Coverage for /pythoncovmergedfiles/medio/medio/usr/local/lib/python3.11/site-packages/IPython/utils/encoding.py: 69%

1""" 

2Utilities for dealing with text encodings 

3""" 

4 

5from __future__ import annotations 

6 

7# ----------------------------------------------------------------------------- 

8# Copyright (C) 2008-2012 The IPython Development Team 

9# 

10# Distributed under the terms of the BSD License. The full license is in 

11# the file COPYING, distributed as part of this software. 

12# ----------------------------------------------------------------------------- 

13 

14# ----------------------------------------------------------------------------- 

15# Imports 

16# ----------------------------------------------------------------------------- 

17import sys 

18import locale 

19import warnings 

20from typing import Any 

21 

22 

23# to deal with the possibility of sys.std* not being a stream at all 

24def get_stream_enc(stream: Any, default: str | None = None) -> str | None: 

25 """Return the given stream's encoding or a default. 

26 

27 There are cases where ``sys.std*`` might not actually be a stream, so 

28 check for the encoding attribute prior to returning it, and return 

29 a default if it doesn't exist or evaluates as False. ``default`` 

30 is None if not provided. 

31 """ 

32 if not hasattr(stream, "encoding") or not stream.encoding: 

33 return default 

34 else: 

35 return stream.encoding 

36 

37 

38_sentinel: object = object() 

39 

40 

41# Less conservative replacement for sys.getdefaultencoding, that will try 

42# to match the environment. 

43# Defined here as central function, so if we find better choices, we 

44# won't need to make changes all over IPython. 

45def getdefaultencoding(prefer_stream: object | bool = _sentinel) -> str: 

46 """Return IPython's guess for the default encoding for bytes as text. 

47 

48 If prefer_stream is True (default), asks for stdin.encoding first, 

49 to match the calling Terminal, but that is often None for subprocesses. 

50 

51 Then fall back on locale.getpreferredencoding(), 

52 which should be a sensible platform default (that respects LANG environment), 

53 and finally to sys.getdefaultencoding() which is the most conservative option, 

54 and usually UTF8 as of Python 3. 

55 """ 

56 if prefer_stream is not _sentinel: 

57 warnings.warn( 

58 "getpreferredencoding(prefer_stream=) argument is deprecated since " 

59 "IPython 9.0, getdefaultencoding() will take no argument in the " 

60 "future. If you rely on `prefer_stream`, please open an issue on " 

61 "the IPython repo.", 

62 DeprecationWarning, 

63 stacklevel=2, 

64 ) 

65 prefer_stream = True 

66 enc: str | None = None 

67 if prefer_stream: 

68 enc = get_stream_enc(sys.stdin) 

69 if not enc or enc == "ascii": 

70 try: 

71 # There are reports of getpreferredencoding raising errors 

72 # in some cases, which may well be fixed, but let's be conservative here. 

73 enc = locale.getpreferredencoding() 

74 except Exception: 

75 pass 

76 enc = enc or sys.getdefaultencoding() 

77 # On windows `cp0` can be returned to indicate that there is no code page. 

78 # Since cp0 is an invalid encoding return instead cp1252 which is the 

79 # Western European default. 

80 if enc == "cp0": 

81 warnings.warn( 

82 "Invalid code page cp0 detected - using cp1252 instead." 

83 "If cp1252 is incorrect please ensure a valid code page " 

84 "is defined for the process.", 

85 RuntimeWarning, 

86 ) 

87 return "cp1252" 

88 return enc 

89 

90 

91DEFAULT_ENCODING = getdefaultencoding()