4-byte char from hex to float

[MrL0co]

he there,
I am trying to read out a .md3 file.
i am stuck on a point where i have to convert a 8-bit hexedecimal char to a float.
and change it from little endian to big endian

this is how far i got

    unsigned char tmp[4];
    f32 test;
 
    fread(&tmp, 1, 4, f);
    test = bitshiftf32(tmp);
 
f32 bitshiftf32(unsigned char val[4]) {
    return val[0]|(val[1]<<8)|(val[2]<<16)|(val[3]<<24);
    }
 
 

f being the file it readss out
test being the end result

[newb16]

As c/c++ standart doesn't mandate use of IEEE-754 32-bit floating-point for 'float' you probably have to convert it by yourself - get mantissa, add missing '1' bit, multiply 2 to the power of exponent.
Or, assuming that float is IEEE-754 32 bit, just
unsigned char val[4];
float x = *(float*)val;

[donbock]

Does the MD3 specification tell you how to interpret the 4-byte floating-point value? If so, then you should write code that explicitly decodes it accordingly; rather than casting the byte array into a native type. The casting trick might work for you, but it is brittle -- it might stop working when you upgrade to the next version of your compiler.

[weaknessforcats]

f32 bitshiftf32(unsigned char val[4]) { 
    return val[0]|(val[1]<<8)|(val[2]<<16)|(val[3]<<24); 
    } 

val is an unsigned char. Probably 8 bits. So you can shift it left up to 7 places. What does shifting it left 16 or 24 do?

[MrL0co]

tnx newb16 that works