[SOLVED] Beginng C programming - how to input decimals from keyboard and calcuate as floats?

keithostertag · 02-17-2012, 10:21 AM

I've been modifying this basic conversion formula program to study basic data types. I can't seem to figure out how to properly input decimals (like 2.5) from the keyboard and use those values. Here's one version of the program:

Code:

#include<stdio.h>
#include <stdlib.h>
void fahr_celsius(float, float, float);

int main(int argc, char *argv[]) {
int i;
if (argc == 4){
	for (i=1;i<argc;i++)
		printf("Argument %d: %s\n", i, argv[i]);

puts("\nFarhenheit-Celsius conversion table\n");
fahr_celsius(atol(argv[1]), atol(argv[2]), atol(argv[3]));
}
else
puts("\nWrong number or type arguments!!\n");
return 0; 
}

void fahr_celsius(float min, float max, float step) {
float fahr;

printf("min = %3.2f max = %3.2f step =%3.2f\n", min, max, step);
/* for (fahr = min; fahr <= max; fahr +=step) 
	printf("%3.2f %8.2f\n", fahr, (5./9.) * (fahr - 32.0)); */
	putchar('\n');
}

Notice I've commented out the calculation until I figure out how to pass the values properly.. and there's no input checking yet, etc.

Here's what the output looks like:

Code:

$convrt 0.8 20.6 2.5

Argument 1: 0.8
Argument 2: 20.6
Argument 3: 2.5

Farhenheit-Celsius conversion table

min = 0.00 max = 20.00 step =2.00

There doesn't seem to be a ASCII-to-float conversion function available, or is there? What's the proper way to do this? Should the main prototype be

Code:

int main(argc, *float argv[])

instead, or some other simliar strategy?

Bonus question: when I get around to checking the input for valid arguments (no lettters or other non-numeric characters), should I check each character using getc or each string using string functions?

Thanks,
Keith Ostertag

Guttorm · 02-17-2012, 10:39 AM

Hi

You're close. "atol" converts to long and "atof" converts to double. You can cast it to float like this:

Code:

float f = (float) atof(argv[1]);

Edit:
Also, if reading from files/strings/input you can use fscanf, sscanf or scanf. You can think of it as the "opposite" of printf. It has the advantage that it returns the number of successful conversions, while atof does not detect errors. It also has lots of other options. See man "scanf".

Code:

float f;
if (sscanf(argv[1], "%f", &f) != 1) {
  printf("Not a number");
}

keithostertag · 02-17-2012, 10:51 AM

Thanks Guttorm- yes, that works. The call to the function now reads:

Code:

fahr_celsius((float) atof(argv[1]), (float) atof(argv[2]), (float) atof(argv[3]));

and it produces the expected output. I'll have to ponder that a bit...

Is that the best way to approach it? Seems a bit crude... I wonder if there is a more direct or simpler way? I'm a rank beginner, so I wouldn't know...

Keith

EDIT: Actually it appears the cast isn't necessary?

Code:

fahr_celsius(atof(argv[1]), atof(argv[2]), atof(argv[3]));

works as well.

Guttorm · 02-17-2012, 11:07 AM

Hi

I think the float casting is not really necessary, but the compiler can make a warning because you lose precision when converting from double to float. Maybe simpler to just use double instead of float? Then you get more precision as well. You can use sscanf if you want to check for invalid input.

keithostertag · 02-17-2012, 11:11 AM

OK, thanks. I'll play with it some more. Marked as solved.

Keith