bpo-28180: Fix the implementation of PEP 538 on Android (GH-4334)

Include/pylifecycle.h

@@ -137,6 +137,7 @@ PyAPI_FUNC(int) _PyOS_URandomNonblock(void *buffer, Py_ssize_t size);
 #ifndef Py_LIMITED_API
 PyAPI_FUNC(void) _Py_CoerceLegacyLocale(void);
 PyAPI_FUNC(int) _Py_LegacyLocaleDetected(void);
+PyAPI_FUNC(char *) _Py_SetLocaleFromEnv(int category);
 #endif
 
 #ifdef __cplusplus

Lib/test/test_c_locale_coercion.py

@@ -6,7 +6,6 @@ import os
 import sys
 import sysconfig
 import shutil
-import subprocess
 from collections import namedtuple
 
 import test.support
@@ -18,9 +17,12 @@ from test.support.script_helper import (
 # Set our expectation for the default encoding used in the C locale
 # for the filesystem encoding and the standard streams
 
-# AIX uses iso8859-1 in the C locale, other *nix platforms use ASCII
+# While most *nix platforms default to ASCII in the C locale, some use a
+# different encoding.
 if sys.platform.startswith("aix"):
     C_LOCALE_STREAM_ENCODING = "iso8859-1"
+elif test.support.is_android:
+    C_LOCALE_STREAM_ENCODING = "utf-8"
 else:
     C_LOCALE_STREAM_ENCODING = "ascii"
 
@@ -301,6 +303,19 @@ class LocaleCoercionTests(_LocaleHandlingTestCase):
                 # See https://bugs.python.org/issue30672 for discussion
                 if locale_to_set == "POSIX":
                     continue
+
+                # Platforms using UTF-8 in the C locale do not print
+                # CLI_COERCION_WARNING when all the locale envt variables are
+                # not set or set to the empty string.
+                _expected_warnings = expected_warnings
+                for _env_var in base_var_dict:
+                    if base_var_dict[_env_var]:
+                        break
+                else:
+                    if (C_LOCALE_STREAM_ENCODING == "utf-8" and
+                           locale_to_set == "" and coerce_c_locale == "warn"):
+                        _expected_warnings = None
+
                 with self.subTest(env_var=env_var,
                                   nominal_locale=locale_to_set,
                                   PYTHONCOERCECLOCALE=coerce_c_locale):
@@ -312,7 +327,7 @@ class LocaleCoercionTests(_LocaleHandlingTestCase):
                     self._check_child_encoding_details(var_dict,
                                                        fs_encoding,
                                                        stream_encoding,
-                                                       expected_warnings,
+                                                       _expected_warnings,
                                                        coercion_expected)
 
     def test_test_PYTHONCOERCECLOCALE_not_set(self):

Misc/NEWS.d/next/Core and Builtins/2017-11-12-11-44-22.bpo-28180.HQX000.rst

+A new internal ``_Py_SetLocaleFromEnv(category)`` helper function has been
+added in order to improve the consistency of behaviour across different
+``libc`` implementations (e.g. Android doesn't support setting the locale from
+the environment by default).

Modules/readline.c

@@ -1245,7 +1245,7 @@ call_readline(FILE *sys_stdin, FILE *sys_stdout, const char *prompt)
     char *saved_locale = strdup(setlocale(LC_CTYPE, NULL));
     if (!saved_locale)
         Py_FatalError("not enough memory to save locale");
-    setlocale(LC_CTYPE, "");
+    _Py_SetLocaleFromEnv(LC_CTYPE);
 #endif
 
     if (sys_stdin != rl_instream || sys_stdout != rl_outstream) {

Programs/python.c

@@ -54,15 +54,8 @@ main(int argc, char **argv)
         return 1;
     }
 
-#ifdef __ANDROID__
-    /* Passing "" to setlocale() on Android requests the C locale rather
-     * than checking environment variables, so request C.UTF-8 explicitly
-     */
-    setlocale(LC_ALL, "C.UTF-8");
-#else
     /* Reconfigure the locale to the default for this process */
-    setlocale(LC_ALL, "");
-#endif
+    _Py_SetLocaleFromEnv(LC_ALL);
 
     /* The legacy C locale assumes ASCII as the default text encoding, which
      * causes problems not only for the CPython runtime, but also other

Python/pylifecycle.c

@@ -459,7 +459,7 @@ _coerce_default_locale_settings(const _LocaleCoercionTarget *target)
     const char *newloc = target->locale_name;
 
     /* Reset locale back to currently configured defaults */
-    setlocale(LC_ALL, "");
+    _Py_SetLocaleFromEnv(LC_ALL);
 
