add morphisms

- unicode
- value delimited seq
- zigzag encoding (still lacking any way to get to signed integer yet'..)
This commit is contained in:
Michael Sippel 2025-03-15 18:49:06 +01:00
parent 630948139b
commit d0118b56b1
Signed by: senvas
GPG key ID: F96CF119C34B64A6
3 changed files with 190 additions and 0 deletions

View file

@ -0,0 +1,93 @@
```
```
morph_string_as_ascii_to_utf8 ()
<Seq ~ <ValueTerminated '\0'> Char~Ascii~x86.UInt8>
--> <Seq Char~Unicode>
~ UTF-8
~ <Seq~<ValueTerminated 0> x86.UInt8>
```
while( *src ) { *dst++ = *src++; }
*dst = 0;
return 0;
```
morph_string_as_utf8_to_ascii ()
<Seq Char~Unicode>
~ UTF-8
~ <Seq~<ValueTerminated 0> x86.UInt8>
--> <Seq ~ <ValueTerminated '\0'> Char~Ascii~x86.UInt8>
```
while( *src ) {
if( *src < 128 ) {
*dst++ = *src++;
} else {
fprintf(stderr, "(morph UTF-8 to Ascii) ignore multi-byte character\n");
while( *++src >= 128 );
}
}
*dst = 0;
return 0;
```
morph_string_as_ascii_to_utf32 ()
<Seq ~ <ValueTerminated '\0'> Char~Ascii~x86.UInt8>
--> <Seq Char~Unicode>
~ UTF-32LE
~ <Seq~<ValueTerminated 0> x86.UInt32>
```
while( *src ) { *dst++ = *src++; }
*dst = 0;
return 0;
```
morph_string_as_utf8_to_utf32 ()
<Seq Char~Unicode>
~ UTF-8
~ <Seq~<ValueTerminated 0> x86.UInt8>
--> <Seq Char~Unicode>
~ UTF-32LE
~ <Seq~<ValueTerminated 0> x86.UInt32>
```
bool has_multibyte = false;
uint32_t val = 0;
while( *src ) {
uint8_t tag = (*src >> 6)&0b11;
switch( tag ) {
// single byte
case 0b00:
case 0b01:
if( has_multibyte ) {
*dst++ = val;
has_multibyte = false;
}
*dst++ = *src++;
break;
// start multibyte
case 0b11:
if( has_multibyte ) {
*dst++ = val;
}
has_multibyte = true;
val = (*src++) & 0b111111;
break;
// continue multibyte
case 0b10:
val <<= 6;
val |= (*src++) & 0b111111;
break;
}
}
if( has_multibyte )
*dst++ = val;
*dst++ = 0;
return 0;
```