/* utf8.c
*
- * Copyright (c) 1998-1999, Larry Wall
+ * Copyright (c) 1998-2000, Larry Wall
*
* You may distribute under the terms of either the GNU General Public
* License or the Artistic License, as specified in the README file.
*d++ = (( uv & 0x3f) | 0x80);
return d;
}
-#ifdef Quad_t
- if (uv < 0x2000000000)
+#ifdef HAS_QUAD
+ if (uv < 0x1000000000LL)
#endif
{
*d++ = 0xfe; /* Can't match U+FEFF! */
*d++ = (( uv & 0x3f) | 0x80);
return d;
}
-#ifdef Quad_t
+#ifdef HAS_QUAD
{
*d++ = 0xff; /* Can't match U+FFFE! */
+ *d++ = 0x80; /* 6 Reserved bits */
+ *d++ = (((uv >> 60) & 0x0f) | 0x80); /* 2 Reserved bits */
+ *d++ = (((uv >> 54) & 0x3f) | 0x80);
+ *d++ = (((uv >> 48) & 0x3f) | 0x80);
+ *d++ = (((uv >> 42) & 0x3f) | 0x80);
*d++ = (((uv >> 36) & 0x3f) | 0x80);
*d++ = (((uv >> 30) & 0x3f) | 0x80);
*d++ = (((uv >> 24) & 0x3f) | 0x80);
#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;
}
if (!(uv & 0x40)) {
- warn("Malformed UTF-8 character");
+ dTHR;
+ if (ckWARN_d(WARN_UTF8))
+ Perl_warner(aTHX_ WARN_UTF8, "Malformed UTF-8 character");
if (retlen)
*retlen = 1;
return *s;
else if (!(uv & 0x08)) { len = 4; uv &= 0x07; }
else if (!(uv & 0x04)) { len = 5; uv &= 0x03; }
else if (!(uv & 0x02)) { len = 6; uv &= 0x01; }
- else if (!(uv & 0x01)) { len = 7; uv &= 0x00; }
- else len = 8; /* whoa! */
+ else if (!(uv & 0x01)) { len = 7; uv = 0; }
+ else { len = 13; uv = 0; } /* whoa! */
if (retlen)
*retlen = len;
s++;
while (len--) {
if ((*s & 0xc0) != 0x80) {
- warn("Malformed UTF-8 character");
+ dTHR;
+ if (ckWARN_d(WARN_UTF8))
+ Perl_warner(aTHX_ WARN_UTF8, "Malformed UTF-8 character");
if (retlen)
*retlen -= len + 1;
return 0xfffd;
return uv;
}
-/* utf8_distance(a,b) is intended to be a - b in pointer arithmetic */
+/* utf8_distance(a,b) returns the number of UTF8 characters between
+ the pointers a and b */
I32
Perl_utf8_distance(pTHX_ U8 *a, U8 *b)
return s;
}
-/* XXX NOTHING CALLS THE FOLLOWING TWO ROUTINES YET!!! */
/*
- * Convert native or reversed UTF-16 to UTF-8.
+=for apidoc Am|U8 *|utf8_to_bytes|U8 *s|STRLEN *len
+
+Converts a string C<s> of length C<len> from UTF8 into byte encoding.
+Unlike C<bytes_to_utf8>, this over-writes the original string, and
+updates len to contain the new length.
+Returns zero on failure leaving the string and len unchanged
+
+=cut
+*/
+
+U8 *
+Perl_utf8_to_bytes(pTHX_ U8* s, STRLEN *len)
+{
+ dTHR;
+ U8 *send;
+ U8 *d;
+ U8 *save;
+
+ send = s + *len;
+ d = save = s;
+
+ /* ensure valid UTF8 and chars < 256 before updating string */
+ while (s < send) {
+ U8 c = *s++;
+ if (c >= 0x80 &&
+ ( (s >= send) || ((*s++ & 0xc0) != 0x80) || ((c & 0xfe) != 0xc2)))
+ return 0;
+ }
+ s = save;
+ while (s < send) {
+ if (*s < 0x80)
+ *d++ = *s++;
+ else {
+ I32 ulen;
+ *d++ = (U8)utf8_to_uv(s, &ulen);
+ s += ulen;
+ }
+ }
+ *d = '\0';
+ *len = d - save;
+ 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 (big-endian) or reversed (little-endian) 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. */
U8*
-Perl_utf16_to_utf8(pTHX_ U16* p, U8* d, I32 bytelen)
+Perl_utf16_to_utf8(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen)
{
- U16* pend = p + bytelen / 2;
+ U8* pend;
+ U8* dstart = d;
+
+ if (bytelen & 1)
+ Perl_croak(aTHX_ "panic: utf16_to_utf8: odd bytelen");
+
+ pend = p + bytelen;
+
while (p < pend) {
- UV uv = *p++;
+ UV uv = (p[0] << 8) + p[1]; /* UTF-16BE */
+ p += 2;
if (uv < 0x80) {
*d++ = uv;
continue;
continue;
}
if (uv >= 0xd800 && uv < 0xdbff) { /* surrogates */
- int low = *p++;
- if (low < 0xdc00 || low >= 0xdfff) {
