#endif
}
+/* Tests if some arbitrary number of bytes begins in a valid UTF-8 character.
+ * The actual number of bytes in the UTF-8 character will be returned if it
+ * is valid, otherwise 0. */
+int
+Perl_is_utf8_char(pTHX_ U8 *s)
+{
+ U8 u = *s;
+ int slen, len;
+
+ if (!(u & 0x80))
+ return 1;
+
+ if (!(u & 0x40))
+ return 0;
+
+ if (!(u & 0x20)) { len = 2; }
+ else if (!(u & 0x10)) { len = 3; }
+ else if (!(u & 0x08)) { len = 4; }
+ else if (!(u & 0x04)) { len = 5; }
+ else if (!(u & 0x02)) { len = 6; }
+ else if (!(u & 0x01)) { len = 7; }
+ else { len = 13; } /* whoa! */
+
+ slen = len - 1;
+ s++;
+ while (slen--) {
+ if ((*s & 0xc0) != 0x80)
+ return 0;
+ s++;
+ }
+ return len;
+}
+
+/*
+=for apidoc Am|bool_utf8_string|U8 *s|STRLEN len
+
+Returns true if first C<len> bytes of the given string form valid a UTF8
+string, false otherwise.
+
+=cut
+*/
+
+bool
+Perl_is_utf8_string(pTHX_ U8 *s, STRLEN len)
+{
+ U8* x=s;
+ U8* send=s+len;
+ int c;
+ while (x < send) {
+ c = is_utf8_char(x);
+ x += c;
+ if (!c || x > send)
+ return 0;
+ }
+ return 1;
+}
+
UV
Perl_utf8_to_uv(pTHX_ U8* s, I32* retlen)
{
return s;
}
-/* XXX NOTHING CALLS THE FOLLOWING TWO ROUTINES YET!!! */
+/*
+=for apidoc Am|U8 *|utf8_to_bytes|U8 *s|STRLEN len
+
+Converts a string C<s> of length C<len> from UTF8 into ASCII encoding.
+Unlike C<bytes_to_utf8>, this over-writes the original string.
+Returns zero on failure after converting as much as possible.
+
+=cut
+*/
+
+U8 *
+Perl_utf8_to_bytes(pTHX_ U8* s, STRLEN len)
+{
+ dTHR;
+ U8 *send;
+ U8 *d;
+ U8 *save;
+
+ send = s + len;
+ d = save = s;
+ while (s < send) {
+ if (*s < 0x80)
+ *d++ = *s++;
+ else {
+ I32 ulen;
+ UV uv = utf8_to_uv(s, &ulen);
+ if (uv > 255) {
+ *d = '\0';
+ return 0;
+ }
+ s += ulen;
+ *d++ = (U8)uv;
+ }
+ }
+ *d = '\0';
+ return save;
+}
+
+/*
+=for apidoc Am|U8 *|bytes_to_utf8|U8 *s|STRLEN *len
+
+Converts a string C<s> of length C<len> from ASCII into UTF8 encoding.
+Returns a pointer to the newly-created string, and sets C<len> to
+reflect the new length.
+
+=cut
+*/
+
+U8*
+Perl_bytes_to_utf8(pTHX_ U8* s, STRLEN *len)
+{
+ dTHR;
+ U8 *send;
+ U8 *d;
+ U8 *dst;
+ send = s + (*len);
+
+ Newz(801, d, (*len) * 2 + 1, U8);
+ dst = d;
+
+ while (s < send) {
+ if (*s < 0x80)
+ *d++ = *s++;
+ else {
+ UV uv = *s++;
+ *d++ = (( uv >> 6) | 0xc0);
+ *d++ = (( uv & 0x3f) | 0x80);
+ }
+ }
+ *d = '\0';
+ *len = d-dst;
+ return dst;
+}
+
/*
* Convert native or reversed UTF-16 to UTF-8.
*
* Destination must be pre-extended to 3/2 source. Do not use in-place.
* We optimize for native, for obvious reasons. */
+/* There are several problems with utf16_to_utf8().
+ * (1) U16 is not necessarily *exactly* two bytes.
+ * (2) Secondly, no check is made for odd length.
+ * (3) Thirdly, the "Malformed UTF-16 surrogate" should probably be
+ * a hard error (and it should be listed in perldiag).
+ * (4) The tests (in comp/t/require.t) are a joke: the UTF16 BOM
+ * really ought to be followed by valid UTF16 characters.
