bpo-34485: Add _PyCoreConfig.stdio_encoding (GH-8881)
* Add stdio_encoding and stdio_errors fields to _PyCoreConfig.
* Add unit tests on stdio_encoding and stdio_errors.
diff --git a/Python/coreconfig.c b/Python/coreconfig.c
index 99d703c..00037d9 100644
--- a/Python/coreconfig.c
+++ b/Python/coreconfig.c
@@ -1,6 +1,9 @@
#include "Python.h"
#include "internal/pystate.h"
#include <locale.h>
+#ifdef HAVE_LANGINFO_H
+# include <langinfo.h>
+#endif
#define DECODE_LOCALE_ERR(NAME, LEN) \
@@ -89,8 +92,8 @@
* mechanism that attempts to figure out an appropriate IO encoding
*/
-char *_Py_StandardStreamEncoding = NULL;
-char *_Py_StandardStreamErrors = NULL;
+static char *_Py_StandardStreamEncoding = NULL;
+static char *_Py_StandardStreamErrors = NULL;
int
Py_SetStandardStreamEncoding(const char *encoding, const char *errors)
@@ -205,6 +208,9 @@
CLEAR(config->dll_path);
#endif
CLEAR(config->base_exec_prefix);
+
+ CLEAR(config->stdio_encoding);
+ CLEAR(config->stdio_errors);
#undef CLEAR
#undef CLEAR_WSTRLIST
}
@@ -216,6 +222,15 @@
_PyCoreConfig_Clear(config);
#define COPY_ATTR(ATTR) config->ATTR = config2->ATTR
+#define COPY_STR_ATTR(ATTR) \
+ do { \
+ if (config2->ATTR != NULL) { \
+ config->ATTR = _PyMem_RawStrdup(config2->ATTR); \
+ if (config->ATTR == NULL) { \
+ return -1; \
+ } \
+ } \
+ } while (0)
#define COPY_WSTR_ATTR(ATTR) \
do { \
if (config2->ATTR != NULL) { \
@@ -287,6 +302,8 @@
COPY_ATTR(quiet);
COPY_ATTR(user_site_directory);
COPY_ATTR(buffered_stdio);
+ COPY_STR_ATTR(stdio_encoding);
+ COPY_STR_ATTR(stdio_errors);
#ifdef MS_WINDOWS
COPY_ATTR(legacy_windows_fs_encoding);
COPY_ATTR(legacy_windows_stdio);
@@ -932,6 +949,161 @@
}
+static const char *
+get_stdio_errors(const _PyCoreConfig *config)
+{
+#ifndef MS_WINDOWS
+ const char *loc = setlocale(LC_CTYPE, NULL);
+ if (loc != NULL) {
+ /* surrogateescape is the default in the legacy C and POSIX locales */
+ if (strcmp(loc, "C") == 0 || strcmp(loc, "POSIX") == 0) {
+ return "surrogateescape";
+ }
+
+#ifdef PY_COERCE_C_LOCALE
+ /* surrogateescape is the default in locale coercion target locales */
+ if (_Py_IsLocaleCoercionTarget(loc)) {
+ return "surrogateescape";
+ }
+#endif
+ }
+
+ return "strict";
+#else
+ /* On Windows, always use surrogateescape by default */
+ return "surrogateescape";
+#endif
+}
+
+
+_PyInitError
+_Py_get_locale_encoding(char **locale_encoding)
+{
+#ifdef MS_WINDOWS
+ char encoding[20];
+ PyOS_snprintf(encoding, sizeof(encoding), "cp%d", GetACP());
+#elif defined(__ANDROID__)
+ const char *encoding = "UTF-8";
+#else
+ const char *encoding = nl_langinfo(CODESET);
+ if (!encoding || encoding[0] == '\0') {
+ return _Py_INIT_USER_ERR("failed to get the locale encoding: "
+ "nl_langinfo(CODESET) failed");
+ }
+#endif
+ *locale_encoding = _PyMem_RawStrdup(encoding);
+ if (*locale_encoding == NULL) {
+ return _Py_INIT_NO_MEMORY();
+ }
+ return _Py_INIT_OK();
+}
+
+
+static _PyInitError
+config_init_stdio_encoding(_PyCoreConfig *config)
+{
