.SSLError:
- self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_SSLv23)
-
- class HTTPSHandlerV3(compat_urllib_request.HTTPSHandler):
- def https_open(self, req):
- return self.do_open(HTTPSConnectionV3, req)
- return HTTPSHandlerV3(**kwargs)
- elif hasattr(ssl, 'create_default_context'): # Python >= 3.4
- context = ssl.create_default_context(ssl.Purpose.CLIENT_AUTH)
- context.options &= ~ssl.OP_NO_SSLv3 # Allow older, not-as-secure SSLv3
+def make_HTTPS_handler(params, **kwargs):
+ opts_no_check_certificate = params.get('nocheckcertificate', False)
+ if hasattr(ssl, 'create_default_context'): # Python >= 3.4 or 2.7.9
+ context = ssl.create_default_context(ssl.Purpose.SERVER_AUTH)
if opts_no_check_certificate:
+ context.check_hostname = False
context.verify_mode = ssl.CERT_NONE
- return compat_urllib_request.HTTPSHandler(context=context, **kwargs)
+ try:
+ return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
+ except TypeError:
+ # Python 2.7.8
+ # (create_default_context present but HTTPSHandler has no context=)
+ pass
+
+ if sys.version_info < (3, 2):
+ return YoutubeDLHTTPSHandler(params, **kwargs)
else: # Python < 3.4
- context = ssl.SSLContext(ssl.PROTOCOL_SSLv23)
+ context = ssl.SSLContext(ssl.PROTOCOL_TLSv1)
context.verify_mode = (ssl.CERT_NONE
if opts_no_check_certificate
else ssl.CERT_REQUIRED)
context.set_default_verify_paths()
- try:
- context.load_default_certs()
- except AttributeError:
- pass # Python < 3.4
- return compat_urllib_request.HTTPSHandler(context=context, **kwargs)
+ return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
class ExtractorError(Exception):
return ''.join(traceback.format_tb(self.traceback))
+class UnsupportedError(ExtractorError):
+ def __init__(self, url):
+ super(UnsupportedError, self).__init__(
+ 'Unsupported URL: %s' % url, expected=True)
+ self.url = url
+
+
class RegexNotFoundError(ExtractorError):
"""Error when a regex didn't match"""
pass
self.expected = expected
+def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
+ hc = http_class(*args, **kwargs)
+ source_address = ydl_handler._params.get('source_address')
+ if source_address is not None:
+ sa = (source_address, 0)
+ if hasattr(hc, 'source_address'): # Python 2.7+
+ hc.source_address = sa
+ else: # Python 2.6
+ def _hc_connect(self, *args, **kwargs):
+ sock = compat_socket_create_connection(
+ (self.host, self.port), self.timeout, sa)
+ if is_https:
+ self.sock = ssl.wrap_socket(
+ sock, self.key_file, self.cert_file,
+ ssl_version=ssl.PROTOCOL_TLSv1)
+ else:
+ self.sock = sock
+ hc.connect = functools.partial(_hc_connect, hc)
+
+ return hc
+
+
class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
"""Handler for HTTP requests and responses.
public domain.
"""
+ def __init__(self, params, *args, **kwargs):
+ compat_urllib_request.HTTPHandler.__init__(self, *args, **kwargs)
+ self._params = params
+
+ def http_open(self, req):
+ return self.do_open(functools.partial(
+ _create_http_connection, self, compat_http_client.HTTPConnection, False),
+ req)
+
@staticmethod
def deflate(data):
try:
def http_request(self, req):
for h, v in std_headers.items():
- if h not in req.headers:
+ # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
+ # The dict keys are capitalized because of this bug by urllib
+ if h.capitalize() not in req.headers:
req.add_header(h, v)
if 'Youtubedl-no-compression' in req.headers:
if 'Accept-encoding' in req.headers:
del req.headers['Accept-encoding']
del req.headers['Youtubedl-no-compression']
- if 'Youtubedl-user-agent' in req.headers:
- if 'User-agent' in req.headers:
- del req.headers['User-agent']
- req.headers['User-agent'] = req.headers['Youtubedl-user-agent']
- del req.headers['Youtubedl-user-agent']
if sys.version_info < (2, 7) and '#' in req.get_full_url():
# Python 2.6 is brain-dead when it comes to fragments
https_response = http_response
+class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
+ def __init__(self, params, https_conn_class=None, *args, **kwargs):
+ compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)
+ self._https_conn_class = https_conn_class or compat_http_client.HTTPSConnection
+ self._params = params
+
+ def https_open(self, req):
+ kwargs = {}
+ if hasattr(self, '_context'): # python > 2.6
+ kwargs['context'] = self._context
+ if hasattr(self, '_check_hostname'): # python 3.x
+ kwargs['check_hostname'] = self._check_hostname
