mirrors/cosmopolitan
1
0
Fork 0
mirror of https://github.com/jart/cosmopolitan.git synced 2025-07-04 18:28:30 +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

60
libc/str/tprecode16to8.c
View file

@ -18,38 +18,62 @@
│ 02110-1301 USA │
╚─────────────────────────────────────────────────────────────────────────────*/
#include "libc/conv/conv.h"
#include "libc/intrin/packsswb.h"
#include "libc/intrin/pandn.h"
#include "libc/intrin/pcmpgtw.h"
#include "libc/intrin/pmovmskb.h"
#include "libc/str/str.h"
#include "libc/str/tpenc.h"
#include "libc/str/utf16.h"
static const int16_t kDel16[8] = {127, 127, 127, 127, 127, 127, 127, 127};
/**
* Transcodes UTF-16 to UTF-8.
*
* @param dst is output buffer
* @param dstsize is bytes in dst
* @param src is NUL-terminated UTF-16 input string
* @return number of bytes written excluding NUL
* @return ax bytes written excluding nul
* @return dx index of character after nul word in src
*/
size_t tprecode16to8(char *dst, size_t dstsize, const char16_t *src) {
size_t i;
axdx_t tprecode16to8(char *dst, size_t dstsize, const char16_t *src) {
axdx_t r;
uint64_t w;
wint_t x, y;
i = 0;
if (dstsize) {
for (;;) {
if (!(x = *src++)) break;
if (IsUtf16Cont(x)) continue;
if (!IsUcs2(x)) {
if (!(y = *src++)) break;
x = MergeUtf16(x, y);
}
w = tpenc(x);
while (w && i + 1 < dstsize) {
dst[i++] = w & 0xFF;
w >>= 8;
int16_t v1[8], v2[8], v3[8], vz[8];
r.ax = 0;
r.dx = 0;
for (;;) {
if (!IsTiny() && !((uintptr_t)(src + r.dx) & 15)) {
/* 10x speedup for ascii */
memset(vz, 0, 16);
while (r.ax + 8 < dstsize) {
memcpy(v1, src + r.dx, 16);
pcmpgtw(v2, v1, vz);
pcmpgtw(v3, v1, kDel16);
pandn((void *)v2, (void *)v3, (void *)v2);
if (pmovmskb((void *)v2) != 0xFFFF) break;
packsswb((void *)v1, v1, v1);
memcpy(dst + r.ax, v1, 8);
r.ax += 8;
r.dx += 8;
}
}
dst[i] = 0;
if (!(x = src[r.dx++])) break;
if (IsUtf16Cont(x)) continue;
if (!IsUcs2(x)) {
if (!(y = src[r.dx++])) break;
x = MergeUtf16(x, y);
}
w = tpenc(x);
while (w && r.ax + 1 < dstsize) {
dst[r.ax++] = w & 0xFF;
w >>= 8;
}
}
return i;
if (r.ax < dstsize) {
dst[r.ax] = 0;
}
return r;
}