Thread: Console application - input handling

Hello!

Yes, I'm having a little trouble with the streams. I will make myself clear before I get to my problem: If you think you need to see my code to help, then you probably can't help me!

Okay!

I need to know how to handle multiple forms of input. That being, say I need to work with integers, but the user sends in a character, and then instead of blowing up, it stops and tells the user it only accepts in integers, and then requests for input again.

I need it to use integers, because the integer is also being sent to functions to be evaluated as integers. It is unreasonable for me to test each character, and then return their integer value, so I need to just throw out character input.

I can't get it done! And, for my class project, I need it to be done. My teacher hasn't exactly been helpful.

This actually has been an ongoing problem for me to solve over the last 10 months I've been working with c++.. Kind of annoying! Thanks for any help you can offer!

Thank you!

I had a method down, and was working on it right now, although this one is just stupid in comparison to such a small snippet of code that solves it.


I was going to have it input into strings, and test to see if it holds integer values, and then assemble the values into an integer. It would break down every value using substr methods (I thought the string class could return exact given positions, but I can't find that, so I built a function to make it easier to return exact positions), and then perform checks on the returned string, which would be a single character, adding to an integer variable as it goes along.

Like, if the user sent 543, it would know it's only a 3 digit number to start with. After it performed checks on what character, it would basically add each position like so:

Code:

integer += 5 * 100;
integer += 4 * 10;
integer += 3;

Although, that doesn't yet account for handling 0 input, which doesn't matter anyways, since I don't have to bother with it anymore!

Thanks again!

Okay, that didn't work for me unfortunately, after testing it. I got the same results I always do, endless loop.. Not sure why it breaks cin, but it does. I did manage a new one, not including my other idea. I used cin.peek(). It allows the user to put things into the stream, however, the command itself doesn't take from the stream, it leaves it IN the stream, and returns the value of what is in the stream as a character.

Code:

int getInput()
{
	char input;
	int integer = 0;

	//do while integer is 0
	do {
		//get input from the user
		input = cin.peek();
		
		//if input is a digit, put it into an integer variable
		if (isdigit(input)) { cin>> integer; }

		//otherwise, empty the stream
		else { cin.ignore(80, '\n'); }
	} while (!(integer));

	return integer;
}

I wish I had known cin.peek() took in input, I only thought it was just to see if anything was left in the stream! I'd have had this done long ago!

Thanks again for the help.

This is a common problem. The most common way to deal with it (or at least my favorite) is this:

Code:

int integer = 0;
while (!(cin >> integer))
{
    cin.clear(); // clear the fail bit
    cin.ignore(1000, '\n'); // ignore the characters
    // Give an error message and prompt the user for an integer again here.
}

This is a better general solution IMO because it allows 0 and negative numbers to be input. It also works for any datatype that can make cin fail.