UV
Perl_to_uni_upper(pTHX_ UV c, U8* p, STRLEN *lenp)
{
- U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
- uvchr_to_utf8(tmpbuf, c);
- return to_utf8_upper(tmpbuf, p, lenp);
+ uvchr_to_utf8(p, c);
+ return to_utf8_upper(p, p, lenp);
}
UV
Perl_to_uni_title(pTHX_ UV c, U8* p, STRLEN *lenp)
{
- U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
- uvchr_to_utf8(tmpbuf, c);
- return to_utf8_title(tmpbuf, p, lenp);
+ uvchr_to_utf8(p, c);
+ return to_utf8_title(p, p, lenp);
}
UV
Perl_to_uni_lower(pTHX_ UV c, U8* p, STRLEN *lenp)
{
- U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
- uvchr_to_utf8(tmpbuf, c);
- return to_utf8_lower(tmpbuf, p, lenp);
+ uvchr_to_utf8(p, c);
+ return to_utf8_lower(p, p, lenp);
}
UV
Perl_to_uni_fold(pTHX_ UV c, U8* p, STRLEN *lenp)
{
- U8 tmpbuf[UTF8_MAXLEN_FOLD+1];
- uvchr_to_utf8(tmpbuf, c);
- return to_utf8_fold(tmpbuf, p, lenp);
+ uvchr_to_utf8(p, c);
+ return to_utf8_fold(p, p, lenp);
}
/* for now these all assume no locale info available for Unicode > 255 */
conversion result to. The "lenp" is a pointer to the length
of the result.
-The "swash" is a pointer to the swash to use.
+The "swashp" is a pointer to the swash to use.
-The "normal" is a string like "ToLower" which means the swash
-$utf8::ToLower, which is stored in lib/unicore/To/Lower.pl,
-and loaded by SWASHGET, using lib/utf8_heavy.pl.
+Both the special and normal mappings are stored lib/unicore/To/Foo.pl,
+and loaded by SWASHGET, using lib/utf8_heavy.pl. The special (usually,
+but not always, a multicharacter mapping), is tried first.
-The "special" is a string like "utf8::ToSpecLower", which means
-the hash %utf8::ToSpecLower, which is stored in the same file,
-lib/unicore/To/Lower.pl, and also loaded by SWASHGET. The access
-to the hash is by Perl_to_utf8_case().
+The "special" is a string like "utf8::ToSpecLower", which means the
+hash %utf8::ToSpecLower. The access to the hash is through
+Perl_to_utf8_case().
-=cut
- */
+The "normal" is a string like "ToLower" which means the swash
+%utf8::ToLower.
+
+=cut */
UV
Perl_to_utf8_case(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp, SV **swashp, char *normal, char *special)
{
- UV uv0, uv1, uv2;
- U8 tmpbuf[UTF8_MAXLEN_FOLD+1], *d;
- char *s = NULL;
- STRLEN len;
- bool has_utf8 = FALSE;
+ UV uv0, uv1;
+ U8 tmpbuf[UTF8_MAXLEN_FOLD+1];
+ STRLEN len = 0;
- if (!*swashp)
- *swashp = swash_init("utf8", normal, &PL_sv_undef, 4, 0);
uv0 = utf8_to_uvchr(p, 0);
/* The NATIVE_TO_UNI() and UNI_TO_NATIVE() mappings
* are necessary in EBCDIC, they are redundant no-ops
* in ASCII-ish platforms, and hopefully optimized away. */
uv1 = NATIVE_TO_UNI(uv0);
uvuni_to_utf8(tmpbuf, uv1);
- uv2 = swash_fetch(*swashp, tmpbuf, TRUE);
- if (uv2) {
- /* It was "normal" (a single character mapping). */
- d = uvuni_to_utf8(ustrp, uv2);
- has_utf8 = !UNI_IS_INVARIANT(uv2);
- }
- else {
- /* It might be "special" (sometimes, but not always,
+
+ if (!*swashp) /* load on-demand */
+ *swashp = swash_init("utf8", normal, &PL_sv_undef, 4, 0);
+
+ if (special) {
+ /* It might be "special" (sometimes, but not always,
* a multicharacter mapping) */
HV *hv;
SV *keysv;
HE *he;
SV *val;
-
+
if ((hv = get_hv(special, FALSE)) &&
(keysv = sv_2mortal(Perl_newSVpvf(aTHX_ "%04"UVXf, uv1))) &&
(he = hv_fetch_ent(hv, keysv, FALSE, 0)) &&
(val = HeVAL(he))) {
+ char *s;
s = SvPV(val, len);
if (len == 1)
- d = uvuni_to_utf8(ustrp, NATIVE_TO_UNI(*(U8*)s));
+ len = uvuni_to_utf8(ustrp, NATIVE_TO_UNI(*(U8*)s)) - ustrp;
else {
+#ifdef EBCDIC
+ /* If we have EBCDIC we need to remap the characters
+ * since any characters in the low 256 are Unicode
+ * code points, not EBCDIC. */
+ U8 *t = (U8*)s, *tend = t + len, *d;
+
+ d = tmpbuf;
+ if (SvUTF8(val)) {
+ STRLEN tlen = 0;
+
+ while (t < tend) {
+ UV c = utf8_to_uvchr(t, &tlen);
+ if (tlen > 0) {
+ d = uvchr_to_utf8(d, UNI_TO_NATIVE(c));
+ t += tlen;
+ }
+ else
+ break;
+ }
+ }
+ else {
+ while (t < tend) {
+ d = uvchr_to_utf8(d, UNI_TO_NATIVE(*t));
+ t++;
+ }
+ }
+ len = d - tmpbuf;
+ Copy(tmpbuf, ustrp, len, U8);
+#else
Copy(s, ustrp, len, U8);
- d = ustrp + len;
+#endif
}
- if (SvUTF8(val))
- has_utf8 = TRUE;
- }
- else {
- /* It was not "special", either. */
- d = uvuni_to_utf8(ustrp, uv1);
- has_utf8 = !UNI_IS_INVARIANT(uv1);
}
}
- len = d - ustrp;
-
-#ifdef EBCDIC
- {
- /* If we have EBCDIC we need to remap the characters since
- * any characters in the low 256 are in Unicode code points,
- * not EBCDIC. */
- U8 *t, *tend;
+ if (!len && *swashp) {
+ UV uv2 = swash_fetch(*swashp, tmpbuf, TRUE);
- d = tmpbuf;
- if (has_utf8) {
- STRLEN tlen = 0;
+ if (uv2) {
+ /* It was "normal" (a single character mapping). */
+ UV uv3 = UNI_TO_NATIVE(uv2);
- for (t = ustrp, tend = t + len;
- t < tend; t += tlen) {
- UV c = utf8_to_uvchr(t, &tlen);
-
- if (tlen > 0)
- d = uvchr_to_utf8(d, UNI_TO_NATIVE(c));
- else
- break;
- }
- } else {
- for (t = ustrp, tend = t + len;
- t < tend; t++)
- d = uvchr_to_utf8(d, UNI_TO_NATIVE(*t));
+ len = uvchr_to_utf8(ustrp, uv3) - ustrp;
}
- len = d - tmpbuf;
- Copy(tmpbuf, ustrp, len, U8);
}
-#endif
+
+ if (!len) /* Neither: just copy. */
+ len = uvchr_to_utf8(ustrp, uv0) - ustrp;
if (lenp)
*lenp = len;
- return utf8_to_uvchr(ustrp, 0);
+ return len ? utf8_to_uvchr(ustrp, 0) : 0;
}
/*
projects / p5sagit/p5-mst-13.2.git / blobdiff