- warn("Malformed UTF-16 surrogate");
- p--;
- uv = 0xfffd;
- }
+ dTHR;
+ UV low = *p++;
+ if (low < 0xdc00 || low >= 0xdfff)
+ Perl_croak(aTHX_ "Malformed UTF-16 surrogate");
uv = ((uv - 0xd800) << 10) + (low - 0xdc00) + 0x10000;
}
if (uv < 0x10000) {
continue;
}
}
+ *newlen = d - dstart;
return d;
}
/* Note: this one is slightly destructive of the source. */
U8*
-Perl_utf16_to_utf8_reversed(pTHX_ U16* p, U8* d, I32 bytelen)
+Perl_utf16_to_utf8_reversed(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen)
{
U8* s = (U8*)p;
U8* send = s + bytelen;
s[1] = tmp;
s += 2;
}
- return utf16_to_utf8(p, d, bytelen);
+ return utf16_to_utf8(p, d, bytelen, newlen);
}
/* for now these are all defined (inefficiently) in terms of the utf8 versions */
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[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[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[UTF8_MAXLEN];
+ uv_to_utf8(tmpbuf, (UV)c);
+ return is_utf8_cntrl(tmpbuf);
+}
+
+bool
+Perl_is_uni_graph(pTHX_ U32 c)
+{
+ 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[UTF8_MAXLEN];
+ uv_to_utf8(tmpbuf, (UV)c);
+ return is_utf8_punct(tmpbuf);
+}
+
+bool
+Perl_is_uni_xdigit(pTHX_ U32 c)
+{
+ 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_uni_alnumc_lc(pTHX_ U32 c)
+{
+ return is_uni_alnumc(c); /* XXX no locale support yet */
+}
+
+bool
Perl_is_uni_idfirst_lc(pTHX_ U32 c)
{
return is_uni_idfirst(c); /* XXX no locale support yet */
}
bool
+Perl_is_uni_ascii_lc(pTHX_ U32 c)
+{
+ return is_uni_ascii(c); /* XXX no locale support yet */
+}
+
+bool
Perl_is_uni_space_lc(pTHX_ U32 c)
{
return is_uni_space(c); /* XXX no locale support yet */
}
bool
+Perl_is_uni_cntrl_lc(pTHX_ U32 c)
+{
+ return is_uni_cntrl(c); /* XXX no locale support yet */
+}
+
+bool
+Perl_is_uni_graph_lc(pTHX_ U32 c)
+{
+ return is_uni_graph(c); /* XXX no locale support yet */
+}
+
+bool
Perl_is_uni_print_lc(pTHX_ U32 c)
{
return is_uni_print(c); /* XXX no locale support yet */
}
+bool
+Perl_is_uni_punct_lc(pTHX_ U32 c)
+{
+ return is_uni_punct(c); /* XXX no locale support yet */
+}
+
+bool
+Perl_is_uni_xdigit_lc(pTHX_ U32 c)
+{
+ return is_uni_xdigit(c); /* XXX no locale support yet */
+}
+
U32
Perl_to_uni_upper_lc(pTHX_ U32 c)
{
return to_uni_lower(c); /* XXX no locale support yet */
}
-
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);
+/* return is_utf8_alpha(p) || is_utf8_digit(p); */
+#ifdef SURPRISINGLY_SLOWER /* probably because alpha is usually true */
+ if (!PL_utf8_alnum)
+ PL_utf8_alnum = swash_init("utf8", "",
+ sv_2mortal(newSVpv("+utf8::IsAlpha\n+utf8::IsDigit\n005F\n",0)), 0, 0);
+ return swash_fetch(PL_utf8_alnum, p);
+#endif
+}
+
+bool
Perl_is_utf8_idfirst(pTHX_ U8 *p)
{
return *p == '_' || is_utf8_alpha(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);
PL_curcop->op_private = PL_hints;
}
if (!SvROK(retval) || SvTYPE(SvRV(retval)) != SVt_PVHV)
- croak("SWASHNEW didn't return an HV ref");
+ Perl_croak(aTHX_ "SWASHNEW didn't return an HV ref");
return retval;
}
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;
svp = hv_store(hv, (char*)ptr, klen, retval, 0);
if (!svp || !(tmps = (U8*)SvPV(*svp, slen)) || slen < 8)
- croak("SWASHGET didn't return result of proper length");
+ Perl_croak(aTHX_ "SWASHGET didn't return result of proper length");
}
PL_last_swash_hv = hv;
off <<= 2;
return (tmps[off] << 24) + (tmps[off+1] << 16) + (tmps[off+2] << 8) + tmps[off + 3] ;
}
- croak("panic: swash_fetch");
+ Perl_croak(aTHX_ "panic: swash_fetch");
return 0;
}