float to byte array?

What would you like in those bytes?

The x86 op code fbstp will convert a floating point type into an array of packed BCDs.

Edited by - Magmai Kai Holmlor on December 30, 2001 5:03:11 AM

bleh...i was hoping to be able to do it without asm...what im used to doing with integers is to just shift the bits and put it into a byte. since you cant do this with a float, im trying to find a way.

Example 1:

union { float F; int I; } MyIntFloat;

Whatever you set F to, I shares the same memory, so you can treat it like an integer. You could also do "unsigned char B[4];" instead of the integer (remember that x86 CPU''s are little endian). Example 2:

float F; int I = *((int *)&F);

I now equals the data within F. There are more methods, but they aren''t as good as those two.

[Resist Windows XP''s Invasive Production Activation Technology!]

what are you trying to do this for? you can do this without asm. just look up the IEEE spec for floats and doubles. its just a matter of following the spec. you can use arrays.

example PSEUDO code:

unsigned char floatArray[sizeof(float)]; float theFloat; // now you can manipulate by: memcpy(floatArray, &theFloat, sizeof(float)); // or if you are brave use type casting (a bit messy) ((unsigned char *)(&theFloat))[idx];

just remeber floating point numbers have a format to follow. so make sure you look it up. it owuld be best to explain what you need this for.

you can do the same with doubles

Edited by - a person on December 30, 2001 6:20:16 PM

thanks guys, this is just what i was looking for