     /* Set the relevant locale environment variable */
     if (setenv("LC_CTYPE", newloc, 1)) {
@@ -472,7 +472,7 @@ _coerce_default_locale_settings(const _LocaleCoercionTarget *target)
     }
 
     /* Reconfigure with the overridden environment variables */
-    setlocale(LC_ALL, "");
+    _Py_SetLocaleFromEnv(LC_ALL);
 }
 #endif
 
@@ -503,13 +503,14 @@ _Py_CoerceLegacyLocale(void)
                 const char *new_locale = setlocale(LC_CTYPE,
                                                    target->locale_name);
                 if (new_locale != NULL) {
-#if !defined(__APPLE__) && defined(HAVE_LANGINFO_H) && defined(CODESET)
+#if !defined(__APPLE__) && !defined(__ANDROID__) && \
+    defined(HAVE_LANGINFO_H) && defined(CODESET)
                     /* Also ensure that nl_langinfo works in this locale */
                     char *codeset = nl_langinfo(CODESET);
                     if (!codeset || *codeset == '\0') {
                         /* CODESET is not set or empty, so skip coercion */
                         new_locale = NULL;
-                        setlocale(LC_CTYPE, "");
+                        _Py_SetLocaleFromEnv(LC_CTYPE);
                         continue;
                     }
 #endif
@@ -524,6 +525,65 @@ _Py_CoerceLegacyLocale(void)
 #endif
 }
 
+/* _Py_SetLocaleFromEnv() is a wrapper around setlocale(category, "") to
+ * isolate the idiosyncrasies of different libc implementations. It reads the
+ * appropriate environment variable and uses its value to select the locale for
+ * 'category'. */
+char *
+_Py_SetLocaleFromEnv(int category)
+{
+#ifdef __ANDROID__
+    const char *locale;
+    const char **pvar;
+#ifdef PY_COERCE_C_LOCALE
+    const char *coerce_c_locale;
+#endif
+    const char *utf8_locale = "C.UTF-8";
+    const char *env_var_set[] = {
+        "LC_ALL",
+        "LC_CTYPE",
+        "LANG",
+        NULL,
+    };
+
+    /* Android setlocale(category, "") doesn't check the environment variables
+     * and incorrectly sets the "C" locale at API 24 and older APIs. We only
+     * check the environment variables listed in env_var_set. */
+    for (pvar=env_var_set; *pvar; pvar++) {
+        locale = getenv(*pvar);
+        if (locale != NULL && *locale != '\0') {
+            if (strcmp(locale, utf8_locale) == 0 ||
+                    strcmp(locale, "en_US.UTF-8") == 0) {
+                return setlocale(category, utf8_locale);
+            }
+            return setlocale(category, "C");
+        }
+    }
+
+    /* Android uses UTF-8, so explicitly set the locale to C.UTF-8 if none of
+     * LC_ALL, LC_CTYPE, or LANG is set to a non-empty string.
+     * Quote from POSIX section "8.2 Internationalization Variables":
+     * "4. If the LANG environment variable is not set or is set to the empty
+     * string, the implementation-defined default locale shall be used." */
+
+#ifdef PY_COERCE_C_LOCALE
+    coerce_c_locale = getenv("PYTHONCOERCECLOCALE");
+    if (coerce_c_locale == NULL || strcmp(coerce_c_locale, "0") != 0) {
+        /* Some other ported code may check the environment variables (e.g. in
+         * extension modules), so we make sure that they match the locale
+         * configuration */
+        if (setenv("LC_CTYPE", utf8_locale, 1)) {
+            fprintf(stderr, "Warning: failed setting the LC_CTYPE "
+                            "environment variable to %s\n", utf8_locale);
+        }
+    }
+#endif
+    return setlocale(category, utf8_locale);
+#else /* __ANDROID__ */
+    return setlocale(category, "");
+#endif /* __ANDROID__ */
+}
+
 
 /* Global initializations.  Can be undone by Py_Finalize().  Don't
    call this twice without an intervening Py_Finalize() call.
@@ -599,19 +659,12 @@ void _Py_InitializeCore(const _PyCoreConfig *config)
         exit(1);
     }
 
-#ifdef __ANDROID__
-    /* Passing "" to setlocale() on Android requests the C locale rather
-     * than checking environment variables, so request C.UTF-8 explicitly
-     */
-    setlocale(LC_CTYPE, "C.UTF-8");
-#else
 #ifndef MS_WINDOWS
     /* Set up the LC_CTYPE locale, so we can obtain
        the locale's charset without having to switch
        locales. */
-    setlocale(LC_CTYPE, "");
+    _Py_SetLocaleFromEnv(LC_CTYPE);
     _emit_stderr_warning_for_legacy_locale();
-#endif
 #endif
 
     if ((p = Py_GETENV("PYTHONDEBUG")) && *p != '\0')