Passing content of float32 to a int32 in C

Question

Passing content of float32 to a int32 in C

Problem:

I would like to pass all the content of a float32 variable to a int32 however doing the following:

float32 source;
int32 destination;

source = 23.59463;         // Hex = 0x41BCC1CE 
destination = 0;           // Hex = 0x00000000
destination = source;      // Destination = 0x00000017 = 23

This does not pass all the content. I have tried casting, and tried to get the content indirectly but then I get issues with pointer types.

Question:

1) How can I grab all the content of the float32 into the int32?

c

embedded

asked on Stack Overflow Nov 15, 2013 by

VolT

4 Answers

If you want a byte-by-byte copy, that's

assert(sizeof(source) == sizeof(destination));
memcpy(&destination, &source, sizeof(source));

answered on Stack Overflow Nov 15, 2013 by

Fred Foo • edited Nov 15, 2013 by

Fred Foo

As an alternative answer, there's always fun with pointers:

destination = *(int32*)&source;

answered on Stack Overflow Nov 15, 2013 by

Ross

When you do the assignment destination = source the source variable is casted to an integer, and thereby truncated.

One way of overcoming this problem is to use a union:

union U
{
    float32 f;
    int32   i;
};

union U u;
u.f = 23.59463;
printf("%08x\n", u.i);

Note: If you intend to send this integer (over network or serial communication) then you have to remember that different platform have different endianness.

answered on Stack Overflow Nov 15, 2013 by

Some programmer dude

If "all the content" means all the digits, you can multiply source to get fixed point destination.

destination = source * 100000;//23.59463 * 100000 = 2359463

answered on Stack Overflow Nov 15, 2013 by

user2995769

User contributions licensed under CC BY-SA 3.0