/* Convert the multibyte character at S, which is no longer
than N characters, to its `wchar_t' representation, placing
this n *PWC and returning its length.
Attention: this function should NEVER be intentionally used.
The interface is completely stupid. The state is shared between
all conversion functions. You should use instead the restartable
version `mbrtowc'. */
int
mbtowc (wchar_t *pwc, const char *s, size_t n)
{
int result;
static mbstate_t state;
/* If S is NULL the function has to return null or not null
depending on the encoding having a state depending encoding or
not. */
if (s == NULL)
{
const struct gconv_fcts *fcts;
/* Get the conversion functions. */
fcts = get_gconv_fcts (_NL_CURRENT_DATA (LC_CTYPE));
/* This is an extension in the Unix standard which does not directly
violate ISO C. */
memset (&state, '\0', sizeof state);
result = fcts->towc->__stateful;
}
else if (*s == '\0')
{
if (pwc != NULL)
*pwc = L'\0';
result = 0;
}
else
{
result = __mbrtowc (pwc, s, n, &state);
/* The `mbrtowc' functions tell us more than we need. Fold the -1
and -2 result into -1. */
if (result < 0)
result = -1;
}
return result;
}
/* Convert WCHAR into its multibyte character representation,
putting this in S and returning its length.
Attention: this function should NEVER be intentionally used.
The interface is completely stupid. The state is shared between
all conversion functions. You should use instead the restartable
version `wcrtomb'. */
int
wctomb (char *s, wchar_t wchar)
{
/* If S is NULL the function has to return null or not null
depending on the encoding having a state depending encoding or
not. */
if (s == NULL)
{
const struct gconv_fcts *fcts;
/* Get the conversion functions. */
fcts = get_gconv_fcts (_NL_CURRENT_DATA (LC_CTYPE));
/* This is an extension in the Unix standard which does not directly
violate ISO C. */
memset (&__wctomb_state, '\0', sizeof __wctomb_state);
return fcts->tomb->__stateful;
}
return __wcrtomb (s, wchar, &__wctomb_state);
}
/* Return the length of the multibyte character (if there is one)
at S which is no longer than N characters.
The ISO C standard says that the `mblen' function must not change
the state of the `mbtowc' function. */
int
mblen (const char *s, size_t n)
{
int result;
/* If S is NULL the function has to return null or not null
depending on the encoding having a state depending encoding or
not. */
if (s == NULL)
{
const struct gconv_fcts *fcts;
/* Get the conversion functions. */
fcts = get_gconv_fcts (_NL_CURRENT_DATA (LC_CTYPE));
/* Reset the state. */
memset (&state, '\0', sizeof state);
result = fcts->towc->__stateful;
}
else if (*s == '\0')
/* According to the ISO C 89 standard this is the expected behaviour. */
result = 0;
else
{
memset (&state, '\0', sizeof state);
result = __mbrtowc (NULL, s, n, &state);
/* The `mbrtowc' functions tell us more than we need. Fold the -1
and -2 result into -1. */
if (result < 0)
result = -1;
}
return result;
}
wctype_t
__wctype (const char *property)
{
const char *names;
unsigned int result;
size_t proplen = strlen (property);
size_t i;
names = _NL_CURRENT (LC_CTYPE, _NL_CTYPE_CLASS_NAMES);
for (result = 0; ; result++)
{
size_t nameslen = strlen (names);
if (proplen == nameslen && memcmp (property, names, proplen) == 0)
break;
names += nameslen + 1;
if (names[0] == '\0')
return 0;
}
i = _NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_CLASS_OFFSET) + result;
return (wctype_t) _NL_CURRENT_DATA (LC_CTYPE)->values[i].string;
}
wctrans_t
__wctrans (const char *property)
{
const char *names;
size_t cnt;
size_t i;
names = _NL_CURRENT (LC_CTYPE, _NL_CTYPE_MAP_NAMES);
cnt = 0;
while (names[0] != '\0')
{
if (strcmp (property, names) == 0)
break;
names = strchr (names, '\0') + 1;
++cnt;
}
if (names[0] == '\0')
return 0;
i = _NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_MAP_OFFSET) + cnt;
return (wctrans_t) _NL_CURRENT_DATA (LC_CTYPE)->values[i].string;
}
size_t
__wcrtomb (char *s, wchar_t wc, mbstate_t *ps)
{
char buf[MB_CUR_MAX];
struct __gconv_step_data data;
int status;
size_t result;
