Use wcscoll for _locale.strcoll if available.

5 files changed, 92 insertions, 8 deletions

diff --git a/Modules/_localemodule.c b/Modules/_localemodule.c
index d535ee8..2077be8 100644
--- a/Modules/_localemodule.c
+++ b/Modules/_localemodule.c
@@ -25,6 +25,10 @@ This software comes with no warranty. Use at your own risk.
 #include <libintl.h>
 #endif
 
+#ifdef HAVE_WCHAR_H
+#include <wchar.h>
+#endif
+
 #if defined(MS_WINDOWS)
 #define WIN32_LEAN_AND_MEAN
 #include <windows.h>
@@ -325,13 +329,76 @@ PyDoc_STRVAR(strcoll__doc__,
 static PyObject*
 PyLocale_strcoll(PyObject* self, PyObject* args)
 {
-  char *s1,*s2;
-
-  if (!PyArg_ParseTuple(args, "ss:strcoll", &s1, &s2))
-      return NULL;
-  return PyInt_FromLong(strcoll(s1, s2));
+#if !defined(HAVE_WCSCOLL) || !defined(Py_USING_UNICODE)
+    char *s1,*s2;
+    
+    if (!PyArg_ParseTuple(args, "ss:strcoll", &s1, &s2))
+        return NULL;
+    return PyInt_FromLong(strcoll(s1, s2));
+#else
+    PyObject *os1, *os2, *result = NULL;
+    wchar_t *ws1 = NULL, *ws2 = NULL;
+    int rel1 = 0, rel2 = 0, len1, len2;
+    
+    if (!PyArg_ParseTuple(args, "OO:strcoll", &os1, &os2))
+        return NULL;
+    /* If both arguments are byte strings, use strcoll.  */
+    if (PyString_Check(os1) && PyString_Check(os2))
+        return PyInt_FromLong(strcoll(PyString_AS_STRING(os1),
+                                      PyString_AS_STRING(os2)));
+    /* If neither argument is unicode, it's an error.  */
+    if (!PyUnicode_Check(os1) && !PyUnicode_Check(os2)) {
+        PyErr_SetString(PyExc_ValueError, "strcoll arguments must be strings");
+    }
+    /* Convert the non-unicode argument to unicode. */
+    if (!PyUnicode_Check(os1)) {
+        os1 = PyUnicode_FromObject(os1);
+        if (!os1)
+            return NULL;
+        rel1 = 1;
+    }
+    if (!PyUnicode_Check(os2)) {
+        os2 = PyUnicode_FromObject(os2);
+        if (!os2) {
+            Py_DECREF(os1);
+            return NULL;
+        } 
+        rel2 = 1;
+    }
+    /* Convert the unicode strings to wchar[]. */
+    len1 = PyUnicode_GET_SIZE(os1) + 1;
+    len2 = PyUnicode_GET_SIZE(os2) + 1;
+    ws1 = PyMem_MALLOC(len1 * sizeof(wchar_t));
+    if (!ws1) {
+        PyErr_NoMemory();
+        goto done;
+    }
+    if (PyUnicode_AsWideChar((PyUnicodeObject*)os1, ws1, len1) == -1)
+        goto done;
+    ws2 = PyMem_MALLOC(len2 * sizeof(wchar_t));
+    if (!ws2) {
+        PyErr_NoMemory();
+        goto done;
+    }
+    if (PyUnicode_AsWideChar((PyUnicodeObject*)os2, ws2, len2) == -1)
+        goto done;
+    /* Collate the strings. */
+    result = PyInt_FromLong(wcscoll(ws1, ws2));
+  done:
+    /* Deallocate everything. */
+    if (ws1) PyMem_FREE(ws1);
+    if (ws2) PyMem_FREE(ws2);
+    if (rel1) {
+        Py_DECREF(os1);
+    }
+    if (rel2) {
+        Py_DECREF(os2);
+    }
+    return result;
+#endif
 }
 
+
 PyDoc_STRVAR(strxfrm__doc__,
 "string -> string. Returns a string that behaves for cmp locale-aware.");
 
