Some small cleanups for multibyte functions

Author: th-otto

multibyte/internal.c

@@ -17,7 +17,7 @@
              | x )
 #define F(x) ( ( x>=5 ? 0 : \
                  x==0 ? R(0x90,0xc0) : \
-                 x==4 ? R(0x80,0xa0) : \
+                 x==4 ? R(0x80,0x90) : \
                  R(0x80,0xc0) ) \
              | ( R(0x80,0xc0) >> 6 ) \
              | ( R(0x80,0xc0) >> 12 ) \
@@ -32,7 +32,3 @@ const uint32_t bittab[] = {
 	E(0x8),E(0x9),E(0xa),E(0xb),E(0xc),E(0xd),E(0xe),E(0xf),
 	F(0x0),F(0x1),F(0x2),F(0x3),F(0x4)
 };
-
-#ifdef BROKEN_VISIBILITY
-__asm__(".hidden __fsmu8");
-#endif

multibyte/internal.h

@@ -16,8 +16,12 @@ extern const uint32_t bittab[]; /* ATTR_LIBC_VISIBILITY; */
 #define OOB(c,b) (((((b)>>3)-0x10)|(((b)>>3)+((int32_t)(c)>>26))) & ~7)
 
 /* Interval [a,b). Either a must be 80 or b must be c0, lower 3 bits clear. */
-#define R(a,b) ((uint32_t)((a==0x80 ? 0x40-b : -a) << 23))
+#define R(a,b) ((uint32_t)(a==0x80 ? 0x40u-b : 0u-a) << 23)
 #define FAILSTATE R(0x80,0x80)
 
 #define SA 0xc2u
 #define SB 0xf4u
+
+/* Arbitrary encoding for representing code units instead of characters. */
+#define CODEUNIT(c) (0xdfff & (signed char)(c))
+#define IS_CODEUNIT(c) ((unsigned)(c)-0xdf80 < 0x80)

multibyte/mbrlen.c

@@ -13,6 +13,6 @@
 
 size_t mbrlen(const char *__restrict s, size_t n, mbstate_t *__restrict st)
 {
-	static unsigned internal;
-	return mbrtowc(0, s, n, st ? st : (mbstate_t *)&internal);
+	static mbstate_t internal;
+	return mbrtowc(0, s, n, st ? st : &internal);
 }

multibyte/mbrtowc.c

@@ -17,14 +17,15 @@ size_t mbrtowc(wchar_t *__restrict wc, const char *__restrict src, size_t n, mbs
 	unsigned c;
 	const unsigned char *s = (const void *)src;
 	const unsigned N = n;
+	wchar_t dummy;
 
 	if (!st) st = &internal_state;
 	c = st->__opaque1;
 
 	if (!s) {
 		if (c) goto ilseq;
 		return 0;
-	} else if (!wc) wc = (void *)&wc;
+	} else if (!wc) wc = &dummy;
 
 	if (!n) return -2;
 	if (!c) {

multibyte/mbsinit.c

@@ -13,5 +13,5 @@
 
 int mbsinit(const mbstate_t *st)
 {
-	return !st || !*(unsigned *)st;
+	return !st || !st->__opaque1;
 }

multibyte/mbsnrtowcs.c

@@ -52,7 +52,7 @@ size_t mbsnrtowcs(wchar_t *__restrict wcs, const char **__restrict src, size_t n
 				break;
 			}
 			/* have to roll back partial character */
-			*(unsigned *)st = 0;
+			st->__opaque1 = 0;
 			break;
 		}
 		s += l; n -= l;

multibyte/mbsrtowcs.c

@@ -17,9 +17,9 @@ size_t mbsrtowcs(wchar_t *__restrict ws, const char **__restrict src, size_t wn,
 	size_t wn0 = wn;
 	unsigned c = 0;
 
-	if (st && (c = *(unsigned *)st)) {
+	if (st && (c = st->__opaque1) != 0) {
 		if (ws) {
-			*(unsigned *)st = 0;
+			st->__opaque1 = 0;
 			goto resume;
 		} else {
 			goto resume0;

multibyte/mbtowc.c

@@ -15,10 +15,11 @@ int mbtowc(wchar_t *__restrict wc, const char *__restrict src, size_t n)
 {
 	unsigned c;
 	const unsigned char *s = (const void *)src;
+	wchar_t dummy;
 
 	if (!s) return 0;
 	if (!n) goto ilseq;
-	if (!wc) wc = (void *)&wc;
+	if (!wc) wc = &dummy;
 
 	if (*s < 0x80) return !!(*wc = *s);
 	if (*s-SA > SB-SA) goto ilseq;