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)
{
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;