+ return self.do_open(functools.partial(
+ _create_http_connection, self, self._https_conn_class, True),
+ req, **kwargs)
+
+
def parse_iso8601(date_str, delimiter='T'):
""" Return a UNIX timestamp from the given date """
return calendar.timegm(dt.timetuple())
-def unified_strdate(date_str):
+def unified_strdate(date_str, day_first=True):
"""Return a string with the date in the format YYYYMMDD"""
if date_str is None:
return None
-
upload_date = None
# Replace commas
date_str = date_str.replace(',', ' ')
# %z (UTC offset) is only supported in python>=3.2
date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
+ # Remove AM/PM + timezone
+ date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
+
format_expressions = [
'%d %B %Y',
'%d %b %Y',
'%b %dst %Y %I:%M%p',
'%b %dnd %Y %I:%M%p',
'%b %dth %Y %I:%M%p',
+ '%Y %m %d',
'%Y-%m-%d',
'%Y/%m/%d',
- '%d.%m.%Y',
- '%d/%m/%Y',
- '%d/%m/%y',
'%Y/%m/%d %H:%M:%S',
- '%d/%m/%Y %H:%M:%S',
'%Y-%m-%d %H:%M:%S',
'%Y-%m-%d %H:%M:%S.%f',
'%d.%m.%Y %H:%M',
'%Y-%m-%dT%H:%M:%S.%f',
'%Y-%m-%dT%H:%M',
]
+ if day_first:
+ format_expressions.extend([
+ '%d.%m.%Y',
+ '%d/%m/%Y',
+ '%d/%m/%y',
+ '%d/%m/%Y %H:%M:%S',
+ ])
+ else:
+ format_expressions.extend([
+ '%m.%d.%Y',
+ '%m/%d/%Y',
+ '%m/%d/%y',
+ '%m/%d/%Y %H:%M:%S',
+ ])
for expression in format_expressions:
try:
upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
Return a datetime object from a string in the format YYYYMMDD or
(now|today)[+-][0-9](day|week|month|year)(s)?"""
today = datetime.date.today()
- if date_str == 'now'or date_str == 'today':
+ if date_str in ('now', 'today'):
return today
+ if date_str == 'yesterday':
+ return today - datetime.timedelta(days=1)
match = re.match('(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
if match is not None:
sign = match.group('sign')
except AttributeError:
# If the output stream doesn't have a fileno, it's virtual
return False
+ except io.UnsupportedOperation:
+ # Some strange Windows pseudo files?
+ return False
if fileno not in WIN_OUTPUT_IDS:
return False
GetStdHandle = ctypes.WINFUNCTYPE(
ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
- ("GetStdHandle", ctypes.windll.kernel32))
+ (b"GetStdHandle", ctypes.windll.kernel32))
h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
WriteConsoleW = ctypes.WINFUNCTYPE(
ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
- ctypes.wintypes.LPVOID)(("WriteConsoleW", ctypes.windll.kernel32))
+ ctypes.wintypes.LPVOID)((b"WriteConsoleW", ctypes.windll.kernel32))
written = ctypes.wintypes.DWORD(0)
- GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)(("GetFileType", ctypes.windll.kernel32))
+ GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b"GetFileType", ctypes.windll.kernel32))
FILE_TYPE_CHAR = 0x0002
FILE_TYPE_REMOTE = 0x8000
GetConsoleMode = ctypes.WINFUNCTYPE(
ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
ctypes.POINTER(ctypes.wintypes.DWORD))(
- ("GetConsoleMode", ctypes.windll.kernel32))
+ (b"GetConsoleMode", ctypes.windll.kernel32))
INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
def not_a_console(handle):
def unsmuggle_url(smug_url, default=None):
- if not '#__youtubedl_smuggle' in smug_url:
+ if '#__youtubedl_smuggle' not in smug_url:
return smug_url, default
url, _, sdata = smug_url.rpartition('#')
jsond = compat_parse_qs(sdata)['__youtubedl_smuggle'][0]
}
units_re = '|'.join(re.escape(u) for u in _UNIT_TABLE)
- m = re.match(r'(?P<num>[0-9]+(?:\.[0-9]*)?)\s*(?P<unit>%s)' % units_re, s)
+ m = re.match(
+ r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)' % units_re, s)
if not m:
return None
- return int(float(m.group('num')) * _UNIT_TABLE[m.group('unit')])
+ num_str = m.group('num').replace(',', '.')
+ mult = _UNIT_TABLE[m.group('unit')]
+ return int(float(num_str) * mult)
def get_term_width():
def parse_duration(s):
- if s is None:
+ if not isinstance(s, compat_basestring):
return None
s = s.strip()
m = re.match(
- r'''(?ix)T?
+ r'''(?ix)(?:P?T)?
+ (?:
+ (?P<only_mins>[0-9.]+)\s*(?:mins?|minutes?)\s*|
+ (?P<only_hours>[0-9.]+)\s*(?:hours?)|
+
(?:
- (?:(?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*)?
+ (?:
+ (?:(?P<days>[0-9]+)\s*(?:[:d]|days?)\s*)?
+ (?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*
+ )?
(?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*
)?
- (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?$''', s)
+ (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?