+ /* If Py_SetStandardStreamEncoding() have been called, use these
+ parameters. */
+ if (config->stdio_encoding == NULL && _Py_StandardStreamEncoding != NULL) {
+ config->stdio_encoding = _PyMem_RawStrdup(_Py_StandardStreamEncoding);
+ if (config->stdio_encoding == NULL) {
+ return _Py_INIT_NO_MEMORY();
+ }
+ }
+
+ if (config->stdio_errors == NULL && _Py_StandardStreamErrors != NULL) {
+ config->stdio_errors = _PyMem_RawStrdup(_Py_StandardStreamErrors);
+ if (config->stdio_errors == NULL) {
+ return _Py_INIT_NO_MEMORY();
+ }
+ }
+
+ if (config->stdio_encoding != NULL && config->stdio_errors != NULL) {
+ return _Py_INIT_OK();
+ }
+
+ /* PYTHONIOENCODING environment variable */
+ const char *opt = _PyCoreConfig_GetEnv(config, "PYTHONIOENCODING");
+ if (opt) {
+ char *pythonioencoding = _PyMem_RawStrdup(opt);
+ if (pythonioencoding == NULL) {
+ return _Py_INIT_NO_MEMORY();
+ }
+
+ char *err = strchr(pythonioencoding, ':');
+ if (err) {
+ *err = '\0';
+ err++;
+ if (!err[0]) {
+ err = NULL;
+ }
+ }
+
+ /* Does PYTHONIOENCODING contain an encoding? */
+ if (pythonioencoding[0]) {
+ if (config->stdio_encoding == NULL) {
+ config->stdio_encoding = _PyMem_RawStrdup(pythonioencoding);
+ if (config->stdio_encoding == NULL) {
+ PyMem_RawFree(pythonioencoding);
+ return _Py_INIT_NO_MEMORY();
+ }
+ }
+
+ /* If the encoding is set but not the error handler,
+ use "strict" error handler by default.
+ PYTHONIOENCODING=latin1 behaves as
+ PYTHONIOENCODING=latin1:strict. */
+ if (!err) {
+ err = "strict";
+ }
+ }
+
+ if (config->stdio_errors == NULL && err != NULL) {
+ config->stdio_errors = _PyMem_RawStrdup(err);
+ if (config->stdio_errors == NULL) {
+ PyMem_RawFree(pythonioencoding);
+ return _Py_INIT_NO_MEMORY();
+ }
+ }
+
+ PyMem_RawFree(pythonioencoding);
+ }
+
+ /* UTF-8 Mode uses UTF-8/surrogateescape */
+ if (config->utf8_mode) {
+ if (config->stdio_encoding == NULL) {
+ config->stdio_encoding = _PyMem_RawStrdup("utf-8");
+ if (config->stdio_encoding == NULL) {
+ return _Py_INIT_NO_MEMORY();
+ }
+ }
+ if (config->stdio_errors == NULL) {
+ config->stdio_errors = _PyMem_RawStrdup("surrogateescape");
+ if (config->stdio_errors == NULL) {
+ return _Py_INIT_NO_MEMORY();
+ }
+ }
+ }
+
+ /* Choose the default error handler based on the current locale. */
+ if (config->stdio_encoding == NULL) {
+ _PyInitError err = _Py_get_locale_encoding(&config->stdio_encoding);
+ if (_Py_INIT_FAILED(err)) {
+ return err;
+ }
+ }
+ if (config->stdio_errors == NULL) {
+ const char *errors = get_stdio_errors(config);
+ config->stdio_errors = _PyMem_RawStrdup(errors);
+ if (config->stdio_errors == NULL) {
+ return _Py_INIT_NO_MEMORY();
+ }
+ }
+
+ return _Py_INIT_OK();
+}
+
+
/* Read configuration settings from standard locations
*
* This function doesn't make any changes to the interpreter state - it
@@ -1044,6 +1216,11 @@
config->argc = 0;
}
+ err = config_init_stdio_encoding(config);
+ if (_Py_INIT_FAILED(err)) {
+ return err;
+ }
+
assert(config->coerce_c_locale >= 0);
assert(config->use_environment >= 0);
