mirrors/cosmopolitan
1
0
Fork 0
mirror of https://github.com/jart/cosmopolitan.git synced 2025-07-02 17:28:30 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Make improvements

Browse source
This commit is contained in:
Justine Tunney 2020-12-01 03:43:40 -08:00
parent 3e4fd4b0ad
commit e44a0cf6f8
256 changed files with 23100 additions and 2294 deletions
Download patch file Download diff file Expand all files Collapse all files

72
libc/str/tprecode8to16.c
View file

@ -17,6 +17,10 @@
│ Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA │
│ 02110-1301 USA │
╚─────────────────────────────────────────────────────────────────────────────*/
#include "libc/intrin/pcmpgtb.h"
#include "libc/intrin/pmovmskb.h"
#include "libc/intrin/punpckhbw.h"
#include "libc/intrin/punpcklbw.h"
#include "libc/str/str.h"
#include "libc/str/thompike.h"
#include "libc/str/utf16.h"
@ -27,34 +31,56 @@
* @param dst is output buffer
* @param dstsize is shorts in dst
* @param src is NUL-terminated UTF-8 input string
* @return number of shorts written excluding NUL
* @return ax shorts written excluding nul
* @return dx index of character after nul word in src
*/
size_t tprecode8to16(char16_t *dst, size_t dstsize, const char *src) {
size_t i;
axdx_t tprecode8to16(char16_t *dst, size_t dstsize, const char *src) {
axdx_t r;
unsigned n;
uint64_t w;
wint_t x, y;
i = 0;
if (dstsize) {
for (;;) {
if (!(x = *src++ & 0xff)) break;
if (ThomPikeCont(x)) continue;
if (!isascii(x)) {
n = ThomPikeLen(x);
x = ThomPikeByte(x);
while (--n) {
if (!(y = *src++ & 0xff)) goto stop;
x = ThomPikeMerge(x, y);
}
}
w = EncodeUtf16(x);
while (w && i + 1 < dstsize) {
dst[i++] = w & 0xFFFF;
w >>= 16;
uint8_t v1[16], v2[16], vz[16];
r.ax = 0;
r.dx = 0;
for (;;) {
if (!IsTiny() && !((uintptr_t)(src + r.dx) & 15)) {
/* 34x speedup for ascii */
memset(vz, 0, 16);
while (r.ax + 16 < dstsize) {
memcpy(v1, src + r.dx, 16);
pcmpgtb((int8_t *)v2, (int8_t *)v1, (int8_t *)vz);
if (pmovmskb(v2) != 0xFFFF) break;
punpcklbw(v2, v1, vz);
punpckhbw(v1, v1, vz);
memcpy(dst + r.ax + 0, v2, 16);
memcpy(dst + r.ax + 8, v1, 16);
r.ax += 16;
r.dx += 16;
}
}
stop:
dst[i] = 0;
x = src[r.dx++] & 0xff;
if (ThomPikeCont(x)) continue;
if (!isascii(x)) {
n = ThomPikeLen(x);
x = ThomPikeByte(x);
while (--n) {
if ((y = src[r.dx++] & 0xff)) {
x = ThomPikeMerge(x, y);
} else {
x = 0;
break;
}
}
}
if (!x) break;
w = EncodeUtf16(x);
while (w && r.ax + 1 < dstsize) {
dst[r.ax++] = w & 0xFFFF;
w >>= 16;
}
}
return i;
if (r.ax < dstsize) {
dst[r.ax] = 0;
}
return r;
}