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_to_bytes as compat_urllib_parse_unquote_to_bytes
79 from urllib.parse import unquote as compat_urllib_parse_unquote
80 from urllib.parse import unquote_plus as compat_urllib_parse_unquote_plus
81 except ImportError: # Python 2
82 _asciire = re.compile('([\x00-\x7f]+)') if sys.version_info < (2, 7) else compat_urllib_parse._asciire
84 # HACK: The following are the correct unquote_to_bytes, unquote and unquote_plus
85 # implementations from cpython 3.4.3's stdlib. Python 2's version
86 # is apparently broken (see https://github.com/rg3/youtube-dl/pull/6244)
88 def compat_urllib_parse_unquote_to_bytes(string):
89 """unquote_to_bytes('abc%20def') -> b'abc def'."""
90 # Note: strings are encoded as UTF-8. This is only an issue if it contains
91 # unescaped non-ASCII characters, which URIs should not.
93 # Is it a string-like object?
96 if isinstance(string, unicode):
97 string = string.encode('utf-8')
98 bits = string.split(b'%')
103 for item in bits[1:]:
105 append(compat_urllib_parse._hextochr[item[:2]])
112 def compat_urllib_parse_unquote(string, encoding='utf-8', errors='replace'):
113 """Replace %xx escapes by their single-character equivalent. The optional
114 encoding and errors parameters specify how to decode percent-encoded
115 sequences into Unicode characters, as accepted by the bytes.decode()
117 By default, percent-encoded sequences are decoded with UTF-8, and invalid
118 sequences are replaced by a placeholder character.
120 unquote('abc%20def') -> 'abc def'.
122 if '%' not in string:
129 bits = _asciire.split(string)
132 for i in range(1, len(bits), 2):
133 append(compat_urllib_parse_unquote_to_bytes(bits[i]).decode(encoding, errors))
137 def compat_urllib_parse_unquote_plus(string, encoding='utf-8', errors='replace'):
138 """Like unquote(), but also replace plus signs by spaces, as required for
139 unquoting HTML form values.
141 unquote_plus('%7e/abc+def') -> '~/abc def'
143 string = string.replace('+', ' ')
144 return compat_urllib_parse_unquote(string, encoding, errors)
147 compat_str = unicode # Python 2
152 compat_basestring = basestring # Python 2
154 compat_basestring = str
157 compat_chr = unichr # Python 2
162 from xml.etree.ElementTree import ParseError as compat_xml_parse_error
163 except ImportError: # Python 2.6
164 from xml.parsers.expat import ExpatError as compat_xml_parse_error
168 from urllib.parse import parse_qs as compat_parse_qs
169 except ImportError: # Python 2
170 # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib.
171 # Python 2's version is apparently totally broken
173 def _parse_qsl(qs, keep_blank_values=False, strict_parsing=False,
174 encoding='utf-8', errors='replace'):
175 qs, _coerce_result = qs, compat_str
176 pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]
178 for name_value in pairs:
179 if not name_value and not strict_parsing:
181 nv = name_value.split('=', 1)
184 raise ValueError("bad query field: %r" % (name_value,))
185 # Handle case of a control-name with no equal sign
186 if keep_blank_values:
190 if len(nv[1]) or keep_blank_values:
191 name = nv[0].replace('+', ' ')
192 name = compat_urllib_parse_unquote(
193 name, encoding=encoding, errors=errors)
194 name = _coerce_result(name)
195 value = nv[1].replace('+', ' ')
196 value = compat_urllib_parse_unquote(
197 value, encoding=encoding, errors=errors)
198 value = _coerce_result(value)
199 r.append((name, value))
202 def compat_parse_qs(qs, keep_blank_values=False, strict_parsing=False,
203 encoding='utf-8', errors='replace'):
205 pairs = _parse_qsl(qs, keep_blank_values, strict_parsing,
206 encoding=encoding, errors=errors)
207 for name, value in pairs:
208 if name in parsed_result:
209 parsed_result[name].append(value)
211 parsed_result[name] = [value]
215 from shlex import quote as shlex_quote
216 except ImportError: # Python < 3.3
218 if re.match(r'^[-_\w./]+$', s):
221 return "'" + s.replace("'", "'\"'\"'") + "'"
231 if sys.version_info >= (3, 0):
232 compat_getenv = os.getenv
233 compat_expanduser = os.path.expanduser
235 # Environment variables should be decoded with filesystem encoding.
236 # Otherwise it will fail if any non-ASCII characters present (see #3854 #3217 #2918)
238 def compat_getenv(key, default=None):
239 from .utils import get_filesystem_encoding
240 env = os.getenv(key, default)
242 env = env.decode(get_filesystem_encoding())
245 # HACK: The default implementations of os.path.expanduser from cpython do not decode
246 # environment variables with filesystem encoding. We will work around this by
247 # providing adjusted implementations.
248 # The following are os.path.expanduser implementations from cpython 2.7.8 stdlib
249 # for different platforms with correct environment variables decoding.
