*
*/
+/* Use UTF-8 as the default script encoding?
+ * Turning this on will break scripts having non-UTF8 binary
+ * data (such as Latin-1) in string literals. */
+#ifdef USE_UTF8_SCRIPTS
+# define USE_UTF8_IN_NAMES (!IN_BYTES)
+#else
+# define USE_UTF8_IN_NAMES (PL_hints & HINT_UTF8)
+#endif
+
#ifdef EBCDIC
/* The equivalent of these macros but implementing UTF-EBCDIC
are in the following header file:
* (that is, the two high bits are set). Otherwise we risk loading in the
* heavy-duty SWASHINIT and SWASHGET routines unnecessarily.
*/
-#define isIDFIRST_lazy_if(p,c) ((IN_BYTE || (!c || (*((U8*)p) < 0xc0))) \
+#define isIDFIRST_lazy_if(p,c) ((IN_BYTES || (!c || (*((U8*)p) < 0xc0))) \
? isIDFIRST(*(p)) \
: isIDFIRST_utf8((U8*)p))
-#define isALNUM_lazy_if(p,c) ((IN_BYTE || (!c || (*((U8*)p) < 0xc0))) \
+#define isALNUM_lazy_if(p,c) ((IN_BYTES || (!c || (*((U8*)p) < 0xc0))) \
? isALNUM(*(p)) \
: isALNUM_utf8((U8*)p))
#define UTF8_MAXLEN 13 /* how wide can a single UTF8 encoded character become */
-/* #define IN_UTF8 (PL_curcop->op_private & HINT_UTF8) */
-#define IN_BYTE (PL_curcop->op_private & HINT_BYTE)
-#define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTE)
+#define IN_BYTES (PL_curcop->op_private & HINT_BYTES)
+#define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES)
#define UTF8_ALLOW_EMPTY 0x0001
#define UTF8_ALLOW_CONTINUATION 0x0002
UTF8_ALLOW_SURROGATE|UTF8_ALLOW_BOM|\
UTF8_ALLOW_FFFF|UTF8_ALLOW_LONG)
#define UTF8_ALLOW_ANY 0x00ff
-#define UTF8_CHECK_ONLY 0x0100
+#define UTF8_CHECK_ONLY 0x0200
#define UNICODE_SURROGATE_FIRST 0xd800
#define UNICODE_SURROGATE_LAST 0xdfff
#define UNICODE_IS_SURROGATE(c) ((c) >= UNICODE_SURROGATE_FIRST && \
(c) <= UNICODE_SURROGATE_LAST)
-#define UNICODE_IS_REPLACEMENT(c) ((c) == UNICODE_REPLACMENT)
+#define UNICODE_IS_REPLACEMENT(c) ((c) == UNICODE_REPLACEMENT)
#define UNICODE_IS_BYTE_ORDER_MARK(c) ((c) == UNICODE_BYTER_ORDER_MARK)
#define UNICODE_IS_ILLEGAL(c) ((c) == UNICODE_ILLEGAL)
-#define UTF8_QUAD_MAX UINT64_C(0x1000000000)
+#ifdef HAS_QUAD
+# define UTF8_QUAD_MAX UINT64_C(0x1000000000)
+#endif
#define UTF8_IS_ASCII(c) UTF8_IS_INVARIANT(c)