zorrooo February 2016
### %c specifier char limit capped in C, how to avoid that?

I am running into a problem when I'm trying to output the fractional part of a hexadecimal number.

Basically, with the following code:

```
float d2h(float decimal, float fraction) {
int i = 0;
char hex[17] = { '0', '1', '2', '3', '4', '5', '6', '7', '8', '9',
'A', 'B', 'C', 'D', 'E', 'F', '\0' };
int temp = (int) decimal;
int j;
for (j = 0; j < 60; j++) {
fraction = fraction - (int) fraction;
displayFractionHex[j] = hex[(int) (fraction * 16)];
fraction *= 16;
}
for (i = sizeof(displayDecimalHex) - 1; i >= 0; i--) {
displayDecimalHex[i] = hex[temp % 16];
temp /= 16;
}
int n;
if ((n = strspn(displayDecimalHex, "0")) != 0
&& displayDecimalHex[n] != '\0') // strspn() - Reference
printf("%s", &displayDecimalHex[n]);
printf(".");
for (i = 0; i < 20; i++) {
printf("%c", displayFractionHex[i]);
}
}
```

I get the output `222.74BC0000000000000000`

, and I obviously want the fractional part of the hex number to display more characters.

However, when I try change the last `printf`

statement to a `%s`

specifier instead of `%c`

, my program crashes.

I have attempted copying the array to another one, unfortunately to no avail.

OP commented:

expected output is something along the lines of 222.74bc6a7ef9d...

Typical `float`

has about 24 binary digits of precision as does `222.74BC000000...`

".

For higher precision use, `double`

.

Use `printf("%a\n%a\n", (float) 546.456, 546.456);`

for example to see the difference.

```
0x1.113a5ep+9
0x1.113a5e353f7cfp+9
```

To be clear, "thus the decimal float is 546 and the fractional float is 0.456" is incorrect. The input to `d2h(float decimal)`

was *not* `546.456`

, but the closest representable `float`

to it exactly `546.4559936523437500...`

and *that* `float`

is exactly `222.74BC`

in hex FP.

Asked in February 2016

Viewed 1,485 times

Voted 6

Answered 1 times

Viewed 1,485 times

Voted 6

Answered 1 times