Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/compat.py

   1 from __future__ import unicode_literals
   2
   3 import collections
   4 import getpass
   5 import optparse
   6 import os
   7 import re
   8 import shutil
   9 import socket
  10 import subprocess
  11 import sys
  12 import itertools
  13
  14
  15 try:
  16     import urllib.request as compat_urllib_request
  17 except ImportError:  # Python 2
  18     import urllib2 as compat_urllib_request
  19
  20 try:
  21     import urllib.error as compat_urllib_error
  22 except ImportError:  # Python 2
  23     import urllib2 as compat_urllib_error
  24
  25 try:
  26     import urllib.parse as compat_urllib_parse
  27 except ImportError:  # Python 2
  28     import urllib as compat_urllib_parse
  29
  30 try:
  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
  34
  35 try:
  36     import urllib.parse as compat_urlparse
  37 except ImportError:  # Python 2
  38     import urlparse as compat_urlparse
  39
  40 try:
  41     import http.cookiejar as compat_cookiejar
  42 except ImportError:  # Python 2
  43     import cookielib as compat_cookiejar
  44
  45 try:
  46     import html.entities as compat_html_entities
  47 except ImportError:  # Python 2
  48     import htmlentitydefs as compat_html_entities
  49
  50 try:
  51     import http.client as compat_http_client
  52 except ImportError:  # Python 2
  53     import httplib as compat_http_client
  54
  55 try:
  56     from urllib.error import HTTPError as compat_HTTPError
  57 except ImportError:  # Python 2
  58     from urllib2 import HTTPError as compat_HTTPError
  59
  60 try:
  61     from urllib.request import urlretrieve as compat_urlretrieve
  62 except ImportError:  # Python 2
  63     from urllib import urlretrieve as compat_urlretrieve
  64
  65
  66 try:
  67     from subprocess import DEVNULL
  68     compat_subprocess_get_DEVNULL = lambda: DEVNULL
  69 except ImportError:
  70     compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w')
  71
  72 try:
  73     import http.server as compat_http_server
  74 except ImportError:
  75     import BaseHTTPServer as compat_http_server
  76
  77 try:
  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 = (compat_urllib_parse._asciire if hasattr(compat_urllib_parse, '_asciire')
  83                 else re.compile('([\x00-\x7f]+)'))
  84
  85     # HACK: The following are the correct unquote_to_bytes, unquote and unquote_plus
  86     # implementations from cpython 3.4.3's stdlib. Python 2's version
  87     # is apparently broken (see https://github.com/rg3/youtube-dl/pull/6244)
  88
  89     def compat_urllib_parse_unquote_to_bytes(string):
  90         """unquote_to_bytes('abc%20def') -> b'abc def'."""
  91         # Note: strings are encoded as UTF-8. This is only an issue if it contains
  92         # unescaped non-ASCII characters, which URIs should not.
  93         if not string:
  94             # Is it a string-like object?
  95             string.split
  96             return b''
  97         if isinstance(string, unicode):
  98             string = string.encode('utf-8')
  99         bits = string.split(b'%')
 100         if len(bits) == 1:
 101             return string
 102         res = [bits[0]]
 103         append = res.append
 104         for item in bits[1:]:
 105             try:
 106                 append(compat_urllib_parse._hextochr[item[:2]])
 107                 append(item[2:])
 108             except KeyError:
 109                 append(b'%')
 110                 append(item)
 111         return b''.join(res)
 112
 113     def compat_urllib_parse_unquote(string, encoding='utf-8', errors='replace'):
 114         """Replace %xx escapes by their single-character equivalent. The optional
 115         encoding and errors parameters specify how to decode percent-encoded
 116         sequences into Unicode characters, as accepted by the bytes.decode()
 117         method.
 118         By default, percent-encoded sequences are decoded with UTF-8, and invalid
 119         sequences are replaced by a placeholder character.
 120
 121         unquote('abc%20def') -> 'abc def'.
