#include <sys/param.h>
__FBSDID("$FreeBSD: src/lib/libc/locale/utf8.c,v 1.16 2007/10/15 09:51:30 ache Exp $");
#include "xlocale_private.h"
#include <errno.h>
#include <limits.h>
#include <runetype.h>
#include <stdlib.h>
#include <string.h>
#include <wchar.h>
#include "mblocal.h"
#define UTF2_MB_CUR_MAX 3
static size_t _UTF2_mbrtowc(wchar_t * __restrict, const char * __restrict,
size_t, mbstate_t * __restrict, locale_t);
static int _UTF2_mbsinit(const mbstate_t *, locale_t);
static size_t _UTF2_mbsnrtowcs(wchar_t * __restrict,
const char ** __restrict, size_t, size_t,
mbstate_t * __restrict, locale_t);
static size_t _UTF2_wcrtomb(char * __restrict, wchar_t,
mbstate_t * __restrict, locale_t);
static size_t _UTF2_wcsnrtombs(char * __restrict, const wchar_t ** __restrict,
size_t, size_t, mbstate_t * __restrict, locale_t);
typedef struct {
wchar_t ch;
int want;
wchar_t lbound;
} _UTF2State;
int
_UTF2_init(struct __xlocale_st_runelocale *xrl)
{
xrl->__mbrtowc = _UTF2_mbrtowc;
xrl->__wcrtomb = _UTF2_wcrtomb;
xrl->__mbsinit = _UTF2_mbsinit;
xrl->__mbsnrtowcs = _UTF2_mbsnrtowcs;
xrl->__wcsnrtombs = _UTF2_wcsnrtombs;
xrl->__mb_cur_max = UTF2_MB_CUR_MAX;
xrl->__mb_sb_limit = 128;
return (0);
}
static int
_UTF2_mbsinit(const mbstate_t *ps, locale_t loc)
{
return (ps == NULL || ((const _UTF2State *)ps)->want == 0);
}
static size_t
_UTF2_mbrtowc(wchar_t * __restrict pwc, const char * __restrict s, size_t n,
mbstate_t * __restrict ps, locale_t loc)
{
_UTF2State *us;
int ch, i, mask, want;
wchar_t lbound, wch;
us = (_UTF2State *)ps;
if (us->want < 0 || us->want > 6) {
errno = EINVAL;
return ((size_t)-1);
}
if (s == NULL) {
s = "";
n = 1;
pwc = NULL;
}
if (n == 0)
return ((size_t)-2);
if (us->want == 0 && ((ch = (unsigned char)*s) & ~0x7f) == 0) {
if (pwc != NULL)
*pwc = ch;
return (ch != '\0' ? 1 : 0);
}
if (us->want == 0) {
ch = (unsigned char)*s;
if ((ch & 0x80) == 0) {
mask = 0x7f;
want = 1;
lbound = 0;
} else if ((ch & 0xe0) == 0xc0) {
mask = 0x1f;
want = 2;
lbound = 0x80;
} else if ((ch & 0xf0) == 0xe0) {
mask = 0x0f;
want = 3;
lbound = 0x800;
} else {
errno = EILSEQ;
return ((size_t)-1);
}
} else {
want = us->want;
lbound = us->lbound;
}
if (us->want == 0)
wch = (unsigned char)*s++ & mask;
else
wch = us->ch;
for (i = (us->want == 0) ? 1 : 0; i < MIN(want, n); i++) {
if ((*s & 0xc0) != 0x80) {
errno = EILSEQ;
return ((size_t)-1);
}
wch <<= 6;
wch |= *s++ & 0x3f;
}
if (i < want) {
us->want = want - i;
us->lbound = lbound;
us->ch = wch;
return ((size_t)-2);
}
if (wch < lbound) {
errno = EILSEQ;
return ((size_t)-1);
}
if (pwc != NULL)
*pwc = wch;
us->want = 0;
return (wch == L'\0' ? 0 : want);
}
static size_t
_UTF2_mbsnrtowcs(wchar_t * __restrict dst, const char ** __restrict src,
