1 from __future__ import unicode_literals
16 import urllib.request as compat_urllib_request
17 except ImportError: # Python 2
18 import urllib2 as compat_urllib_request
21 import urllib.error as compat_urllib_error
22 except ImportError: # Python 2
23 import urllib2 as compat_urllib_error
26 import urllib.parse as compat_urllib_parse
27 except ImportError: # Python 2
28 import urllib as compat_urllib_parse
31 from urllib.parse import urlparse as compat_urllib_parse_urlparse
32 except ImportError: # Python 2
33 from urlparse import urlparse as compat_urllib_parse_urlparse
36 import urllib.parse as compat_urlparse
37 except ImportError: # Python 2
38 import urlparse as compat_urlparse
41 import http.cookiejar as compat_cookiejar
42 except ImportError: # Python 2
43 import cookielib as compat_cookiejar
46 import html.entities as compat_html_entities
47 except ImportError: # Python 2
48 import htmlentitydefs as compat_html_entities
51 import http.client as compat_http_client
52 except ImportError: # Python 2
53 import httplib as compat_http_client
56 from urllib.error import HTTPError as compat_HTTPError
57 except ImportError: # Python 2
58 from urllib2 import HTTPError as compat_HTTPError
61 from urllib.request import urlretrieve as compat_urlretrieve
62 except ImportError: # Python 2
63 from urllib import urlretrieve as compat_urlretrieve
67 from subprocess import DEVNULL
68 compat_subprocess_get_DEVNULL = lambda: DEVNULL
70 compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w')
73 import http.server as compat_http_server
75 import BaseHTTPServer as compat_http_server
78 from urllib.parse import unquote as compat_urllib_parse_unquote
80 def compat_urllib_parse_unquote(string, encoding='utf-8', errors='replace'):
83 res = string.split('%')
90 # pct_sequence: contiguous sequence of percent-encoded bytes, decoded
97 pct_sequence += item[:2].decode('hex')
100 # This segment was just a single percent-encoded character.
101 # May be part of a sequence of code units, so delay decoding.
102 # (Stored in pct_sequence).
106 # Encountered non-percent-encoded characters. Flush the current
108 string += pct_sequence.decode(encoding, errors) + rest
111 # Flush the final pct_sequence
112 string += pct_sequence.decode(encoding, errors)
116 compat_str = unicode # Python 2
121 compat_basestring = basestring # Python 2
123 compat_basestring = str
126 compat_chr = unichr # Python 2
131 from xml.etree.ElementTree import ParseError as compat_xml_parse_error
132 except ImportError: # Python 2.6
133 from xml.parsers.expat import ExpatError as compat_xml_parse_error
137 from urllib.parse import parse_qs as compat_parse_qs
138 except ImportError: # Python 2
139 # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib.
140 # Python 2's version is apparently totally broken
142 def _parse_qsl(qs, keep_blank_values=False, strict_parsing=False,
143 encoding='utf-8', errors='replace'):
144 qs, _coerce_result = qs, compat_str
145 pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]
147 for name_value in pairs:
148 if not name_value and not strict_parsing:
150 nv = name_value.split('=', 1)
153 raise ValueError("bad query field: %r" % (name_value,))
154 # Handle case of a control-name with no equal sign
155 if keep_blank_values:
159 if len(nv[1]) or keep_blank_values:
160 name = nv[0].replace('+', ' ')
161 name = compat_urllib_parse_unquote(
162 name, encoding=encoding, errors=errors)
163 name = _coerce_result(name)
164 value = nv[1].replace('+', ' ')
165 value = compat_urllib_parse_unquote(
166 value, encoding=encoding, errors=errors)
167 value = _coerce_result(value)
168 r.append((name, value))
171 def compat_parse_qs(qs, keep_blank_values=False, strict_parsing=False,
172 encoding='utf-8', errors='replace'):
174 pairs = _parse_qsl(qs, keep_blank_values, strict_parsing,
175 encoding=encoding, errors=errors)
176 for name, value in pairs:
177 if name in parsed_result:
178 parsed_result[name].append(value)
180 parsed_result[name] = [value]
184 from shlex import quote as shlex_quote
185 except ImportError: # Python < 3.3
187 if re.match(r'^[-_\w./]+$', s):
190 return "'" + s.replace("'", "'\"'\"'") + "'"
200 if sys.version_info >= (3, 0):
201 compat_getenv = os.getenv
202 compat_expanduser = os.path.expanduser
204 # Environment variables should be decoded with filesystem encoding.
205 # Otherwise it will fail if any non-ASCII characters present (see #3854 #3217 #2918)
207 def compat_getenv(key, default=None):
208 from .utils import get_filesystem_encoding
209 env = os.getenv(key, default)
211 env = env.decode(get_filesystem_encoding())
214 # HACK: The default implementations of os.path.expanduser from cpython do not decode
215 # environment variables with filesystem encoding. We will work around this by
216 # providing adjusted implementations.
217 # The following are os.path.expanduser implementations from cpython 2.7.8 stdlib
218 # for different platforms with correct environment variables decoding.
