1 from __future__ import unicode_literals
17 import urllib.request as compat_urllib_request
18 except ImportError: # Python 2
19 import urllib2 as compat_urllib_request
22 import urllib.error as compat_urllib_error
23 except ImportError: # Python 2
24 import urllib2 as compat_urllib_error
27 import urllib.parse as compat_urllib_parse
28 except ImportError: # Python 2
29 import urllib as compat_urllib_parse
32 from urllib.parse import urlparse as compat_urllib_parse_urlparse
33 except ImportError: # Python 2
34 from urlparse import urlparse as compat_urllib_parse_urlparse
37 import urllib.parse as compat_urlparse
38 except ImportError: # Python 2
39 import urlparse as compat_urlparse
42 import http.cookiejar as compat_cookiejar
43 except ImportError: # Python 2
44 import cookielib as compat_cookiejar
47 import http.cookies as compat_cookies
48 except ImportError: # Python 2
49 import Cookie as compat_cookies
52 import html.entities as compat_html_entities
53 except ImportError: # Python 2
54 import htmlentitydefs as compat_html_entities
57 import http.client as compat_http_client
58 except ImportError: # Python 2
59 import httplib as compat_http_client
62 from urllib.error import HTTPError as compat_HTTPError
63 except ImportError: # Python 2
64 from urllib2 import HTTPError as compat_HTTPError
67 from urllib.request import urlretrieve as compat_urlretrieve
68 except ImportError: # Python 2
69 from urllib import urlretrieve as compat_urlretrieve
73 from subprocess import DEVNULL
74 compat_subprocess_get_DEVNULL = lambda: DEVNULL
76 compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w')
79 import http.server as compat_http_server
81 import BaseHTTPServer as compat_http_server
84 compat_str = unicode # Python 2
89 from urllib.parse import unquote_to_bytes as compat_urllib_parse_unquote_to_bytes
90 from urllib.parse import unquote as compat_urllib_parse_unquote
91 from urllib.parse import unquote_plus as compat_urllib_parse_unquote_plus
92 except ImportError: # Python 2
93 _asciire = (compat_urllib_parse._asciire if hasattr(compat_urllib_parse, '_asciire')
94 else re.compile('([\x00-\x7f]+)'))
96 # HACK: The following are the correct unquote_to_bytes, unquote and unquote_plus
97 # implementations from cpython 3.4.3's stdlib. Python 2's version
98 # is apparently broken (see https://github.com/rg3/youtube-dl/pull/6244)
100 def compat_urllib_parse_unquote_to_bytes(string):
101 """unquote_to_bytes('abc%20def') -> b'abc def'."""
102 # Note: strings are encoded as UTF-8. This is only an issue if it contains
103 # unescaped non-ASCII characters, which URIs should not.
105 # Is it a string-like object?
108 if isinstance(string, compat_str):
109 string = string.encode('utf-8')
110 bits = string.split(b'%')
115 for item in bits[1:]:
117 append(compat_urllib_parse._hextochr[item[:2]])
124 def compat_urllib_parse_unquote(string, encoding='utf-8', errors='replace'):
125 """Replace %xx escapes by their single-character equivalent. The optional
126 encoding and errors parameters specify how to decode percent-encoded
127 sequences into Unicode characters, as accepted by the bytes.decode()
129 By default, percent-encoded sequences are decoded with UTF-8, and invalid
130 sequences are replaced by a placeholder character.
132 unquote('abc%20def') -> 'abc def'.
134 if '%' not in string:
141 bits = _asciire.split(string)
144 for i in range(1, len(bits), 2):
145 append(compat_urllib_parse_unquote_to_bytes(bits[i]).decode(encoding, errors))
149 def compat_urllib_parse_unquote_plus(string, encoding='utf-8', errors='replace'):
150 """Like unquote(), but also replace plus signs by spaces, as required for
151 unquoting HTML form values.
153 unquote_plus('%7e/abc+def') -> '~/abc def'
155 string = string.replace('+', ' ')
156 return compat_urllib_parse_unquote(string, encoding, errors)
159 compat_basestring = basestring # Python 2
161 compat_basestring = str
164 compat_chr = unichr # Python 2
169 from xml.etree.ElementTree import ParseError as compat_xml_parse_error
170 except ImportError: # Python 2.6
171 from xml.parsers.expat import ExpatError as compat_xml_parse_error
175 from urllib.parse import parse_qs as compat_parse_qs
176 except ImportError: # Python 2
177 # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib.
