#endif
#endif
-
-#define HALF_UPGRADE(start,end) { \
- U8* newstr; \
- STRLEN len; \
- len = end-start; \
- newstr = bytes_to_utf8(start, &len); \
- Copy(newstr,start,len,U8*); \
- end = start + len; \
- }
-
+#define HALF_UTF8_UPGRADE(start,end) \
+ STMT_START { \
+ U8* NeWsTr; \
+ STRLEN LeN = (end) - (start); \
+ NeWsTr = bytes_to_utf8(start, &LeN); \
+ Safefree(start); \
+ (start) = NeWsTr; \
+ (end) = (start) + LeN; \
+ } STMT_END
STATIC I32
S_do_trans_simple(pTHX_ SV *sv)
/* First, take care of non-UTF8 input strings, because they're easy */
if (!sutf) {
- while (s < send) {
+ while (s < send) {
if ((ch = tbl[*s]) >= 0) {
- matches++;
- *s++ = ch;
- } else
- s++;
- }
- SvSETMAGIC(sv);
+ matches++;
+ *s++ = ch;
+ }
+ else
+ s++;
+ }
+ SvSETMAGIC(sv);
return matches;
}
else
d = uv_to_utf8(d,ch);
s += ulen;
- } else { /* No match -> copy */
+ }
+ else { /* No match -> copy */
while (ulen--)
*d++ = *s++;
}
}
- *d='\0';
+ *d = '\0';
sv_setpvn(sv, (const char*)dstart, d - dstart);
SvUTF8_on(sv);
SvLEN_set(sv, 2*len+1);
while (s < send) {
if (hasutf && *s & 0x80)
- s+=UTF8SKIP(s);
+ s += UTF8SKIP(s);
else {
UV c;
I32 ulen;
c = *s;
if (c < 0x100 && tbl[c] >= 0)
matches++;
- s+=ulen;
+ s += ulen;
}
}
while (s < send) {
if (hasutf && *s & 0x80)
- s+=UTF8SKIP(s);
+ s += UTF8SKIP(s);
else {
if ((ch = tbl[*s]) >= 0) {
*d = ch;
else
p = d++;
}
- else if (ch == -1) /* -1 is unmapped character */
+ else if (ch == -1) /* -1 is unmapped character */
*d++ = *s; /* -2 is delete character */
s++;
}
else {
while (s < send) {
if (hasutf && *s & 0x80)
- s+=UTF8SKIP(s);
+ s += UTF8SKIP(s);
else {
if ((ch = tbl[*s]) >= 0) {
*d = ch;
matches++;
d++;
}
- else if (ch == -1) /* -1 is unmapped character */
+ else if (ch == -1) /* -1 is unmapped character */
*d++ = *s; /* -2 is delete character */
s++;
}
}
}
- matches += send - d; /* account for disappeared chars */
+ matches += send - d; /* account for disappeared chars */
*d = '\0';
SvCUR_set(sv, d - (U8*)SvPVX(sv));
SvSETMAGIC(sv);
if ((uv = swash_fetch(rv, s)) < none) {
s += UTF8SKIP(s);
matches++;
- if (uv & 0x80 && !isutf++)
- HALF_UPGRADE(dstart,d);
+ if ((uv & 0x80) && !isutf++)
+ HALF_UTF8_UPGRADE(dstart,d);
d = uv_to_utf8(d, uv);
}
else if (uv == none) {
int i;
- i = UTF8SKIP(s);
+ i = UTF8SKIP(s);
if (i > 1 && !isutf++)
- HALF_UPGRADE(dstart,d);
+ HALF_UTF8_UPGRADE(dstart,d);
while(i--)
- *d++ = *s++;
+ *d++ = *s++;
}
else if (uv == extra) {
int i;
- i = UTF8SKIP(s);
+ i = UTF8SKIP(s);
s += i;
matches++;
if (i > 1 && !isutf++)
- HALF_UPGRADE(dstart,d);
+ HALF_UTF8_UPGRADE(dstart,d);
d = uv_to_utf8(d, final);
}
else
if (uv < none) {
matches++;
if (uv != puv) {
- if (uv & 0x80 && !isutf++)
- HALF_UPGRADE(dst,d);
- d = uv_to_utf8(d, uv);
+ if ((uv & 0x80) && !isutf++)
+ HALF_UTF8_UPGRADE(dst,d);
+ d = uv_to_utf8(d, uv);
puv = uv;
}
- s += UTF8SKIP(s);
+ s += UTF8SKIP(s);
continue;
}
else if (uv == none) { /* "none" is unmapped character */
- I32 ulen;
- *d++ = (U8)utf8_to_uv(s, &ulen);
- s += ulen;
+ I32 ulen;
+ *d++ = (U8)utf8_to_uv(s, &ulen);
+ s += ulen;
puv = 0xfeedface;
continue;
}
else if (uv == extra && !del) {
matches++;
if (uv != puv) {
- d = uv_to_utf8(d, final);
+ d = uv_to_utf8(d, final);
puv = final;
}
- s += UTF8SKIP(s);