size_t dummy;
const struct gconv_fcts *fcts;
/* Set information for this step. */
data.__invocation_counter = 0;
data.__internal_use = 1;
data.__flags = __GCONV_IS_LAST;
data.__statep = ps ?: &state;
data.__trans = NULL;
/* A first special case is if S is NULL. This means put PS in the
initial state. */
if (s == NULL)
{
s = buf;
wc = L'\0';
}
/* Tell where we want to have the result. */
data.__outbuf = (unsigned char *) s;
data.__outbufend = (unsigned char *) s + MB_CUR_MAX;
/* Get the conversion functions. */
fcts = get_gconv_fcts (_NL_CURRENT_DATA (LC_CTYPE));
__gconv_fct fct = fcts->tomb->__fct;
#ifdef PTR_DEMANGLE
if (fcts->tomb->__shlib_handle != NULL)
PTR_DEMANGLE (fct);
#endif
/* If WC is the NUL character we write into the output buffer the byte
sequence necessary for PS to get into the initial state, followed
by a NUL byte. */
if (wc == L'\0')
{
status = DL_CALL_FCT (fct, (fcts->tomb, &data, NULL, NULL,
NULL, &dummy, 1, 1));
if (status == __GCONV_OK || status == __GCONV_EMPTY_INPUT)
*data.__outbuf++ = '\0';
}
else
{
/* Do a normal conversion. */
const unsigned char *inbuf = (const unsigned char *) &wc;
status = DL_CALL_FCT (fct,
(fcts->tomb, &data, &inbuf,
inbuf + sizeof (wchar_t), NULL, &dummy, 0, 1));
}
/* There must not be any problems with the conversion but illegal input
characters. The output buffer must be large enough, otherwise the
definition of MB_CUR_MAX is not correct. All the other possible
errors also must not happen. */
assert (status == __GCONV_OK || status == __GCONV_EMPTY_INPUT
|| status == __GCONV_ILLEGAL_INPUT
|| status == __GCONV_INCOMPLETE_INPUT
|| status == __GCONV_FULL_OUTPUT);
if (status == __GCONV_OK || status == __GCONV_EMPTY_INPUT
|| status == __GCONV_FULL_OUTPUT)
result = data.__outbuf - (unsigned char *) s;
else
{
result = (size_t) -1;
__set_errno (EILSEQ);
}
return result;
}
wint_t
__btowc (int c)
{
const struct gconv_fcts *fcts;
/* If the parameter does not fit into one byte or it is the EOF value
we can give the answer now. */
if (c < SCHAR_MIN || c > UCHAR_MAX || c == EOF)
return WEOF;
/* We know that only ASCII compatible encodings are used for the
locale and that the wide character encoding is ISO 10646. */
if (isascii (c))
return (wint_t) c;
/* Get the conversion functions. */
fcts = get_gconv_fcts (_NL_CURRENT_DATA (LC_CTYPE));
__gconv_btowc_fct btowc_fct = fcts->towc->__btowc_fct;
#ifdef PTR_DEMANGLE
if (fcts->towc->__shlib_handle != NULL)
PTR_DEMANGLE (btowc_fct);
#endif
if (__builtin_expect (fcts->towc_nsteps == 1, 1)
&& __builtin_expect (btowc_fct != NULL, 1))
{
/* Use the shortcut function. */
return DL_CALL_FCT (btowc_fct, (fcts->towc, (unsigned char) c));
}
else
{
/* Fall back to the slow but generic method. */
wchar_t result;
struct __gconv_step_data data;
unsigned char inbuf[1];
const unsigned char *inptr = inbuf;
size_t dummy;
int status;
/* Tell where we want the result. */
data.__outbuf = (unsigned char *) &result;
data.__outbufend = data.__outbuf + sizeof (wchar_t);
data.__invocation_counter = 0;
data.__internal_use = 1;
data.__flags = __GCONV_IS_LAST;
data.__statep = &data.__state;
/* Make sure we start in the initial state. */
memset (&data.__state, '\0', sizeof (mbstate_t));
/* Create the input string. */
inbuf[0] = c;
__gconv_fct fct = fcts->towc->__fct;
#ifdef PTR_DEMANGLE
if (fcts->towc->__shlib_handle != NULL)
PTR_DEMANGLE (fct);
#endif
status = DL_CALL_FCT (fct, (fcts->towc, &data, &inptr, inptr + 1,
NULL, &dummy, 0, 1));
if (status != __GCONV_OK && status != __GCONV_FULL_OUTPUT
&& status != __GCONV_EMPTY_INPUT)
/* The conversion failed. */
result = WEOF;
return result;
}
}
size_t
__mbrtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *ps)
{
wchar_t buf[1];
struct __gconv_step_data data;
int status;
size_t result;
size_t dummy;
const unsigned char *inbuf, *endbuf;
char *outbuf = (char *) (pwc ?: buf);
const struct gconv_fcts *fcts;
/* Set information for this step. */
data.__invocation_counter = 0;
data.__internal_use = 1;
data.__flags = __GCONV_IS_LAST;
data.__statep = ps ?: &state;
data.__trans = NULL;