178 # Python 2's version is apparently totally broken
180 def _parse_qsl(qs, keep_blank_values=False, strict_parsing=False,
181 encoding='utf-8', errors='replace'):
182 qs, _coerce_result = qs, compat_str
183 pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]
185 for name_value in pairs:
186 if not name_value and not strict_parsing:
188 nv = name_value.split('=', 1)
191 raise ValueError("bad query field: %r" % (name_value,))
192 # Handle case of a control-name with no equal sign
193 if keep_blank_values:
197 if len(nv[1]) or keep_blank_values:
198 name = nv[0].replace('+', ' ')
199 name = compat_urllib_parse_unquote(
200 name, encoding=encoding, errors=errors)
201 name = _coerce_result(name)
202 value = nv[1].replace('+', ' ')
203 value = compat_urllib_parse_unquote(
204 value, encoding=encoding, errors=errors)
205 value = _coerce_result(value)
206 r.append((name, value))
209 def compat_parse_qs(qs, keep_blank_values=False, strict_parsing=False,
210 encoding='utf-8', errors='replace'):
212 pairs = _parse_qsl(qs, keep_blank_values, strict_parsing,
213 encoding=encoding, errors=errors)
214 for name, value in pairs:
215 if name in parsed_result:
216 parsed_result[name].append(value)
218 parsed_result[name] = [value]
222 from shlex import quote as shlex_quote
223 except ImportError: # Python < 3.3
225 if re.match(r'^[-_\w./]+$', s):
228 return "'" + s.replace("'", "'\"'\"'") + "'"
231 if sys.version_info >= (2, 7, 3):
232 compat_shlex_split = shlex.split
234 # Working around shlex issue with unicode strings on some python 2
235 # versions (see http://bugs.python.org/issue1548891)
236 def compat_shlex_split(s, comments=False, posix=True):
237 if isinstance(s, compat_str):
238 s = s.encode('utf-8')
239 return shlex.split(s, comments, posix)
249 if sys.version_info >= (3, 0):
250 compat_getenv = os.getenv
251 compat_expanduser = os.path.expanduser
253 # Environment variables should be decoded with filesystem encoding.
254 # Otherwise it will fail if any non-ASCII characters present (see #3854 #3217 #2918)
256 def compat_getenv(key, default=None):
257 from .utils import get_filesystem_encoding
258 env = os.getenv(key, default)
260 env = env.decode(get_filesystem_encoding())
263 # HACK: The default implementations of os.path.expanduser from cpython do not decode
264 # environment variables with filesystem encoding. We will work around this by
265 # providing adjusted implementations.
266 # The following are os.path.expanduser implementations from cpython 2.7.8 stdlib
267 # for different platforms with correct environment variables decoding.
269 if os.name == 'posix':
270 def compat_expanduser(path):
271 """Expand ~ and ~user constructions. If user or $HOME is unknown,
273 if not path.startswith('~'):
275 i = path.find('/', 1)
279 if 'HOME' not in os.environ:
281 userhome = pwd.getpwuid(os.getuid()).pw_dir
283 userhome = compat_getenv('HOME')
287 pwent = pwd.getpwnam(path[1:i])
290 userhome = pwent.pw_dir
291 userhome = userhome.rstrip('/')
292 return (userhome + path[i:]) or '/'
293 elif os.name == 'nt' or os.name == 'ce':
294 def compat_expanduser(path):
295 """Expand ~ and ~user constructs.
