Thread: How do I check if an input file exist, if the numbers in the file are int or fl

First off. This is my first c project (and one of the first programming projects in any language), so if my syntax looks strange or I do some strange stuff I appolgize.

I have a program where three input files are used. Each file contains a few numbers, some are integers and some are floats. I've managed to check if the files exist, but not the other stuff. If any of these demands are not met the program should end and display an error message accordingly. I've seen questions around asking one one of these things, but not combined. The answers I have seen also demands quit a bit of space. I was hoping to get this done without on as few lines as possible. Help is much appreciated.

One of the scripts I've been working on can be seen below. It's gives an error message if the input file is not there, but I have removed my attempts at checking for int vs. float and if they are within range.

-------------------------------------------------------------------------------------------

Code:

void visualreader(int *x_range, int *y_range, int *z_range, int *x_cells, int *y_cells, int *z_cells, int *time){

//Scans the input file containing parameters for the visualizing the velocity vector field
FILE *fp = fopen("visual_input.dat", "r");

        if(fp){
                fscanf(fp, "%d %d %d\n", x_range, y_range, z_range);
                fscanf(fp, "%d %d %d\n",x_cells, y_cells, z_cells);
                fscanf(fp, "%d\n",time);

                fclose( fp );}
        
        else{
                perror("error opening the file" );
        }
}

There is no "read this file and tell me if it's a float or an int" function, but there are some ways that aren't too bad. I'm assuming the file is a text file and the numbers are separated by spaces, something like

Code:

12 345.67 890
654 32.10 9.87

If that's the case, you can use the %s modifier with fscanf. That reads a "string". Perhaps "word" would be a better description, as it reads everything up to a white space. You must use a char array to store the result. Then, I would use the strtol (string to long) and strtof (string to float) functions to try and convert what you read into a number. Unless specified otherwise, I would actually use double instead of float for your program as it is more precise. If you do use double, use strtod instead of strtof. Something like:

Code:

char *endp;  // this is for strtol and strtof
char buf[32];  // this stores the string we got from scanf
errno = 0;
x = strtol(buf, &endp, 10);  // the 10 is for base 10 numbers.  strtol can handle base 2-36.
if (errno == 0 && *endp == '\0') {
    // hooray! we have an integer
    // errno is zero, so strtol had some success
    // we also check that endp is pointing to the null terminator for buf
    // to ensure that strtol didn't stop early because it encountered an invalid character
    // in buf such as a decimal point, or a letter or other garbage
}

So that's how you use strtol/strtof to check for success/failure. Since integer is stricter format than float, we should check it first. That is, every valid integer is a valid float, but not vice-versa. The general flow is something like:

Code:

read a string into buf using fscanf with the "%s" modifier
if the string is a valid integer
    do some integer stuff
else
    if the string is a valid float
        do some float stuff
    else
        we have invalid data in our input