bpo-28604: Fix localeconv() for different LC_MONETARY (GH-10606) (GH-10619) (GH-10621)

locale.localeconv() now sets temporarily the LC_CTYPE locale to the
LC_MONETARY locale if the two locales are different and monetary
strings are non-ASCII. This temporary change affects other threads.

Changes:

* locale.localeconv() can now set LC_CTYPE to LC_MONETARY to decode
  monetary fields.
* Add LocaleInfo.grouping_buffer: copy localeconv() grouping string
  since it can be replaced anytime if a different thread calls
  localeconv().

(cherry picked from commit 02e6bf7)
(cherry picked from commit 6eff6b8)

Author: vstinner

Doc/library/locale.rst

@@ -148,10 +148,8 @@ The :mod:`locale` module defines the following exception and functions:
    +--------------+-----------------------------------------+
 
    The function sets temporarily the ``LC_CTYPE`` locale to the ``LC_NUMERIC``
-   locale to decode ``decimal_point`` and ``thousands_sep`` byte strings if
-   they are non-ASCII or longer than 1 byte, and the ``LC_NUMERIC`` locale is
-   different than the ``LC_CTYPE`` locale. This temporary change affects other
-   threads.
+   locale or the ``LC_MONETARY`` locale if locales are different and numeric or
+   monetary strings are non-ASCII. This temporary change affects other threads.
 
    .. versionchanged:: 3.6.5
       The function now sets temporarily the ``LC_CTYPE`` locale to the

Misc/NEWS.d/next/Library/2018-11-20-13-34-01.bpo-28604.iiih5h.rst

@@ -0,0 +1,3 @@
+:func:`locale.localeconv` now sets temporarily the ``LC_CTYPE`` locale to the
+``LC_MONETARY`` locale if the two locales are different and monetary strings
+are non-ASCII. This temporary change affects other threads.

Modules/_localemodule.c

@@ -128,6 +128,82 @@ PyLocale_setlocale(PyObject* self, PyObject* args)
     return result_object;
 }
 
+static int
+locale_is_ascii(const char *str)
+{
+    return (strlen(str) == 1 && ((unsigned char)str[0]) <= 127);
+}
+
+static int
+locale_decode_monetary(PyObject *dict, struct lconv *lc)
+{
+    int change_locale;
+    change_locale = (!locale_is_ascii(lc->int_curr_symbol)
+                     || !locale_is_ascii(lc->currency_symbol)
+                     || !locale_is_ascii(lc->mon_decimal_point)
+                     || !locale_is_ascii(lc->mon_thousands_sep));
+
+    /* Keep a copy of the LC_CTYPE locale */
+    char *oldloc = NULL, *loc = NULL;
+    if (change_locale) {
+        oldloc = setlocale(LC_CTYPE, NULL);
+        if (!oldloc) {
+            PyErr_SetString(PyExc_RuntimeWarning,
+                            "failed to get LC_CTYPE locale");
+            return -1;
+        }
+
+        oldloc = _PyMem_Strdup(oldloc);
+        if (!oldloc) {
+            PyErr_NoMemory();
+            return -1;
+        }
+
+        loc = setlocale(LC_MONETARY, NULL);
+        if (loc != NULL && strcmp(loc, oldloc) == 0) {
+            loc = NULL;
+        }
+
+        if (loc != NULL) {
+            /* Only set the locale temporarily the LC_CTYPE locale
+               to the LC_MONETARY locale if the two locales are different and
+               at least one string is non-ASCII. */
+            setlocale(LC_CTYPE, loc);
+        }
+    }
+
+    int res = -1;
+
+#define RESULT_STRING(ATTR) \
+    do { \
+        PyObject *obj; \
+        obj = PyUnicode_DecodeLocale(lc->ATTR, NULL); \
+        if (obj == NULL) { \
+            goto done; \
+        } \
+        if (PyDict_SetItemString(dict, Py_STRINGIFY(ATTR), obj) < 0) { \
+            Py_DECREF(obj); \
+            goto done; \
+        } \
+        Py_DECREF(obj); \
+    } while (0)
+
+    RESULT_STRING(int_curr_symbol);
+    RESULT_STRING(currency_symbol);
+    RESULT_STRING(mon_decimal_point);
+    RESULT_STRING(mon_thousands_sep);
+#undef RESULT_STRING
+
+    res = 0;
+
+done:
+    if (loc != NULL) {
+        setlocale(LC_CTYPE, oldloc);
+    }
+    PyMem_Free(oldloc);
+    return res;
+}
+
 PyDoc_STRVAR(localeconv__doc__,
 "() -> dict. Returns numeric and monetary locale-specific parameters.");
 