 122         """
 123         if '%' not in string:
 124             string.split
 125             return string
 126         if encoding is None:
 127             encoding = 'utf-8'
 128         if errors is None:
 129             errors = 'replace'
 130         bits = _asciire.split(string)
 131         res = [bits[0]]
 132         append = res.append
 133         for i in range(1, len(bits), 2):
 134             append(compat_urllib_parse_unquote_to_bytes(bits[i]).decode(encoding, errors))
 135             append(bits[i + 1])
 136         return ''.join(res)
 137
 138     def compat_urllib_parse_unquote_plus(string, encoding='utf-8', errors='replace'):
 139         """Like unquote(), but also replace plus signs by spaces, as required for
 140         unquoting HTML form values.
 141
 142         unquote_plus('%7e/abc+def') -> '~/abc def'
 143         """
 144         string = string.replace('+', ' ')
 145         return compat_urllib_parse_unquote(string, encoding, errors)
 146
 147 try:
 148     compat_str = unicode  # Python 2
 149 except NameError:
 150     compat_str = str
 151
 152 try:
 153     compat_basestring = basestring  # Python 2
 154 except NameError:
 155     compat_basestring = str
 156
 157 try:
 158     compat_chr = unichr  # Python 2
 159 except NameError:
 160     compat_chr = chr
 161
 162 try:
 163     from xml.etree.ElementTree import ParseError as compat_xml_parse_error
 164 except ImportError:  # Python 2.6
 165     from xml.parsers.expat import ExpatError as compat_xml_parse_error
 166
 167
 168 try:
 169     from urllib.parse import parse_qs as compat_parse_qs
 170 except ImportError:  # Python 2
 171     # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib.
 172     # Python 2's version is apparently totally broken
 173
 174     def _parse_qsl(qs, keep_blank_values=False, strict_parsing=False,
 175                    encoding='utf-8', errors='replace'):
 176         qs, _coerce_result = qs, compat_str
 177         pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]
 178         r = []
 179         for name_value in pairs:
 180             if not name_value and not strict_parsing:
 181                 continue
 182             nv = name_value.split('=', 1)
 183             if len(nv) != 2:
 184                 if strict_parsing:
 185                     raise ValueError("bad query field: %r" % (name_value,))
 186                 # Handle case of a control-name with no equal sign
 187                 if keep_blank_values:
 188                     nv.append('')
 189                 else:
 190                     continue
 191             if len(nv[1]) or keep_blank_values:
 192                 name = nv[0].replace('+', ' ')
 193                 name = compat_urllib_parse_unquote(
 194                     name, encoding=encoding, errors=errors)
 195                 name = _coerce_result(name)
 196                 value = nv[1].replace('+', ' ')
 197                 value = compat_urllib_parse_unquote(
 198                     value, encoding=encoding, errors=errors)
 199                 value = _coerce_result(value)
 200                 r.append((name, value))
 201         return r
 202
 203     def compat_parse_qs(qs, keep_blank_values=False, strict_parsing=False,
 204                         encoding='utf-8', errors='replace'):
 205         parsed_result = {}
 206         pairs = _parse_qsl(qs, keep_blank_values, strict_parsing,
 207                            encoding=encoding, errors=errors)
 208         for name, value in pairs:
 209             if name in parsed_result:
 210                 parsed_result[name].append(value)
 211             else:
 212                 parsed_result[name] = [value]
 213         return parsed_result
 214
 215 try:
 216     from shlex import quote as shlex_quote
 217 except ImportError:  # Python < 3.3
 218     def shlex_quote(s):
 219         if re.match(r'^[-_\w./]+$', s):
 220             return s
 221         else:
 222             return "'" + s.replace("'", "'\"'\"'") + "'"
 223
 224
 225 def compat_ord(c):
 226     if type(c) is int:
 227         return c
 228     else:
 229         return ord(c)
 230
 231
 232 if sys.version_info >= (3, 0):
 233     compat_getenv = os.getenv
 234     compat_expanduser = os.path.expanduser
 235 else:
 236     # Environment variables should be decoded with filesystem encoding.