diff --git a/Python/pylifecycle.c b/Python/pylifecycle.c
index 88403f4..9f6757f 100644
--- a/Python/pylifecycle.c
+++ b/Python/pylifecycle.c
@@ -184,27 +184,6 @@
return NULL;
}
-static _PyInitError
-get_locale_encoding(char **locale_encoding)
-{
-#ifdef MS_WINDOWS
- char encoding[20];
- PyOS_snprintf(encoding, sizeof(encoding), "cp%d", GetACP());
-#elif defined(__ANDROID__)
- const char *encoding = "UTF-8";
-#else
- const char *encoding = nl_langinfo(CODESET);
- if (!encoding || encoding[0] == '\0') {
- return _Py_INIT_USER_ERR("failed to get the locale encoding: "
- "nl_langinfo(CODESET) failed");
- }
-#endif
- *locale_encoding = _PyMem_RawStrdup(encoding);
- if (*locale_encoding == NULL) {
- return _Py_INIT_NO_MEMORY();
- }
- return _Py_INIT_OK();
-}
static _PyInitError
initimport(PyInterpreterState *interp, PyObject *sysmod)
@@ -340,35 +319,20 @@
{NULL}
};
-static const char *
-get_stdio_errors(void)
+
+int
+_Py_IsLocaleCoercionTarget(const char *ctype_loc)
{
-#ifndef MS_WINDOWS
- const char *ctype_loc = setlocale(LC_CTYPE, NULL);
- if (ctype_loc != NULL) {
- /* surrogateescape is the default in the legacy C and POSIX locales */
- if (strcmp(ctype_loc, "C") == 0 || strcmp(ctype_loc, "POSIX") == 0) {
- return "surrogateescape";
+ const _LocaleCoercionTarget *target = NULL;
+ for (target = _TARGET_LOCALES; target->locale_name; target++) {
+ if (strcmp(ctype_loc, target->locale_name) == 0) {
+ return 1;
}
-
-#ifdef PY_COERCE_C_LOCALE
- /* surrogateescape is the default in locale coercion target locales */
- const _LocaleCoercionTarget *target = NULL;
- for (target = _TARGET_LOCALES; target->locale_name; target++) {
- if (strcmp(ctype_loc, target->locale_name) == 0) {
- return "surrogateescape";
- }
- }
-#endif
}
-
- return "strict";
-#else
- /* On Windows, always use surrogateescape by default */
- return "surrogateescape";
-#endif
+ return 0;
}
+
#ifdef PY_COERCE_C_LOCALE
static const char C_LOCALE_COERCION_WARNING[] =
"Python detected LC_CTYPE=C: LC_CTYPE coerced to %.20s (set another locale "
@@ -1533,8 +1497,10 @@
Py_HasFileSystemDefaultEncoding = 1;
}
else {
+ extern _PyInitError _Py_get_locale_encoding(char **locale_encoding);
+
char *locale_encoding;
- _PyInitError err = get_locale_encoding(&locale_encoding);
+ _PyInitError err = _Py_get_locale_encoding(&locale_encoding);
if (_Py_INIT_FAILED(err)) {
return err;
}
@@ -1740,13 +1706,16 @@
PyObject *std = NULL;
int fd;
PyObject * encoding_attr;
- char *pythonioencoding = NULL;
- const char *encoding, *errors;
- char *locale_encoding = NULL;
- char *codec_name = NULL;
_PyInitError res = _Py_INIT_OK();
- extern char *_Py_StandardStreamEncoding;
- extern char *_Py_StandardStreamErrors;
+ _PyCoreConfig *config = &interp->core_config;
+
+ char *codec_name = get_codec_name(config->stdio_encoding);
+ if (codec_name == NULL) {
+ return _Py_INIT_ERR("failed to get the Python codec name "
+ "of the stdio encoding");
+ }
+ PyMem_RawFree(config->stdio_encoding);
+ config->stdio_encoding = codec_name;
