mask a random 64 bit integer to a floating point range

randomFP.c

/*
    random64() is a random integer function
    this code uses undefined behavior but it'll be alright
    make sure to include stdint
*/
#include <stdint.h>

// mask to (-2,2] 
// this is just an example, change the exponent bits for other ranges
inline double _22() {
    uint64_t u64 = 0x4010000000000001ULL | (random64() >> 12);
    return *(double *)&u64 - 6.0;
}

// mask to [0,1)
inline double _01() {
    uint64_t u64 = 0x3FF0000000000000ULL | (random64() >> 12);
    return *(double *)&u64 - 1.0;
}