297 If user or $HOME is unknown, do nothing."""
301 while i < n and path[i] not in '/\\':
304 if 'HOME' in os.environ:
305 userhome = compat_getenv('HOME')
306 elif 'USERPROFILE' in os.environ:
307 userhome = compat_getenv('USERPROFILE')
308 elif 'HOMEPATH' not in os.environ:
312 drive = compat_getenv('HOMEDRIVE')
315 userhome = os.path.join(drive, compat_getenv('HOMEPATH'))
318 userhome = os.path.join(os.path.dirname(userhome), path[1:i])
320 return userhome + path[i:]
322 compat_expanduser = os.path.expanduser
325 if sys.version_info < (3, 0):
327 from .utils import preferredencoding
328 print(s.encode(preferredencoding(), 'xmlcharrefreplace'))
331 assert isinstance(s, compat_str)
336 subprocess_check_output = subprocess.check_output
337 except AttributeError:
338 def subprocess_check_output(*args, **kwargs):
339 assert 'input' not in kwargs
340 p = subprocess.Popen(*args, stdout=subprocess.PIPE, **kwargs)
341 output, _ = p.communicate()
344 raise subprocess.CalledProcessError(ret, p.args, output=output)
347 if sys.version_info < (3, 0) and sys.platform == 'win32':
348 def compat_getpass(prompt, *args, **kwargs):
349 if isinstance(prompt, compat_str):
350 from .utils import preferredencoding
351 prompt = prompt.encode(preferredencoding())
352 return getpass.getpass(prompt, *args, **kwargs)
354 compat_getpass = getpass.getpass
356 # Old 2.6 and 2.7 releases require kwargs to be bytes
360 _testfunc(**{'x': 0})
362 def compat_kwargs(kwargs):
363 return dict((bytes(k), v) for k, v in kwargs.items())
365 compat_kwargs = lambda kwargs: kwargs
368 if sys.version_info < (2, 7):
369 def compat_socket_create_connection(address, timeout, source_address=None):
372 for res in socket.getaddrinfo(host, port, 0, socket.SOCK_STREAM):
373 af, socktype, proto, canonname, sa = res
376 sock = socket.socket(af, socktype, proto)
377 sock.settimeout(timeout)
379 sock.bind(source_address)
382 except socket.error as _:
389 raise socket.error("getaddrinfo returns an empty list")
391 compat_socket_create_connection = socket.create_connection
394 # Fix https://github.com/rg3/youtube-dl/issues/4223
395 # See http://bugs.python.org/issue9161 for what is broken
396 def workaround_optparse_bug9161():
397 op = optparse.OptionParser()
398 og = optparse.OptionGroup(op, 'foo')
402 real_add_option = optparse.OptionGroup.add_option
404 def _compat_add_option(self, *args, **kwargs):
406 v.encode('ascii', 'replace') if isinstance(v, compat_str)
408 bargs = [enc(a) for a in args]
410 (k, enc(v)) for k, v in kwargs.items())
411 return real_add_option(self, *bargs, **bkwargs)
412 optparse.OptionGroup.add_option = _compat_add_option
414 if hasattr(shutil, 'get_terminal_size'): # Python >= 3.3
415 compat_get_terminal_size = shutil.get_terminal_size
417 _terminal_size = collections.namedtuple('terminal_size', ['columns', 'lines'])
419 def compat_get_terminal_size():
420 columns = compat_getenv('COLUMNS', None)
422 columns = int(columns)
425 lines = compat_getenv('LINES', None)
432 sp = subprocess.Popen(
434 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
435 out, err = sp.communicate()
436 lines, columns = map(int, out.split())
439 return _terminal_size(columns, lines)
442 itertools.count(start=0, step=1)
443 compat_itertools_count = itertools.count
444 except TypeError: # Python 2.6
445 def compat_itertools_count(start=0, step=1):
451 if sys.version_info >= (3, 0):
452 from tokenize import tokenize as compat_tokenize_tokenize
454 from tokenize import generate_tokens as compat_tokenize_tokenize
463 'compat_get_terminal_size',
466 'compat_html_entities',
467 'compat_http_client',
468 'compat_http_server',
469 'compat_itertools_count',
474 'compat_shlex_split',
475 'compat_socket_create_connection',
477 'compat_subprocess_get_DEVNULL',
478 'compat_tokenize_tokenize',
479 'compat_urllib_error',
480 'compat_urllib_parse',
481 'compat_urllib_parse_unquote',
482 'compat_urllib_parse_unquote_plus',
483 'compat_urllib_parse_unquote_to_bytes',
484 'compat_urllib_parse_urlparse',
485 'compat_urllib_request',
487 'compat_urlretrieve',
488 'compat_xml_parse_error',
490 'subprocess_check_output',
491 'workaround_optparse_bug9161',