size_t nms, size_t len, mbstate_t * __restrict ps, locale_t loc)
{
_UTF2State *us;
const char *s;
size_t nchr;
wchar_t wc;
size_t nb;
us = (_UTF2State *)ps;
s = *src;
nchr = 0;
if (dst == NULL) {
if (nms > 0 && us->want > 0 && (signed char)*s > 0) {
errno = EILSEQ;
return ((size_t)-1);
}
for (;;) {
if (nms > 0 && (signed char)*s > 0)
nb = 1;
else if ((nb = _UTF2_mbrtowc(&wc, s, nms, ps, loc)) ==
(size_t)-1)
return ((size_t)-1);
else if (nb == 0 || nb == (size_t)-2)
return (nchr);
s += nb;
nms -= nb;
nchr++;
}
}
if (nms > 0 && len > 0 && us->want > 0 && (signed char)*s > 0) {
errno = EILSEQ;
return ((size_t)-1);
}
while (len-- > 0) {
if (nms > 0 && (signed char)*s > 0) {
*dst = (wchar_t)*s;
nb = 1;
} else if ((nb = _UTF2_mbrtowc(dst, s, nms, ps, loc)) ==
(size_t)-1) {
*src = s;
return ((size_t)-1);
} else if (nb == (size_t)-2) {
*src = s + nms;
return (nchr);
} else if (nb == 0) {
*src = NULL;
return (nchr);
}
s += nb;
nms -= nb;
nchr++;
dst++;
}
*src = s;
return (nchr);
}
static size_t
_UTF2_wcrtomb(char * __restrict s, wchar_t wc, mbstate_t * __restrict ps, locale_t loc)
{
_UTF2State *us;
unsigned char lead;
int i, len;
us = (_UTF2State *)ps;
if (us->want != 0) {
errno = EINVAL;
return ((size_t)-1);
}
if (s == NULL)
return (1);
if ((wc & ~0x7f) == 0) {
*s = (char)wc;
return (1);
}
if ((wc & ~0x7f) == 0) {
lead = 0;
len = 1;
} else if ((wc & ~0x7ff) == 0) {
lead = 0xc0;
len = 2;
} else if ((wc & ~0xffff) == 0) {
lead = 0xe0;
len = 3;
} else {
errno = EILSEQ;
return ((size_t)-1);
}
for (i = len - 1; i > 0; i--) {
s[i] = (wc & 0x3f) | 0x80;
wc >>= 6;
}
*s = (wc & 0xff) | lead;
return (len);
}
static size_t
_UTF2_wcsnrtombs(char * __restrict dst, const wchar_t ** __restrict src,
size_t nwc, size_t len, mbstate_t * __restrict ps, locale_t loc)
{
_UTF2State *us;
char buf[MB_LEN_MAX];
const wchar_t *s;
size_t nbytes;
size_t nb;
us = (_UTF2State *)ps;
if (us->want != 0) {
errno = EINVAL;
return ((size_t)-1);
}
s = *src;
nbytes = 0;
if (dst == NULL) {
while (nwc-- > 0) {
if (0 <= *s && *s < 0x80)
nb = 1;
else if ((nb = _UTF2_wcrtomb(buf, *s, ps, loc)) ==
(size_t)-1)
return ((size_t)-1);
if (*s == L'\0')
return (nbytes + nb - 1);
s++;
nbytes += nb;
}
return (nbytes);
}
while (len > 0 && nwc-- > 0) {
if (0 <= *s && *s < 0x80) {
nb = 1;
*dst = *s;
} else if (len > (size_t)UTF2_MB_CUR_MAX) {
if ((nb = _UTF2_wcrtomb(dst, *s, ps, loc)) == (size_t)-1) {
*src = s;
return ((size_t)-1);
}
} else {
if ((nb = _UTF2_wcrtomb(buf, *s, ps, loc)) == (size_t)-1) {
*src = s;
return ((size_t)-1);
}
if (nb > (int)len)
break;
memcpy(dst, buf, nb);
}
if (*s == L'\0') {
*src = NULL;
return (nbytes + nb - 1);
}
s++;
dst += nb;
len -= nb;
nbytes += nb;
}
*src = s;
return (nbytes);
}