220 if os.name == 'posix':
221 def compat_expanduser(path):
222 """Expand ~ and ~user constructions. If user or $HOME is unknown,
224 if not path.startswith('~'):
226 i = path.find('/', 1)
230 if 'HOME' not in os.environ:
232 userhome = pwd.getpwuid(os.getuid()).pw_dir
234 userhome = compat_getenv('HOME')
238 pwent = pwd.getpwnam(path[1:i])
241 userhome = pwent.pw_dir
242 userhome = userhome.rstrip('/')
243 return (userhome + path[i:]) or '/'
244 elif os.name == 'nt' or os.name == 'ce':
245 def compat_expanduser(path):
246 """Expand ~ and ~user constructs.
248 If user or $HOME is unknown, do nothing."""
252 while i < n and path[i] not in '/\\':
255 if 'HOME' in os.environ:
256 userhome = compat_getenv('HOME')
257 elif 'USERPROFILE' in os.environ:
258 userhome = compat_getenv('USERPROFILE')
259 elif 'HOMEPATH' not in os.environ:
263 drive = compat_getenv('HOMEDRIVE')
266 userhome = os.path.join(drive, compat_getenv('HOMEPATH'))
269 userhome = os.path.join(os.path.dirname(userhome), path[1:i])
271 return userhome + path[i:]
273 compat_expanduser = os.path.expanduser
276 if sys.version_info < (3, 0):
278 from .utils import preferredencoding
279 print(s.encode(preferredencoding(), 'xmlcharrefreplace'))
282 assert isinstance(s, compat_str)
287 subprocess_check_output = subprocess.check_output
288 except AttributeError:
289 def subprocess_check_output(*args, **kwargs):
290 assert 'input' not in kwargs
291 p = subprocess.Popen(*args, stdout=subprocess.PIPE, **kwargs)
292 output, _ = p.communicate()
295 raise subprocess.CalledProcessError(ret, p.args, output=output)
298 if sys.version_info < (3, 0) and sys.platform == 'win32':
299 def compat_getpass(prompt, *args, **kwargs):
300 if isinstance(prompt, compat_str):
301 from .utils import preferredencoding
302 prompt = prompt.encode(preferredencoding())
303 return getpass.getpass(prompt, *args, **kwargs)
305 compat_getpass = getpass.getpass
307 # Old 2.6 and 2.7 releases require kwargs to be bytes
311 _testfunc(**{'x': 0})
313 def compat_kwargs(kwargs):
314 return dict((bytes(k), v) for k, v in kwargs.items())
316 compat_kwargs = lambda kwargs: kwargs
319 if sys.version_info < (2, 7):
320 def compat_socket_create_connection(address, timeout, source_address=None):
323 for res in socket.getaddrinfo(host, port, 0, socket.SOCK_STREAM):
324 af, socktype, proto, canonname, sa = res
327 sock = socket.socket(af, socktype, proto)
328 sock.settimeout(timeout)
330 sock.bind(source_address)
333 except socket.error as _:
340 raise socket.error("getaddrinfo returns an empty list")
342 compat_socket_create_connection = socket.create_connection
345 # Fix https://github.com/rg3/youtube-dl/issues/4223
346 # See http://bugs.python.org/issue9161 for what is broken
347 def workaround_optparse_bug9161():
348 op = optparse.OptionParser()
349 og = optparse.OptionGroup(op, 'foo')
353 real_add_option = optparse.OptionGroup.add_option
355 def _compat_add_option(self, *args, **kwargs):
357 v.encode('ascii', 'replace') if isinstance(v, compat_str)
359 bargs = [enc(a) for a in args]
361 (k, enc(v)) for k, v in kwargs.items())
362 return real_add_option(self, *bargs, **bkwargs)
363 optparse.OptionGroup.add_option = _compat_add_option
365 if hasattr(shutil, 'get_terminal_size'): # Python >= 3.3
366 compat_get_terminal_size = shutil.get_terminal_size
368 _terminal_size = collections.namedtuple('terminal_size', ['columns', 'lines'])
370 def compat_get_terminal_size():
371 columns = compat_getenv('COLUMNS', None)
373 columns = int(columns)
376 lines = compat_getenv('LINES', None)
383 sp = subprocess.Popen(
385 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
386 out, err = sp.communicate()
387 lines, columns = map(int, out.split())
390 return _terminal_size(columns, lines)
393 itertools.count(start=0, step=1)
394 compat_itertools_count = itertools.count
395 except TypeError: # Python 2.6
396 def compat_itertools_count(start=0, step=1):
408 'compat_get_terminal_size',
411 'compat_html_entities',
412 'compat_http_client',
413 'compat_http_server',
414 'compat_itertools_count',
419 'compat_socket_create_connection',
421 'compat_subprocess_get_DEVNULL',
422 'compat_urllib_error',
423 'compat_urllib_parse',
424 'compat_urllib_parse_unquote',
425 'compat_urllib_parse_urlparse',
426 'compat_urllib_request',
428 'compat_urlretrieve',
429 'compat_xml_parse_error',
431 'subprocess_check_output',
432 'workaround_optparse_bug9161',