Mirrors/tinyusdz
0
0
Fork 0
mirror of https://github.com/lighttransport/tinyusdz.git synced 2026-01-18 01:11:17 +01:00
Code Issues Packages Projects Releases Wiki Activity

Add utf-8 character validation.

Browse Source 
Add assetpath with utf8 string test scene.
This commit is contained in:
Syoyo Fujita
2023-08-02 19:53:40 +09:00
parent e15796fba5
commit be06996c91
2 changed files with 32 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
src/str-util.cc
View File

@@ -481,31 +481,48 @@ uint32_t to_utf8_code(const std::string &s) {
}
// TODO: endianness.
uint32_t code = 0;
if (s.size() == 1) {
unsigned char s0 = static_cast<unsigned char>(s[0]);
if (s0 > 0x7f) {
return ~0u;
}
code = uint32_t(s0) & 0x7f;
} else if (s.size() == 2) {
// 11bit: 110y-yyyx 10xx-xxxx
unsigned char s0 = static_cast<unsigned char>(s[0]);
unsigned char s1 = static_cast<unsigned char>(s[1]);
code = (uint32_t(s0 & 0x1f) << 6) | (s1 & 0x3f);
if (((s0 & 0xe0) == 0xc0) && ((s1 & 0xc0) == 0x80)) {
code = (uint32_t(s0 & 0x1f) << 6) | (s1 & 0x3f);
} else {
return ~0u;
}
} else if (s.size() == 3) {
// 16bit: 1110-yyyy 10yx-xxxx 10xx-xxxx
unsigned char s0 = static_cast<unsigned char>(s[0]);
unsigned char s1 = static_cast<unsigned char>(s[1]);
unsigned char s2 = static_cast<unsigned char>(s[2]);
code =
(uint32_t(s0 & 0xf) << 12) | (uint32_t(s1 & 0x3f) << 6) | (s2 & 0x3f);
if (((s0 & 0xf0) == 0xe0) && ((s1 & 0xc0) == 0x80) &&
((s2 & 0xc0) == 0x80)) {
code =
(uint32_t(s0 & 0xf) << 12) | (uint32_t(s1 & 0x3f) << 6) | (s2 & 0x3f);
} else {
return ~0u;
}
} else {
// 21bit: 1111-0yyy 10yy-xxxx 10xx-xxxx 10xx-xxxx
unsigned char s0 = static_cast<unsigned char>(s[0]);
unsigned char s1 = static_cast<unsigned char>(s[1]);
unsigned char s2 = static_cast<unsigned char>(s[2]);
unsigned char s3 = static_cast<unsigned char>(s[3]);
code = (uint32_t(s0 & 0x7) << 18) | (uint32_t(s1 & 0x3f) << 12) |
(uint32_t(s2 & 0x3f) << 6) | uint32_t(s3 & 0x3f);
if (((s0 & 0xf8) == 0xf0) && ((s1 & 0xc0) == 0x80) &&
((s2 & 0xc0) == 0x80) && ((s2 & 0xc0) == 0x80)) {
code = (uint32_t(s0 & 0x7) << 18) | (uint32_t(s1 & 0x3f) << 12) |
(uint32_t(s2 & 0x3f) << 6) | uint32_t(s3 & 0x3f);
} else {
return ~0u;
}
}
return code;

9
tests/usda/utf8-assetpath-001.usda Normal file
View File

@@ -0,0 +1,9 @@
#usda 1.0
def Xform "hello" (
displayName = "こんにちは"
)
{
# https://github.com/PixarAnimationStudios/OpenUSD/issues/2560
asset myasset = "響"
}