251 if os.name == 'posix':
252 def compat_expanduser(path):
253 """Expand ~ and ~user constructions. If user or $HOME is unknown,
255 if not path.startswith('~'):
257 i = path.find('/', 1)
261 if 'HOME' not in os.environ:
263 userhome = pwd.getpwuid(os.getuid()).pw_dir
265 userhome = compat_getenv('HOME')
269 pwent = pwd.getpwnam(path[1:i])
272 userhome = pwent.pw_dir
273 userhome = userhome.rstrip('/')
274 return (userhome + path[i:]) or '/'
275 elif os.name == 'nt' or os.name == 'ce':
276 def compat_expanduser(path):
277 """Expand ~ and ~user constructs.
279 If user or $HOME is unknown, do nothing."""
283 while i < n and path[i] not in '/\\':
286 if 'HOME' in os.environ:
287 userhome = compat_getenv('HOME')
288 elif 'USERPROFILE' in os.environ:
289 userhome = compat_getenv('USERPROFILE')
290 elif 'HOMEPATH' not in os.environ:
294 drive = compat_getenv('HOMEDRIVE')
297 userhome = os.path.join(drive, compat_getenv('HOMEPATH'))
300 userhome = os.path.join(os.path.dirname(userhome), path[1:i])
302 return userhome + path[i:]
304 compat_expanduser = os.path.expanduser
307 if sys.version_info < (3, 0):
309 from .utils import preferredencoding
310 print(s.encode(preferredencoding(), 'xmlcharrefreplace'))
313 assert isinstance(s, compat_str)
318 subprocess_check_output = subprocess.check_output
319 except AttributeError:
320 def subprocess_check_output(*args, **kwargs):
321 assert 'input' not in kwargs
322 p = subprocess.Popen(*args, stdout=subprocess.PIPE, **kwargs)
323 output, _ = p.communicate()
326 raise subprocess.CalledProcessError(ret, p.args, output=output)
329 if sys.version_info < (3, 0) and sys.platform == 'win32':
330 def compat_getpass(prompt, *args, **kwargs):
331 if isinstance(prompt, compat_str):
332 from .utils import preferredencoding
333 prompt = prompt.encode(preferredencoding())
334 return getpass.getpass(prompt, *args, **kwargs)
336 compat_getpass = getpass.getpass
338 # Old 2.6 and 2.7 releases require kwargs to be bytes
342 _testfunc(**{'x': 0})
344 def compat_kwargs(kwargs):
345 return dict((bytes(k), v) for k, v in kwargs.items())
347 compat_kwargs = lambda kwargs: kwargs
350 if sys.version_info < (2, 7):
351 def compat_socket_create_connection(address, timeout, source_address=None):
354 for res in socket.getaddrinfo(host, port, 0, socket.SOCK_STREAM):
355 af, socktype, proto, canonname, sa = res
358 sock = socket.socket(af, socktype, proto)
359 sock.settimeout(timeout)
361 sock.bind(source_address)
364 except socket.error as _:
371 raise socket.error("getaddrinfo returns an empty list")
373 compat_socket_create_connection = socket.create_connection
376 # Fix https://github.com/rg3/youtube-dl/issues/4223
377 # See http://bugs.python.org/issue9161 for what is broken
378 def workaround_optparse_bug9161():
379 op = optparse.OptionParser()
380 og = optparse.OptionGroup(op, 'foo')
384 real_add_option = optparse.OptionGroup.add_option
386 def _compat_add_option(self, *args, **kwargs):
388 v.encode('ascii', 'replace') if isinstance(v, compat_str)
390 bargs = [enc(a) for a in args]
392 (k, enc(v)) for k, v in kwargs.items())
393 return real_add_option(self, *bargs, **bkwargs)
394 optparse.OptionGroup.add_option = _compat_add_option
396 if hasattr(shutil, 'get_terminal_size'): # Python >= 3.3
397 compat_get_terminal_size = shutil.get_terminal_size
399 _terminal_size = collections.namedtuple('terminal_size', ['columns', 'lines'])
401 def compat_get_terminal_size():
402 columns = compat_getenv('COLUMNS', None)
404 columns = int(columns)
407 lines = compat_getenv('LINES', None)
414 sp = subprocess.Popen(
416 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
417 out, err = sp.communicate()
418 lines, columns = map(int, out.split())
421 return _terminal_size(columns, lines)
424 itertools.count(start=0, step=1)
425 compat_itertools_count = itertools.count
426 except TypeError: # Python 2.6
427 def compat_itertools_count(start=0, step=1):
439 'compat_get_terminal_size',
442 'compat_html_entities',
443 'compat_http_client',
444 'compat_http_server',
445 'compat_itertools_count',
450 'compat_socket_create_connection',
452 'compat_subprocess_get_DEVNULL',
453 'compat_urllib_error',
454 'compat_urllib_parse',
455 'compat_urllib_parse_unquote',
456 'compat_urllib_parse_unquote_plus',
457 'compat_urllib_parse_unquote_to_bytes',
458 'compat_urllib_parse_urlparse',
459 'compat_urllib_request',
461 'compat_urlretrieve',
462 'compat_xml_parse_error',
464 'subprocess_check_output',
465 'workaround_optparse_bug9161',