mirrors/cosmopolitan
1
0
Fork 0
mirror of https://github.com/jart/cosmopolitan.git synced 2025-06-27 14:58:30 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

isutf8: implement RFC 3629

Browse source 
reject surrogate pairs (U+D800 to U+DFFF)
reject greater than U+10FFFF
This commit is contained in:
wingdeans 2023-12-28 15:41:29 -05:00
parent 8a8fc9a65f
commit a39caefc8b
1 changed files with 15 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

18
libc/str/isutf8.c
View file

@ -27,8 +27,8 @@ static const char kUtf8Dispatch[] = {
1, 1, 1, 1, 1, 1, 1, 1, // 0320 1, 1, 1, 1, 1, 1, 1, 1, // 0320
1, 1, 1, 1, 1, 1, 1, 1, // 0330 1, 1, 1, 1, 1, 1, 1, 1, // 0330
2, 3, 3, 3, 3, 3, 3, 3, // 0340 utf8-3 2, 3, 3, 3, 3, 3, 3, 3, // 0340 utf8-3
3, 3, 3, 3, 3, 3, 3, 3, // 0350 3, 3, 3, 3, 3, 4, 3, 3, // 0350
4, 5, 5, 5, 5, 0, 0, 0, // 0360 utf8-4 5, 6, 6, 6, 7, 0, 0, 0, // 0360 utf8-4
0, 0, 0, 0, 0, 0, 0, 0, // 0370 0, 0, 0, 0, 0, 0, 0, 0, // 0370
}; };
@ -93,6 +93,7 @@ bool32 isutf8(const void *data, size_t size) {
return false; // overlong return false; // overlong
} }
// fallthrough // fallthrough
case 3:
case3: case3:
if (p + 2 <= e && // if (p + 2 <= e && //
(p[0] & 0300) == 0200 && // (p[0] & 0300) == 0200 && //
@ -103,11 +104,17 @@ bool32 isutf8(const void *data, size_t size) {
return false; // missing cont return false; // missing cont
} }
case 4: case 4:
if (p < e && (*p & 040)) {
return false; // utf-16 surrogate
}
goto case3;
case 5:
if (p < e && (*p & 0377) < 0220) { if (p < e && (*p & 0377) < 0220) {
return false; // overlong return false; // overlong
} }
// fallthrough // fallthrough
case 5: case 6:
case6:
if (p + 3 <= e && // if (p + 3 <= e && //
(((uint32_t)(p[+2] & 0377) << 030 | // (((uint32_t)(p[+2] & 0377) << 030 | //
(uint32_t)(p[+1] & 0377) << 020 | // (uint32_t)(p[+1] & 0377) << 020 | //
@ -119,6 +126,11 @@ bool32 isutf8(const void *data, size_t size) {
} else { } else {
return false; // missing cont return false; // missing cont
} }
case 7:
if (p < e && (*p & 0x3F) > 0xF) {
return false; // over limit
}
goto case6;
default: default:
__builtin_unreachable(); __builtin_unreachable();
} }