+ )$''', s)
if not m:
return None
- res = int(m.group('secs'))
+ res = 0
+ if m.group('only_mins'):
+ return float_or_none(m.group('only_mins'), invscale=60)
+ if m.group('only_hours'):
+ return float_or_none(m.group('only_hours'), invscale=60 * 60)
+ if m.group('secs'):
+ res += int(m.group('secs'))
if m.group('mins'):
res += int(m.group('mins')) * 60
- if m.group('hours'):
- res += int(m.group('hours')) * 60 * 60
+ if m.group('hours'):
+ res += int(m.group('hours')) * 60 * 60
+ if m.group('days'):
+ res += int(m.group('days')) * 24 * 60 * 60
if m.group('ms'):
res += float(m.group('ms'))
return res
def get_exe_version(exe, args=['--version'],
- version_re=r'version\s+([0-9._-a-zA-Z]+)',
- unrecognized='present'):
+ version_re=None, unrecognized='present'):
""" Returns the version of the specified executable,
or False if the executable is not present """
try:
- out, err = subprocess.Popen(
+ out, _ = subprocess.Popen(
[exe] + args,
stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
except OSError:
return False
- firstline = out.partition(b'\n')[0].decode('ascii', 'ignore')
- m = re.search(version_re, firstline)
+ if isinstance(out, bytes): # Python 2.x
+ out = out.decode('ascii', 'ignore')
+ return detect_exe_version(out, version_re, unrecognized)
+
+
+def detect_exe_version(output, version_re=None, unrecognized='present'):
+ assert isinstance(output, compat_str)
+ if version_re is None:
+ version_re = r'version\s+([-0-9._a-zA-Z]+)'
+ m = re.search(version_re, output)
if m:
return m.group(1)
else:
def escape_rfc3986(s):
"""Escape non-ASCII characters as suggested by RFC 3986"""
- if sys.version_info < (3, 0) and isinstance(s, unicode):
+ if sys.version_info < (3, 0) and isinstance(s, compat_str):
s = s.encode('utf-8')
return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]")
res = re.sub(r'''(?x)
"(?:[^"\\]*(?:\\\\|\\")?)*"|
'(?:[^'\\]*(?:\\\\|\\')?)*'|
- [a-zA-Z_][a-zA-Z_0-9]*
+ [a-zA-Z_][.a-zA-Z_0-9]*
''', fix_kv, code)
res = re.sub(r',(\s*\])', lambda m: m.group(1), res)
return res
def version_tuple(v):
- return [int(e) for e in v.split('.')]
+ return tuple(int(e) for e in re.split(r'[-.]', v))
def is_outdated_version(version, limit, assume_new=True):
def args_to_str(args):
# Get a short string representation for a subprocess command
return ' '.join(shlex_quote(a) for a in args)
+
+
+def urlhandle_detect_ext(url_handle):
+ try:
+ url_handle.headers
+ getheader = lambda h: url_handle.headers[h]
+ except AttributeError: # Python < 3
+ getheader = url_handle.info().getheader
+
+ cd = getheader('Content-Disposition')
+ if cd:
+ m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd)
+ if m:
+ e = determine_ext(m.group('filename'), default_ext=None)
+ if e:
+ return e
+
+ return getheader('Content-Type').split("/")[1]
+
+
+def age_restricted(content_limit, age_limit):
+ """ Returns True iff the content should be blocked """
+
+ if age_limit is None: # No limit set
+ return False
+ if content_limit is None:
+ return False # Content available for everyone
+ return age_limit < content_limit
+
+
+def is_html(first_bytes):
+ """ Detect whether a file contains HTML by examining its first bytes. """
+
+ BOMS = [
+ (b'\xef\xbb\xbf', 'utf-8'),
+ (b'\x00\x00\xfe\xff', 'utf-32-be'),
+ (b'\xff\xfe\x00\x00', 'utf-32-le'),
+ (b'\xff\xfe', 'utf-16-le'),
+ (b'\xfe\xff', 'utf-16-be'),
+ ]
+ for bom, enc in BOMS:
+ if first_bytes.startswith(bom):
+ s = first_bytes[len(bom):].decode(enc, 'replace')
+ break
+ else:
+ s = first_bytes.decode('utf-8', 'replace')
+
+ return re.match(r'^\s*<', s)
+
+
+def determine_protocol(info_dict):
+ protocol = info_dict.get('protocol')
+ if protocol is not None:
+ return protocol
+
+ url = info_dict['url']
+ if url.startswith('rtmp'):
+ return 'rtmp'
+ elif url.startswith('mms'):
+ return 'mms'
+ elif url.startswith('rtsp'):
+ return 'rtsp'
+
+ ext = determine_ext(url)
+ if ext == 'm3u8':
+ return 'm3u8'
+ elif ext == 'f4m':
+ return 'f4m'
+
+ return compat_urllib_parse_urlparse(url).scheme
+
+
+def render_table(header_row, data):
+ """ Render a list of rows, each as a list of values """
+ table = [header_row] + data
+ max_lens = [max(len(compat_str(v)) for v in col) for col in zip(*table)]
+ format_str = ' '.join('%-' + compat_str(ml + 1) + 's' for ml in max_lens[:-1]) + '%s'
+ return '\n'.join(format_str % tuple(row) for row in table)