Input problem when using user-defined functions.

I'm scratching my head over this one. My input for feet is always recorded as 0 when I print within the calc() function, but it is correct when printed in the main() function. What's going on?

Code:

#include <stdio.h>

float input (void)
{
    float x;

    scanf("%f", &x);
    getchar();

    return x;
}

float calc ()
{
    float y, feet, inches;

    y = 30.48 * feet + 30.48 * (inches / 12);

    printf("%f and %f", feet, inches);
    printf("%f", y);

    return y;
}

void output ()
{
    int meters;
    float centimeters;

    printf("\nThe value is equivalent to %d meters and %.2f centimeters.", meters, centimeters);
}

int main (void)
{
    int meters;
    float feet, inches, metric, centimeters;
    char ans;

    printf("This program converts a length in feet and inches into meters and centimeters.\n");

do
{
    printf("Insert length in feet: ");
    feet = input ();
    printf("Insert length in inches: ");
    inches = input ();

    metric = calc (feet, inches);

    meters = metric / 100;
    centimeters = (metric - meters) * 100;

    output (meters, centimeters);

    printf("\nWould you like to try again? [Y/N]");
    scanf("%c", &ans);
    getchar();

} while (ans == 'y' || ans == 'Y');

    return 0;
}

In the calc function, the local variables feet and inches were not given an initial value before use.

I'm a bit confused, calc is a function which takes no parameters but your calling it with two parameters from main?

ssharish

Thanks! When do we need to initialize the values? Other functions seem to work fine without initialization.

Originally Posted by ssharish2005

I'm a bit confused, calc is a function which takes no parameters but your calling it with two parameters from main?

ssharish

From what I read, using () implies unknown number of arguments, whereas having no argument is denoted by (void).

Correct me if I'm wrong.

Originally Posted by 843

From what I read, using () implies unknown number of arguments, whereas having no argument is denoted by (void).

Yeah, but here you have a function definition, so the number of parameters is known to be zero. If you used those empty parentheses with a function declaration that was not also a function definition, then indeed the number of parameters would be unknown at that point.

I initialized all the values and rearrange some functions but now I'm getting a bunch of 'incompatible' error messages.

Code:

#include <stdio.h>

float input (void)
{
    float x;

    scanf("%f", &x);
    getchar();

    return x;
}

float calc (float feet, float inches)
{
    return 30.48 * feet + 30.48 * (inches / 12);
}

int meters (float metric)
{
    return metric / 100;
}

float centimeters (float metric, int meters)
{
    return (metric - meters) * 100;
}

void output (int meters, float centimeters)
{
    printf("\nThe value is equivalent to %d meters and %.2f centimeters.", meters, centimeters);
}

int main (void)
{
    int m;
    float feet, inches, metric, cm;
    char ans;

    printf("This program converts a length in feet and inches into meters and centimeters.\n");

do
{
    printf("Insert length in feet: ");
    feet = input ();
    printf("Insert length in inches: ");
    inches = input ();

    metric = calc (feet, inches);

    m = meters (metric);
    cm = centimeters (metric, meters);

    output (meters, centimeters);

    printf("\nWould you like to try again? [Y/N]");
    scanf("%c", &ans);
    getchar();

} while (ans == 'y' || ans == 'Y');

    return 0;
}

EDIT: Ah, I see the problem now...

Originally Posted by laserlight

...function declaration that was not also a function definition...

Could you please clarify this part?

>From what I read, using () implies unknown number of arguments, whereas having no argument is denoted by (void).
Yeah i know that and your right.

This is interesting, the compiler wouldn't complain if you send parameters to a function which dosn't take any parameters.

ssharish

Got it! I didn't realize the functions' arguments are private to those functions only.

Code:

#include <stdio.h>

float input (void)
{
    float x;

    scanf("%f", &x);
    getchar();

    return x;
}

float calc (float x, float y)
{
    return 30.48 * x + 30.48 * (y / 12);
}

int meters (float x)
{
    return x / 100;
}

float centimeters (float x, int y)
{
    return x - y * 100;
}

void output (int x, float y)
{
    printf("\nThe value is equivalent to %d meters and %.2f centimeters.", x, y);
}

int main (void)
{
    int m;
    float feet, inches, metric, cm;
    char ans;

    printf("This program converts a length in feet and inches into meters and centimeters.\n");

do
{
    printf("Insert length in feet: ");
    feet = input ();
    printf("Insert length in inches: ");
    inches = input ();

    metric = calc (feet, inches);

    m = meters (metric);
    cm = centimeters (metric, m);

    output (m, cm);

    printf("\nWould you like to try again? [Y/N]");
    scanf("%c", &ans);
    getchar();

} while (ans == 'y' || ans == 'Y');

    return 0;
}

Originally Posted by 843

Could you please clarify this part?

Take your most recent code as an example. You could have written:

Code:

float calc();

/* ... */

int main(void)
{
    /* ... */

    metric = calc(feet, inches);

    /* ... */
}

/* ... */

float calc(float x, float y)
{
    return 30.48 * x + 30.48 * (y / 12);
}

But then you might as well just leave the parameters in the function prototype (which is a declaration that is not also a definition).

What's the point of a prototype since you can just declare the function and its definition?

Originally Posted by 843

What's the point of a prototype since you can just declare the function and its definition?

You may wish to use the function across different source files, so its prototype may go into a header file.