Question

What is the correct way to make a constructor’s argument accessible to different methods within a class?

For example, in the code snippet below, I want to make N accessible within a method called aMethod, without changing aMethod’s existing argument signature. Is myArray.length the best alternative?

public class MyClass{

  private int[][] myArray;

  public MyClass(int N){

    if(N <= 0) 
      throw new IndexOutOfBoundsException("Input Error: N <= 0");  

    myArray = new int[N][N];            
  }

  public void aMethod(int i, int j){

    // N won't work here. Is myArray.length the best alternative?       
    if(i <= 1 || i > N) 
      throw new IndexOutOfBoundsException("Row index i out of bounds");
    if(j <= 1 || j > N) 
      throw new IndexOutOfBoundsException("Column index j out of bounds");            
  }
}

EDIT 1
I’m testing for inputs greater than 0 so if a user enters 0 for i or 0 for j, the input is invalid.

Answer 1

Just create a field for it, like you did for the array.

 public class MyClass{

    private int[][] myArray;
    private int myArraySize;

    public MyClass(int N){

      if(N <= 0) 
        throw new IndexOutOfBoundsException("Input Error: N <= 0");  

      myArray = new int[N][N];
      myArraySize = N;            
    }

    ...
 }

However in this case I wouldn’t do that, I’d change aMethod() instead:

public void aMethod(int i, int j){

    // N won't work here. Is myArray.length the best alternative       
    if(i < 0 || i >= myArray.length ) 
      throw new IndexOutOfBoundsException("Index i out of bounds");
    if(j < 0 || j >= myArray[i].length) 
      throw new IndexOutOfBoundsException("Column index j out of bounds");            
}

(I also changed the check to allow [0..N-1] instead of [1..N], as arrays are indexed from 0.)