@@ -15,7 +15,7 @@ wmain(int argc, wchar_t **argv)
15 15
}
16 16
#else
17 17 18 -
/* Helpers to better handle the legacy C locale
18 +
/* Access private pylifecycle helper API to better handle the legacy C locale
19 19
*
20 20
* The legacy C locale assumes ASCII as the default text encoding, which
21 21
* causes problems not only for the CPython runtime, but also other
@@ -27,107 +27,8 @@ wmain(int argc, wchar_t **argv)
27 27
* See the documentation of the PYTHONCOERCECLOCALE setting for more details.
28 28
*
29 29
*/
30 - 31 -
#ifdef PY_COERCE_C_LOCALE
32 -
/* Access private pylifecycle API to check PYTHONCOERCECLOCALE */
33 -
extern int _Py_CLocaleCoercionIsExpected(void);
34 - 35 -
static const char *_C_LOCALE_COERCION_WARNING =
36 -
"Python detected LC_CTYPE=C: %.20s coerced to %.20s (set another locale "
37 -
"or PYTHONCOERCECLOCALE=0 to disable this locale coercion behavior).\n";
38 - 39 -
typedef struct _CandidateLocale {
40 -
const char *locale_name;
41 -
int category;
42 -
} _LocaleCoercionTarget;
43 - 44 -
static _LocaleCoercionTarget _TARGET_LOCALES[] = {
45 -
{ "C.UTF-8", LC_ALL },
46 -
{ "C.utf8", LC_ALL },
47 -
{ "UTF-8", LC_CTYPE },
48 -
{ NULL, 0 }
49 -
};
50 - 51 -
void
52 -
_coerce_default_locale_settings(const _LocaleCoercionTarget *target)
53 -
{
54 -
const char *newloc = target->locale_name;
55 -
int category = target->category;
56 - 57 -
/* Reset locale back to currently configured defaults */
58 -
setlocale(LC_ALL, "");
59 - 60 -
/* Set the relevant locale environment variables */
61 -
if (category == LC_ALL) {
62 -
const char *env_vars_updated = "LC_ALL & LANG";
63 -
if (setenv("LC_ALL", newloc, 1)) {
64 -
fprintf(stderr,
65 -
"Error setting LC_ALL, skipping C locale coercion\n");
66 -
return;
67 -
}
68 -
if (setenv("LANG", newloc, 1)) {
69 -
fprintf(stderr,
70 -
"Error setting LANG during C locale coercion\n");
71 -
env_vars_updated = "LC_ALL";
72 -
}
73 -
fprintf(stderr, _C_LOCALE_COERCION_WARNING, env_vars_updated, newloc);
74 -
} else if (category == LC_CTYPE) {
75 -
if (setenv("LC_CTYPE", newloc, 1)) {
76 -
fprintf(stderr,
77 -
"Error setting LC_CTYPE, skipping C locale coercion\n");
78 -
return;
79 -
}
80 -
fprintf(stderr, _C_LOCALE_COERCION_WARNING, "LC_CTYPE", newloc);
81 -
} else {
82 -
fprintf(stderr, "Locale coercion must target LC_ALL or LC_CTYPE\n");
83 -
return;
84 -
}
85 - 86 -
/* Set standard stream encoding if PYTHONIOENCODING is not set
87 -
*
88 -
* We avoid setting PYTHONIOENCODING, as that can confuse Python 2
89 -
* instances in subprocesses that inherit the environment (as Python
90 -
* 2 has no 'surrogateescape' error handler).
91 -
*
92 -
* If PEP 540 is also implemented, this check will be replaced with
93 -
* unconditionally setting PYTHONUTF8=1
94 -
*/
95 -
const char *io_encoding = getenv("PYTHONIOENCODING");
96 -
if ((io_encoding == NULL) || (strnlen(io_encoding, 1) == 0)) {
97 -
Py_SetStandardStreamEncoding("utf-8", "surrogateescape");
98 -
}
99 - 100 -
/* Reconfigure with the overridden environment variables */
101 -
setlocale(LC_ALL, "");
102 -
}
103 - 104 -
void
105 -
_handle_legacy_c_locale(void)
106 -
{
107 -
/* We ignore the Python -E and -I flags here, as we need to sort out
108 -
* the locale settings *before* we try to do anything with the command
109 -
* line arguments. For cross-platform debugging purposes, we also need
110 -
* to give end users a way to force even scripts that are otherwise
111 -
* isolated from their environment to use the legacy ASCII-centric C
112 -
* locale.
113 -
*/
114 -
if (_Py_CLocaleCoercionIsExpected()) {
115 -
/* PYTHONCOERCECLOCALE is not set, or is not set to exactly "0" */
116 -
const _LocaleCoercionTarget *target = NULL;
117 -
for (target = _TARGET_LOCALES; target->locale_name; target++) {
118 -
const char *reconfigured_locale = setlocale(target->category,
119 -
target->locale_name);
120 -
if (reconfigured_locale != NULL) {
121 -
/* Successfully configured locale, so make it the default */
122 -
_coerce_default_locale_settings(target);
123 -
return;
124 -
}
125 -
}
126 - 127 -
}
128 -
/* No C locale warning here, as Py_Initialize will emit one later */
129 -
}
130 -
#endif
30 +
extern int _Py_LegacyLocaleDetected(void);
31 +
extern void _Py_CoerceLegacyLocale(void);
131 32 132 33
int
133 34
main(int argc, char **argv)
@@ -177,14 +78,9 @@ main(int argc, char **argv)
177 78
setlocale(LC_ALL, "");
178 79
#endif
179 80 180 -
#ifdef PY_COERCE_C_LOCALE
181 -
/* When the LC_CTYPE category still claims to be using the C locale,
182 -
assume configuration error and try for a UTF-8 based locale instead */
183 -
const char *ctype_loc = setlocale(LC_CTYPE, NULL);
184 -
if (ctype_loc != NULL && strcmp(ctype_loc, "C") == 0) {
185 -
_handle_legacy_c_locale();
81 +
if (_Py_LegacyLocaleDetected()) {
82 +
_Py_CoerceLegacyLocale();
186 83
}
187 -
#endif
188 84 189 85
/* Convert from char to wchar_t based on the locale settings */
190 86
for (i = 0; i < argc; i++) {
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4