+ s += UTF8SKIP(s);
continue;
}
- matches++; /* "none+1" is delete character */
- s += UTF8SKIP(s);
+ matches++; /* "none+1" is delete character */
+ s += UTF8SKIP(s);
}
}
else {
}
if (uv < none) {
matches++;
- d = uv_to_utf8(d, uv);
- s += UTF8SKIP(s);
+ d = uv_to_utf8(d, uv);
+ s += UTF8SKIP(s);
continue;
}
else if (uv == none) { /* "none" is unmapped character */
- I32 ulen;
- *d++ = (U8)utf8_to_uv(s, &ulen);
- s += ulen;
+ I32 ulen;
+ *d++ = (U8)utf8_to_uv(s, &ulen);
+ s += ulen;
continue;
}
else if (uv == extra && !del) {
matches++;
- d = uv_to_utf8(d, final);
- s += UTF8SKIP(s);
+ d = uv_to_utf8(d, final);
+ s += UTF8SKIP(s);
continue;
}
- matches++; /* "none+1" is delete character */
- s += UTF8SKIP(s);
+ matches++; /* "none+1" is delete character */
+ s += UTF8SKIP(s);
}
}
if (dst)
switch (PL_op->op_private & ~hasutf & 63) {
case 0:
- if (hasutf)
- return do_trans_simple_utf8(sv);
- else
- return do_trans_simple(sv);
+ if (hasutf)
+ return do_trans_simple_utf8(sv);
+ else
+ return do_trans_simple(sv);
case OPpTRANS_IDENTICAL:
- if (hasutf)
- return do_trans_count_utf8(sv);
- else
- return do_trans_count(sv);
+ if (hasutf)
+ return do_trans_count_utf8(sv);
+ else
+ return do_trans_count(sv);
default:
- if (hasutf)
+ if (hasutf)
return do_trans_complex_utf8(sv);
else
return do_trans_complex(sv);
char *rsave;
bool left_utf = DO_UTF8(left);
bool right_utf = DO_UTF8(right);
+ I32 needlen;
if (left_utf && !right_utf)
sv_utf8_upgrade(right);
rsave = rc = SvPV(right, rightlen);
len = leftlen < rightlen ? leftlen : rightlen;
lensave = len;
- if (SvOK(sv) || SvTYPE(sv) > SVt_PVMG) {
+ if ((left_utf || right_utf) && (sv == left || sv == right)) {
+ needlen = optype == OP_BIT_AND ? len : leftlen + rightlen;
+ Newz(801, dc, needlen + 1, char);
+ }
+ else if (SvOK(sv) || SvTYPE(sv) > SVt_PVMG) {
STRLEN n_a;
dc = SvPV_force(sv, n_a);
if (SvCUR(sv) < len) {
dc = SvGROW(sv, len + 1);
(void)memzero(dc + SvCUR(sv), len - SvCUR(sv) + 1);
}
+ if (optype != OP_BIT_AND && (left_utf || right_utf))
+ dc = SvGROW(sv, leftlen + rightlen + 1);
}
else {
- I32 needlen = ((optype == OP_BIT_AND)
- ? len : (leftlen > rightlen ? leftlen : rightlen));
+ needlen = ((optype == OP_BIT_AND)
+ ? len : (leftlen > rightlen ? leftlen : rightlen));
Newz(801, dc, needlen + 1, char);
(void)sv_usepvn(sv, dc, needlen);
dc = SvPVX(sv); /* sv_usepvn() calls Renew() */
(void)SvPOK_only(sv);
if (left_utf || right_utf) {
UV duc, luc, ruc;
+ char *dcsave = dc;
STRLEN lulen = leftlen;
STRLEN rulen = rightlen;
- STRLEN dulen = 0;
I32 ulen;
- if (optype != OP_BIT_AND)
- dc = SvGROW(sv, leftlen+rightlen+1);
-
switch (optype) {
case OP_BIT_AND:
while (lulen && rulen) {
duc = luc & ruc;
dc = (char*)uv_to_utf8((U8*)dc, duc);
}
- dulen = dc - SvPVX(sv);
- SvCUR_set(sv, dulen);
+ if (sv == left || sv == right)
+ (void)sv_usepvn(sv, dcsave, needlen);
+ SvCUR_set(sv, dc - dcsave);
break;
case OP_BIT_XOR:
while (lulen && rulen) {
dc = (char*)uv_to_utf8((U8*)dc, duc);
}
mop_up_utf:
- dulen = dc - SvPVX(sv);
- SvCUR_set(sv, dulen);
+ if (sv == left || sv == right)
+ (void)sv_usepvn(sv, dcsave, needlen);
+ SvCUR_set(sv, dc - dcsave);
if (rulen)
sv_catpvn(sv, rc, rulen);
else if (lulen)
projects / p5sagit/p5-mst-13.2.git / blobdiff