@@ -709,3 +776,10 @@ init_locale(void)
     }
 #endif
 }
+
+/* 
+Local variables:
+c-basic-offset: 4
+indent-tabs-mode: nil
+End:
+*/
diff --git a/PC/pyconfig.h b/PC/pyconfig.h
index 6ff99b6..0e7912e 100644
--- a/PC/pyconfig.h
+++ b/PC/pyconfig.h
@@ -497,6 +497,9 @@ Py_NO_ENABLE_SHARED to find out.  Also support MS_NO_COREDLL for b/w compat */
 /* Define if you have waitpid.  */
 /* #undef HAVE_WAITPID */
 
+/* Define to 1 if you have the `wcscoll' function. */
+#define HAVE_WCSCOLL 1
+
 /* Define if you have the <dlfcn.h> header file.  */
 /* #undef HAVE_DLFCN_H */
 
@@ -551,6 +554,9 @@ Py_NO_ENABLE_SHARED to find out.  Also support MS_NO_COREDLL for b/w compat */
 /* Define if you have the <utime.h> header file.  */
 /* #define HAVE_UTIME_H 1 */
 
+/* Define if the compiler provides a wchar.h header file. */
+#define HAVE_WCHAR_H 1
+
 /* Define if you have the dl library (-ldl).  */
 /* #undef HAVE_LIBDL */
 
diff --git a/configure b/configure
index b825d15..0103ba1 100755
--- a/configure
+++ b/configure
@@ -1,5 +1,5 @@
 #! /bin/sh
-# From configure.in Revision: 1.376 .
+# From configure.in Revision: 1.377 .
 # Guess values for system-dependent variables and create Makefiles.
 # Generated by GNU Autoconf 2.53 for python 2.3.
 #
@@ -12008,6 +12008,7 @@ echo "${ECHO_T}MACHDEP_OBJS" >&6
 
 
 
+
 for ac_func in alarm chown clock confstr ctermid execv \
  fchdir flock fork fsync fdatasync fpathconf ftime ftruncate \
  gai_strerror getgroups getlogin getpeername getpgid getpid getpwent getwd \
@@ -12018,7 +12019,7 @@ for ac_func in alarm chown clock confstr ctermid execv \
  setlocale setregid setreuid setsid setpgid setuid setvbuf snprintf \
  sigaction siginterrupt sigrelse strftime strptime \
  sysconf tcgetpgrp tcsetpgrp tempnam timegm times tmpfile tmpnam tmpnam_r \
- truncate uname unsetenv utimes waitpid _getpty getpriority
+ truncate uname unsetenv utimes waitpid wcscoll _getpty getpriority
 do
 as_ac_var=`echo "ac_cv_func_$ac_func" | $as_tr_sh`
 echo "$as_me:$LINENO: checking for $ac_func" >&5
diff --git a/configure.in b/configure.in
index a3b8c67..b6a266d 100644
--- a/configure.in
+++ b/configure.in
@@ -1770,7 +1770,7 @@ AC_CHECK_FUNCS(alarm chown clock confstr ctermid execv \
  setlocale setregid setreuid setsid setpgid setuid setvbuf snprintf \
  sigaction siginterrupt sigrelse strftime strptime \
  sysconf tcgetpgrp tcsetpgrp tempnam timegm times tmpfile tmpnam tmpnam_r \
- truncate uname unsetenv utimes waitpid _getpty getpriority)
+ truncate uname unsetenv utimes waitpid wcscoll _getpty getpriority)
 
 # For some functions, having a definition is not sufficient, since
 # we want to take their address.
diff --git a/pyconfig.h.in b/pyconfig.h.in
index 9503ea8..d63c36c 100644
--- a/pyconfig.h.in
+++ b/pyconfig.h.in
@@ -586,6 +586,9 @@
 /* Define if the compiler provides a wchar.h header file. */
 #undef HAVE_WCHAR_H
 
+/* Define to 1 if you have the `wcscoll' function. */
+#undef HAVE_WCSCOLL
+
 /* Define to 1 if you have the `_getpty' function. */
 #undef HAVE__GETPTY