+ * --Mike Guy */
U8*
Perl_utf16_to_utf8(pTHX_ U16* p, U8* d, I32 bytelen)
{
bool
Perl_is_uni_alnum(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_alnum(tmpbuf);
}
bool
Perl_is_uni_alnumc(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_alnumc(tmpbuf);
}
bool
Perl_is_uni_idfirst(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_idfirst(tmpbuf);
}
bool
Perl_is_uni_alpha(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_alpha(tmpbuf);
}
bool
Perl_is_uni_ascii(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_ascii(tmpbuf);
}
bool
Perl_is_uni_space(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_space(tmpbuf);
}
bool
Perl_is_uni_digit(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_digit(tmpbuf);
}
bool
Perl_is_uni_upper(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_upper(tmpbuf);
}
bool
Perl_is_uni_lower(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_lower(tmpbuf);
}
bool
Perl_is_uni_cntrl(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_cntrl(tmpbuf);
}
bool
Perl_is_uni_graph(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_graph(tmpbuf);
}
bool
Perl_is_uni_print(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_print(tmpbuf);
}
bool
Perl_is_uni_punct(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_punct(tmpbuf);
}
bool
Perl_is_uni_xdigit(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return is_utf8_xdigit(tmpbuf);
}
U32
Perl_to_uni_upper(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return to_utf8_upper(tmpbuf);
}
U32
Perl_to_uni_title(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return to_utf8_title(tmpbuf);
}
U32
Perl_to_uni_lower(pTHX_ U32 c)
{
- U8 tmpbuf[10];
+ U8 tmpbuf[UTF8_MAXLEN];
uv_to_utf8(tmpbuf, (UV)c);
return to_utf8_lower(tmpbuf);
}
bool
Perl_is_utf8_alnum(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_alnum)
PL_utf8_alnum = swash_init("utf8", "IsAlnum", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_alnum, p);
bool
Perl_is_utf8_alnumc(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_alnum)
PL_utf8_alnum = swash_init("utf8", "IsAlnumC", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_alnum, p);
bool
Perl_is_utf8_alpha(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_alpha)
PL_utf8_alpha = swash_init("utf8", "IsAlpha", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_alpha, p);
bool
Perl_is_utf8_ascii(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_ascii)
PL_utf8_ascii = swash_init("utf8", "IsAscii", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_ascii, p);
bool
Perl_is_utf8_space(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_space)
PL_utf8_space = swash_init("utf8", "IsSpace", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_space, p);
bool
Perl_is_utf8_digit(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_digit)
PL_utf8_digit = swash_init("utf8", "IsDigit", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_digit, p);
bool
Perl_is_utf8_upper(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_upper)
PL_utf8_upper = swash_init("utf8", "IsUpper", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_upper, p);
bool
Perl_is_utf8_lower(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_lower)
PL_utf8_lower = swash_init("utf8", "IsLower", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_lower, p);
bool
Perl_is_utf8_cntrl(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_cntrl)
PL_utf8_cntrl = swash_init("utf8", "IsCntrl", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_cntrl, p);
bool
Perl_is_utf8_graph(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_graph)
PL_utf8_graph = swash_init("utf8", "IsGraph", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_graph, p);
bool
Perl_is_utf8_print(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_print)
PL_utf8_print = swash_init("utf8", "IsPrint", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_print, p);
bool
Perl_is_utf8_punct(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_punct)
PL_utf8_punct = swash_init("utf8", "IsPunct", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_punct, p);
bool
Perl_is_utf8_xdigit(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_xdigit)
PL_utf8_xdigit = swash_init("utf8", "IsXDigit", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_xdigit, p);
bool
Perl_is_utf8_mark(pTHX_ U8 *p)
{
+ if (!is_utf8_char(p))
+ return FALSE;
if (!PL_utf8_mark)
PL_utf8_mark = swash_init("utf8", "IsM", &PL_sv_undef, 0, 0);
return swash_fetch(PL_utf8_mark, p);
SV* retval;
char tmpbuf[256];
dSP;
+
+ if (!gv_stashpv(pkg, 0)) { /* demand load utf8 */
+ ENTER;
+ Perl_load_module(aTHX_ PERL_LOADMOD_NOIMPORT, newSVpv(pkg,0), Nullsv);
+ LEAVE;
+ }
+ SPAGAIN;
PUSHSTACKi(PERLSI_MAGIC);
PUSHMARK(SP);
EXTEND(SP,5);
if (hv == PL_last_swash_hv &&
klen == PL_last_swash_klen &&
- (!klen || memEQ(ptr,PL_last_swash_key,klen)) )
+ (!klen || memEQ((char *)ptr,(char *)PL_last_swash_key,klen)) )
{
tmps = PL_last_swash_tmps;
slen = PL_last_swash_slen;