Question

public MonomialPolynomial(double... coeffs){
    List<IMonom> monoms = new ArrayList<IMonom>();

    for (int i = 0; i < coeffs.length; i++) {
        // zero coeffs might yaffect the degree
        if(coeffs[i] != 0)
            monoms.add(new Monom(coeffs[i], i));
    }

    super.monoms = monoms;
}

Why do people write double... and not [] double when they mean to array?
Is there a special meaning for this?

Answer 1

double... declares it a “var args” parameter — inside your method it’s identical to double[] but for the caller, it’s much easier to call with varying numbers of arguments (hence var args) without the need to explicitly create an array:

Without var args:

MonomialPolynomial(double[] coeffs) { ...}
...
// explicit array creation necessary
new MonomialPolynomial(new double[] {1, 2, 3)};

With var args:

MonomialPolynomial(double... coeffs) { ...}
...
// array gets created implicitly, so less boilerplate code
new MonomialPolynomial(1, 2, 3);

---

Edit: One thing to watch out for with var args, only the last argument of a method can be a var args argument. This guarantess there’s no ambiguity when calling a method, e.g.

foo(int a, int b...) is unambigious, because the first argument will always be assigned to a and anything after that will be go into of b.

foo(int a..., int... b) on the other hand is ambigious, there’s no way to tell if foo(1, 2, 3) means a={1} and b={2, 3} or a={1, 2} and b={3}.