 237     # Otherwise it will fail if any non-ASCII characters present (see #3854 #3217 #2918)
 238
 239     def compat_getenv(key, default=None):
 240         from .utils import get_filesystem_encoding
 241         env = os.getenv(key, default)
 242         if env:
 243             env = env.decode(get_filesystem_encoding())
 244         return env
 245
 246     # HACK: The default implementations of os.path.expanduser from cpython do not decode
 247     # environment variables with filesystem encoding. We will work around this by
 248     # providing adjusted implementations.
 249     # The following are os.path.expanduser implementations from cpython 2.7.8 stdlib
 250     # for different platforms with correct environment variables decoding.
 251
 252     if os.name == 'posix':
 253         def compat_expanduser(path):
 254             """Expand ~ and ~user constructions.  If user or $HOME is unknown,
 255             do nothing."""
 256             if not path.startswith('~'):
 257                 return path
 258             i = path.find('/', 1)
 259             if i < 0:
 260                 i = len(path)
 261             if i == 1:
 262                 if 'HOME' not in os.environ:
 263                     import pwd
 264                     userhome = pwd.getpwuid(os.getuid()).pw_dir
 265                 else:
 266                     userhome = compat_getenv('HOME')
 267             else:
 268                 import pwd
 269                 try:
 270                     pwent = pwd.getpwnam(path[1:i])
 271                 except KeyError:
 272                     return path
 273                 userhome = pwent.pw_dir
 274             userhome = userhome.rstrip('/')
 275             return (userhome + path[i:]) or '/'
 276     elif os.name == 'nt' or os.name == 'ce':
 277         def compat_expanduser(path):
 278             """Expand ~ and ~user constructs.
 279
 280             If user or $HOME is unknown, do nothing."""
 281             if path[:1] != '~':
 282                 return path
 283             i, n = 1, len(path)
 284             while i < n and path[i] not in '/\\':
 285                 i = i + 1
 286
 287             if 'HOME' in os.environ:
 288                 userhome = compat_getenv('HOME')
 289             elif 'USERPROFILE' in os.environ:
 290                 userhome = compat_getenv('USERPROFILE')
 291             elif 'HOMEPATH' not in os.environ:
 292                 return path
 293             else:
 294                 try:
 295                     drive = compat_getenv('HOMEDRIVE')
 296                 except KeyError:
 297                     drive = ''
 298                 userhome = os.path.join(drive, compat_getenv('HOMEPATH'))
 299
 300             if i != 1:  # ~user
 301                 userhome = os.path.join(os.path.dirname(userhome), path[1:i])
 302
 303             return userhome + path[i:]
 304     else:
 305         compat_expanduser = os.path.expanduser
 306
 307
 308 if sys.version_info < (3, 0):
 309     def compat_print(s):
 310         from .utils import preferredencoding
 311         print(s.encode(preferredencoding(), 'xmlcharrefreplace'))
 312 else:
 313     def compat_print(s):
 314         assert isinstance(s, compat_str)
 315         print(s)
 316
 317
 318 try:
 319     subprocess_check_output = subprocess.check_output
 320 except AttributeError:
 321     def subprocess_check_output(*args, **kwargs):
 322         assert 'input' not in kwargs
 323         p = subprocess.Popen(*args, stdout=subprocess.PIPE, **kwargs)
 324         output, _ = p.communicate()
 325         ret = p.poll()
 326         if ret:
 327             raise subprocess.CalledProcessError(ret, p.args, output=output)
 328         return output
 329
 330 if sys.version_info < (3, 0) and sys.platform == 'win32':
 331     def compat_getpass(prompt, *args, **kwargs):
 332         if isinstance(prompt, compat_str):
 333             from .utils import preferredencoding
 334             prompt = prompt.encode(preferredencoding())
 335         return getpass.getpass(prompt, *args, **kwargs)
 336 else:
 337     compat_getpass = getpass.getpass
 338
 339 # Old 2.6 and 2.7 releases require kwargs to be bytes
 340 try:
 341     def _testfunc(x):
 342         pass
 343     _testfunc(**{'x': 0})
 344 except TypeError:
 345     def compat_kwargs(kwargs):
