在 Java 中转换 float int

Question

I'm new to Java and I don't have a clue how to convert given variable to int. I've done this:

public class Ulamek<T> {
    private T licznik;
    private T mianownik;

    public Ulamek(T licznik, T mianownik) {
        this.licznik = licznik;
        this.mianownik = mianownik;
    }

And I want variables licznik and mianownik always to be type of int even if I create an object Ulamek with float or double. Is there a way to convert them or any other way to do this? I have to do this with the <T> in public class Ulamek<T> .

Answer 1

You can make the class non-generic and use method overloading to support instantiating the class with different types.

Example

public class Ulamek {
    private int licznik;
    private int mianownik;

    public Ulamek(int licznik, int mianownik) {
        this.licznik = licznik;
        this.mianownik = mianownik;
    }

    public Ulamek(double licznik, double mianownik) {
        this.licznik = (int)licznik;
        this.mianownik = (int)mianownik;
    }
}

Answer 2

Here is one possible approach.

    public static <T> int value(T v) {
        int k;
        if (v instanceof Double) {
            k = ((Double)v).intValue();
            return k;
        }
        return -1; // error, only valid for positive values of v.
    }

You would need to structure this around exception handling if v is any Number since any value of k you return could not accurately relate an error.

If v is a primitive, then generics won't be involved and just do a simple cast to int

double f  = 4.;
int i = (int)f;

Answer 3

It looks like you're aiming to create a method (in this case it's a constructor, but the principle applies to methods in general) that accepts any type of number and then convert those numbers to an integer value and store those.

If:

• you want to allow any numeric type and want to avoid method overloading,
• you accept that the numbers may get truncated, rounded or even overflow to a negative value (for instance when you provide a long value that is larger than Integer.MAX_VALUE )
• and you want to avoid having to perform the type conversion when invoking your constructor (therefore having int s as an argument may not be an option in your specific case).

You could use an implementation that leverages Java autoboxing and the fact that these boxed numerical types all extend the same base class : Number . Therefore you could write:

public class Ulamek {
    private int licznik;
    private int mianownik;

    public Ulamek(Number licznik, Number mianownik) {
        this.licznik = Objects
                .requireNonNull(licznik, "licznik cannot be null")
                .intValue();
        this.mianownik = Objects
                .requireNonNull(mianownik, "mianownik cannot be null")
                .intValue();
    }
}

No need for generic type parameters.

Note that I added an explicit requireNonNull there. You didn't specify how to deal with null values, also your target data type ( int ) doesn't have a representation for null . Therefore the safe thing to assume is that null is not acceptable here.

Now you could write:

var a = new Ulamek(42, 43); // int
var b = new Ulamek(42.0, 43.0); // double
var c = new Ulamek(42l, 43l); // long
var d = new Ulamek(new BigDecimal(42.0), new BigDecimal(43.0));

Using this approach, adding range checks before assigning the integer values could prove a bit more challenging should you require those.