use Py_UNICODE_WIDE instead of USE_UCS4_STORAGE and Py_UNICODE_SIZE

tests.

Include/unicodeobject.h

@@ -66,10 +66,11 @@ Copyright (c) Corporation for National Research Initiatives.
 #error Must define Py_UNICODE_SIZE
 #endif
 
-/* experimental UCS-4 support.  enable at your own risk! */
-#undef USE_UCS4_STORAGE
-#if Py_UNICODE_SIZE == 4
-#define USE_UCS4_STORAGE
+/* Setting Py_UNICODE_WIDE enables UCS-4 storage.  Otherwise, Unicode
+   strings are stored as UCS-2 (with limited support for UTF-16) */
+
+#if Py_UNICODE_SIZE >= 4
+#define Py_UNICODE_WIDE
 #endif
 
 /* Set these flags if the platform has "wchar.h", "wctype.h" and the
@@ -81,12 +82,12 @@ Copyright (c) Corporation for National Research Initiatives.
 #ifndef PY_UNICODE_TYPE
 
 /* Windows has a usable wchar_t type (unless we're using UCS-4) */
-# if defined(MS_WIN32) && !defined(USE_UCS4_STORAGE)
+# if defined(MS_WIN32) && Py_UNICODE_SIZE == 2
 #  define HAVE_USABLE_WCHAR_T
 #  define PY_UNICODE_TYPE wchar_t
 # endif
 
-# if defined(USE_UCS4_STORAGE)
+# if defined(Py_UNICODE_WIDE)
 #  define PY_UNICODE_TYPE Py_UCS4
 # endif
 

Modules/sre.h

@@ -13,8 +13,9 @@
 
 #include "sre_constants.h"
 
-/* size of a code word (must be unsigned short or larger) */
-#ifdef USE_UCS4_STORAGE
+/* size of a code word (must be unsigned short or larger, and
+   large enough to hold a Py_UNICODE character) */
+#ifdef Py_UNICODE_WIDE
 #define SRE_CODE unsigned long
 #else
 #define SRE_CODE unsigned short

Objects/unicodectype.c

@@ -68,7 +68,7 @@ Py_UNICODE _PyUnicode_ToTitlecase(register Py_UNICODE ch)
     else
 	ch += ctype->upper;
 
-#ifdef USE_UCS4_STORAGE
+#ifdef Py_UNICODE_WIDE
     /* The database assumes that the values wrap around at 0x10000. */
     if (ch > 0x10000)
 	ch -= 0x10000;
@@ -360,7 +360,7 @@ Py_UNICODE _PyUnicode_ToUppercase(register Py_UNICODE ch)
     const _PyUnicode_TypeRecord *ctype = gettyperecord(ch);
 
     ch += ctype->upper;
-#ifdef USE_UCS4_STORAGE
+#ifdef Py_UNICODE_WIDE
     /* The database assumes that the values wrap around at 0x10000. */
     if (ch > 0x10000)
 	ch -= 0x10000;
@@ -376,7 +376,7 @@ Py_UNICODE _PyUnicode_ToLowercase(register Py_UNICODE ch)
     const _PyUnicode_TypeRecord *ctype = gettyperecord(ch);
 
     ch += ctype->lower;
-#ifdef USE_UCS4_STORAGE
+#ifdef Py_UNICODE_WIDE
     /* The database assumes that the values wrap around at 0x10000. */
     if (ch > 0x10000)
 	ch -= 0x10000;

Objects/unicodeobject.c

@@ -106,7 +106,7 @@ static char unicode_default_encoding[100];
 Py_UNICODE
 PyUnicode_GetMax()
 {
-#ifdef USE_UCS4_STORAGE
+#ifdef Py_UNICODE_WIDE
 	return 0x10FFFF;
 #else
 	/* This is actually an illegal character, so it should
@@ -791,7 +791,7 @@ PyObject *PyUnicode_DecodeUTF8(const char *s,
                 errmsg = "illegal encoding";
 		goto utf8Error;
 	    }
-#if Py_UNICODE_SIZE == 4
+#ifdef Py_UNICODE_WIDE
 	    *p++ = (Py_UNICODE)ch;
 #else
             /*  compute and append the two surrogates: */
@@ -1080,7 +1080,7 @@ PyObject *PyUnicode_DecodeUTF16(const char *s,
 		    ch2 = (ch2 >> 8) | (ch2 << 8);
 #endif
 	    if (0xDC00 <= ch2 && ch2 <= 0xDFFF) {
-#if Py_UNICODE_SIZE == 2
+#ifndef Py_UNICODE_WIDE
 		/* This is valid data (a UTF-16 surrogate pair), but
 		   we are not able to store this information since our
 		   Py_UNICODE type only has 16 bits... this might
@@ -1326,7 +1326,7 @@ PyObject *PyUnicode_DecodeUnicodeEscape(const char *s,
                 *p++ = (Py_UNICODE) chr;
             else if (chr <= 0x10ffff) {
                 /* UCS-4 character. Either store directly, or as surrogate pair. */
-#if Py_UNICODE_SIZE == 4
+#ifdef Py_UNICODE_WIDE
                 *p++ = chr;
 #else
                 chr -= 0x10000L;

Python/bltinmodule.c

@@ -325,7 +325,7 @@ builtin_unichr(PyObject *self, PyObject *args)
 		return PyUnicode_FromUnicode(s, 1);
 	}
 	else {
-#if Py_UNICODE_SIZE == 2
+#ifndef Py_UNICODE_WIDE
 		/* UCS-4 character.  store as two surrogate characters */
 		x -= 0x10000L;
 		s[0] = 0xD800 + (Py_UNICODE) (x >> 10);