 346         return dict((bytes(k), v) for k, v in kwargs.items())
 347 else:
 348     compat_kwargs = lambda kwargs: kwargs
 349
 350
 351 if sys.version_info < (2, 7):
 352     def compat_socket_create_connection(address, timeout, source_address=None):
 353         host, port = address
 354         err = None
 355         for res in socket.getaddrinfo(host, port, 0, socket.SOCK_STREAM):
 356             af, socktype, proto, canonname, sa = res
 357             sock = None
 358             try:
 359                 sock = socket.socket(af, socktype, proto)
 360                 sock.settimeout(timeout)
 361                 if source_address:
 362                     sock.bind(source_address)
 363                 sock.connect(sa)
 364                 return sock
 365             except socket.error as _:
 366                 err = _
 367                 if sock is not None:
 368                     sock.close()
 369         if err is not None:
 370             raise err
 371         else:
 372             raise socket.error("getaddrinfo returns an empty list")
 373 else:
 374     compat_socket_create_connection = socket.create_connection
 375
 376
 377 # Fix https://github.com/rg3/youtube-dl/issues/4223
 378 # See http://bugs.python.org/issue9161 for what is broken
 379 def workaround_optparse_bug9161():
 380     op = optparse.OptionParser()
 381     og = optparse.OptionGroup(op, 'foo')
 382     try:
 383         og.add_option('-t')
 384     except TypeError:
 385         real_add_option = optparse.OptionGroup.add_option
 386
 387         def _compat_add_option(self, *args, **kwargs):
 388             enc = lambda v: (
 389                 v.encode('ascii', 'replace') if isinstance(v, compat_str)
 390                 else v)
 391             bargs = [enc(a) for a in args]
 392             bkwargs = dict(
 393                 (k, enc(v)) for k, v in kwargs.items())
 394             return real_add_option(self, *bargs, **bkwargs)
 395         optparse.OptionGroup.add_option = _compat_add_option
 396
 397 if hasattr(shutil, 'get_terminal_size'):  # Python >= 3.3
 398     compat_get_terminal_size = shutil.get_terminal_size
 399 else:
 400     _terminal_size = collections.namedtuple('terminal_size', ['columns', 'lines'])
 401
 402     def compat_get_terminal_size():
 403         columns = compat_getenv('COLUMNS', None)
 404         if columns:
 405             columns = int(columns)
 406         else:
 407             columns = None
 408         lines = compat_getenv('LINES', None)
 409         if lines:
 410             lines = int(lines)
 411         else:
 412             lines = None
 413
 414         try:
 415             sp = subprocess.Popen(
 416                 ['stty', 'size'],
 417                 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
 418             out, err = sp.communicate()
 419             lines, columns = map(int, out.split())
 420         except Exception:
 421             pass
 422         return _terminal_size(columns, lines)
 423
 424 try:
 425     itertools.count(start=0, step=1)
 426     compat_itertools_count = itertools.count
 427 except TypeError:  # Python 2.6
 428     def compat_itertools_count(start=0, step=1):
 429         n = start
 430         while True:
 431             yield n
 432             n += step
 433
 434 __all__ = [
 435     'compat_HTTPError',
 436     'compat_basestring',
 437     'compat_chr',
 438     'compat_cookiejar',
 439     'compat_expanduser',
 440     'compat_get_terminal_size',
 441     'compat_getenv',
 442     'compat_getpass',
 443     'compat_html_entities',
 444     'compat_http_client',
 445     'compat_http_server',
 446     'compat_itertools_count',
 447     'compat_kwargs',
 448     'compat_ord',
 449     'compat_parse_qs',
 450     'compat_print',
 451     'compat_socket_create_connection',
 452     'compat_str',
 453     'compat_subprocess_get_DEVNULL',
 454     'compat_urllib_error',
 455     'compat_urllib_parse',
 456     'compat_urllib_parse_unquote',
 457     'compat_urllib_parse_unquote_plus',
 458     'compat_urllib_parse_unquote_to_bytes',
 459     'compat_urllib_parse_urlparse',
 460     'compat_urllib_request',
 461     'compat_urlparse',
 462     'compat_urlretrieve',
 463     'compat_xml_parse_error',
 464     'shlex_quote',
 465     'subprocess_check_output',
 466     'workaround_optparse_bug9161',
 467 ]