/* A first special case is if S is NULL. This means put PS in the
initial state. */
if (s == NULL)
{
outbuf = (char *) buf;
s = "";
n = 1;
}
/* Tell where we want the result. */
data.__outbuf = outbuf;
data.__outbufend = outbuf + sizeof (wchar_t);
/* Get the conversion functions. */
fcts = get_gconv_fcts (_NL_CURRENT_DATA (LC_CTYPE));
/* Do a normal conversion. */
inbuf = (const unsigned char *) s;
endbuf = inbuf + n;
if (__builtin_expect (endbuf < inbuf, 0))
endbuf = (const unsigned char *) ~(uintptr_t) 0;
status = DL_CALL_FCT (fcts->towc->__fct,
(fcts->towc, &data, &inbuf, endbuf,
NULL, &dummy, 0, 1));
/* There must not be any problems with the conversion but illegal input
characters. The output buffer must be large enough, otherwise the
definition of MB_CUR_MAX is not correct. All the other possible
errors also must not happen. */
assert (status == __GCONV_OK || status == __GCONV_EMPTY_INPUT
|| status == __GCONV_ILLEGAL_INPUT
|| status == __GCONV_INCOMPLETE_INPUT
|| status == __GCONV_FULL_OUTPUT);
if (status == __GCONV_OK || status == __GCONV_EMPTY_INPUT
|| status == __GCONV_FULL_OUTPUT)
{
if (data.__outbuf != (unsigned char *) outbuf
&& *(wchar_t *) outbuf == L'\0')
{
/* The converted character is the NUL character. */
assert (__mbsinit (data.__statep));
result = 0;
}
else
result = inbuf - (const unsigned char *) s;
}
else if (status == __GCONV_INCOMPLETE_INPUT)
result = (size_t) -2;
else
{
result = (size_t) -1;
__set_errno (EILSEQ);
}
return result;
}
size_t
mbrtoc16 (char16_t *pc16, const char *s, size_t n, mbstate_t *ps)
{
if (ps == NULL)
ps = &state;
/* The standard text does not say that S being NULL means the state
is reset even if the second half of a surrogate still have to be
returned. In fact, the error code description indicates
otherwise. Therefore always first try to return a second
half. */
if (ps->__count & 0x80000000)
{
/* We have to return the second word for a surrogate. */
ps->__count &= 0x7fffffff;
*pc16 = ps->__value.__wch;
ps->__value.__wch = L'\0';
return (size_t) -3;
}
wchar_t wc;
struct __gconv_step_data data;
int status;
size_t result;
size_t dummy;
const unsigned char *inbuf, *endbuf;
unsigned char *outbuf = (unsigned char *) &wc;
const struct gconv_fcts *fcts;
/* Set information for this step. */
data.__invocation_counter = 0;
data.__internal_use = 1;
data.__flags = __GCONV_IS_LAST;
data.__statep = ps;
data.__trans = NULL;
/* A first special case is if S is NULL. This means put PS in the
initial state. */
if (s == NULL)
{
pc16 = NULL;
s = "";
n = 1;
}
/* Tell where we want the result. */
data.__outbuf = outbuf;
data.__outbufend = outbuf + sizeof (wchar_t);
/* Get the conversion functions. */
fcts = get_gconv_fcts (_NL_CURRENT_DATA (LC_CTYPE));
/* Do a normal conversion. */
inbuf = (const unsigned char *) s;
endbuf = inbuf + n;
if (__glibc_unlikely (endbuf < inbuf))
{
endbuf = (const unsigned char *) ~(uintptr_t) 0;
if (endbuf == inbuf)
goto ilseq;
}
__gconv_fct fct = fcts->towc->__fct;
#ifdef PTR_DEMANGLE
if (fcts->towc->__shlib_handle != NULL)
PTR_DEMANGLE (fct);
#endif
status = DL_CALL_FCT (fct, (fcts->towc, &data, &inbuf, endbuf,
NULL, &dummy, 0, 1));
/* There must not be any problems with the conversion but illegal input
characters. The output buffer must be large enough, otherwise the
definition of MB_CUR_MAX is not correct. All the other possible
errors also must not happen. */
assert (status == __GCONV_OK || status == __GCONV_EMPTY_INPUT
|| status == __GCONV_ILLEGAL_INPUT
|| status == __GCONV_INCOMPLETE_INPUT
|| status == __GCONV_FULL_OUTPUT);
if (status == __GCONV_OK || status == __GCONV_EMPTY_INPUT
|| status == __GCONV_FULL_OUTPUT)
{
result = inbuf - (const unsigned char *) s;
if (wc < 0x10000)
{
if (pc16 != NULL)
*pc16 = wc;
if (data.__outbuf != outbuf && wc == L'\0')
{
/* The converted character is the NUL character. */
assert (__mbsinit (data.__statep));
result = 0;
}
}
else
{
/* This is a surrogate. */
if (pc16 != NULL)
*pc16 = 0xd7c0 + (wc >> 10);
ps->__count |= 0x80000000;
ps->__value.__wch = 0xdc00 + (wc & 0x3ff);
}
}