@@ -171,11 +247,10 @@ PyLocale_localeconv(PyObject* self)
         RESULT(#i, x); \
     } while (0)
 
-    /* Monetary information */
-    RESULT_STRING(int_curr_symbol);
-    RESULT_STRING(currency_symbol);
-    RESULT_STRING(mon_decimal_point);
-    RESULT_STRING(mon_thousands_sep);
+    /* Monetary information: LC_MONETARY encoding */
+    if (locale_decode_monetary(result, l) < 0) {
+        goto failed;
+    }
     x = copy_grouping(l->mon_grouping);
     RESULT("mon_grouping", x);
 
@@ -190,7 +265,7 @@ PyLocale_localeconv(PyObject* self)
     RESULT_INT(p_sign_posn);
     RESULT_INT(n_sign_posn);
 
-    /* Numeric information */
+    /* Numeric information: LC_NUMERIC encoding */
     PyObject *decimal_point, *thousands_sep;
     const char *grouping;
     if (_Py_GetLocaleconvNumeric(&decimal_point,
@@ -220,6 +295,10 @@ PyLocale_localeconv(PyObject* self)
   failed:
     Py_XDECREF(result);
     return NULL;
+
+#undef RESULT
+#undef RESULT_STRING
+#undef RESULT_INT
 }
 
 #if defined(HAVE_WCSCOLL)

Python/fileutils.c

@@ -1687,7 +1687,7 @@ _Py_GetLocaleconvNumeric(PyObject **decimal_point, PyObject **thousands_sep,
     if (change_locale) {
         oldloc = setlocale(LC_CTYPE, NULL);
         if (!oldloc) {
-            PyErr_SetString(PyExc_RuntimeWarning, "faild to get LC_CTYPE locale");
+            PyErr_SetString(PyExc_RuntimeWarning, "failed to get LC_CTYPE locale");
             return -1;
         }
 
@@ -1703,7 +1703,7 @@ _Py_GetLocaleconvNumeric(PyObject **decimal_point, PyObject **thousands_sep,
         }
 
         if (loc != NULL) {
-            /* Only set the locale temporarilty the LC_CTYPE locale
+            /* Only set the locale temporarily the LC_CTYPE locale
                if LC_NUMERIC locale is different than LC_CTYPE locale and
                decimal_point and/or thousands_sep are non-ASCII or longer than
                1 byte */

Python/formatter_unicode.c

@@ -397,9 +397,10 @@ typedef struct {
     PyObject *decimal_point;
     PyObject *thousands_sep;
     const char *grouping;
+    char *grouping_buffer;
 } LocaleInfo;
 
-#define STATIC_LOCALE_INFO_INIT {0, 0, 0}
+#define STATIC_LOCALE_INFO_INIT {0, 0, 0, 0}
 
 /* describes the layout for an integer, see the comment in
    calc_number_widths() for details */
@@ -708,11 +709,22 @@ get_locale_info(enum LocaleType type, LocaleInfo *locale_info)
 {
     switch (type) {
     case LT_CURRENT_LOCALE: {
+        const char *grouping;
         if (_Py_GetLocaleconvNumeric(&locale_info->decimal_point,
                                      &locale_info->thousands_sep,
-                                     &locale_info->grouping) < 0) {
+                                     &grouping) < 0) {
             return -1;
         }
+
+        /* localeconv() grouping can become a dangling pointer or point
+           to a different string if another thread calls localeconv() during
+           the string formatting. Copy the string to avoid this risk. */
+        locale_info->grouping_buffer = _PyMem_Strdup(grouping);
+        if (locale_info->grouping_buffer == NULL) {
+            PyErr_NoMemory();
+            return -1;
+        }
+        locale_info->grouping = locale_info->grouping_buffer;
         break;
     }
     case LT_DEFAULT_LOCALE:
@@ -746,6 +758,7 @@ free_locale_info(LocaleInfo *locale_info)
 {
     Py_XDECREF(locale_info->decimal_point);
     Py_XDECREF(locale_info->thousands_sep);
+    PyMem_Free(locale_info->grouping_buffer);
 }
 
 /************************************************************************/