/* Hack to avoid a nasty recursion issue when Python is invoked
in verbose mode: pre-import the Latin-1 and UTF-8 codecs */
@@ -1778,85 +1747,15 @@
}
Py_DECREF(wrapper);
- encoding = _Py_StandardStreamEncoding;
- errors = _Py_StandardStreamErrors;
- if (!encoding || !errors) {
- char *opt = Py_GETENV("PYTHONIOENCODING");
- if (opt && opt[0] != '\0') {
- char *err;
- pythonioencoding = _PyMem_Strdup(opt);
- if (pythonioencoding == NULL) {
- PyErr_NoMemory();
- goto error;
- }
- err = strchr(pythonioencoding, ':');
- if (err) {
- *err = '\0';
- err++;
- if (!err[0]) {
- err = NULL;
- }
- }
-
- /* Does PYTHONIOENCODING contain an encoding? */
- if (pythonioencoding[0]) {
- if (!encoding) {
- encoding = pythonioencoding;
- }
-
- /* If the encoding is set but not the error handler,
- use "strict" error handler by default.
- PYTHONIOENCODING=latin1 behaves as
- PYTHONIOENCODING=latin1:strict. */
- if (!err) {
- err = "strict";
- }
- }
-
- if (!errors && err != NULL) {
- errors = err;
- }
- }
-
- if (interp->core_config.utf8_mode) {
- if (!encoding) {
- encoding = "utf-8";
- }
- if (!errors) {
- errors = "surrogateescape";
- }
- }
-
- if (!errors) {
- /* Choose the default error handler based on the current locale */
- errors = get_stdio_errors();
- }
- }
-
- if (encoding == NULL) {
- _PyInitError err = get_locale_encoding(&locale_encoding);
- if (_Py_INIT_FAILED(err)) {
- return err;
- }
- encoding = locale_encoding;
- }
-
- codec_name = get_codec_name(encoding);
- if (codec_name == NULL) {
- PyErr_SetString(PyExc_RuntimeError,
- "failed to get the Python codec name "
- "of stdio encoding");
- goto error;
- }
- encoding = codec_name;
-
/* Set sys.stdin */
fd = fileno(stdin);
/* Under some conditions stdin, stdout and stderr may not be connected
* and fileno() may point to an invalid file descriptor. For example
* GUI apps don't have valid standard streams by default.
*/
- std = create_stdio(iomod, fd, 0, "<stdin>", encoding, errors);
+ std = create_stdio(iomod, fd, 0, "<stdin>",
+ config->stdio_encoding,
+ config->stdio_errors);
if (std == NULL)
goto error;
PySys_SetObject("__stdin__", std);
@@ -1865,7 +1764,9 @@
/* Set sys.stdout */
fd = fileno(stdout);
- std = create_stdio(iomod, fd, 1, "<stdout>", encoding, errors);
+ std = create_stdio(iomod, fd, 1, "<stdout>",
+ config->stdio_encoding,
+ config->stdio_errors);
if (std == NULL)
goto error;
PySys_SetObject("__stdout__", std);
@@ -1875,7 +1776,9 @@
#if 1 /* Disable this if you have trouble debugging bootstrap stuff */
/* Set sys.stderr, replaces the preliminary stderr */
fd = fileno(stderr);
- std = create_stdio(iomod, fd, 1, "<stderr>", encoding, "backslashreplace");
+ std = create_stdio(iomod, fd, 1, "<stderr>",
+ config->stdio_encoding,
+ "backslashreplace");
if (std == NULL)
goto error;
@@ -1911,9 +1814,6 @@
done:
_Py_ClearStandardStreamEncoding();
- PyMem_RawFree(locale_encoding);
- PyMem_RawFree(codec_name);
- PyMem_Free(pythonioencoding);
Py_XDECREF(bimod);
Py_XDECREF(iomod);
return res;