我有一个递归方法,用一定的算法计算x ^ n,但这在这里并不重要。重要的是我的辅助函数,它跟踪该算法的递归调用。
public class FastPot {
public static double fastPotRek(double x, int n) {
class Aux {
private double aux(double x, int n, int c) {
if (n == 0) {System.out.print("It took "+c+" recursive calls to compute "); return 1;}
else return (n % 2 == 0) ? aux(x*x, n/2, c+1)
: x * aux(x, n-1, c+1);
}
}
Aux a = new Aux();
return a.aux(x, n, 0);
}
}
为了组织这一点,我想在aux
内声明fastPotRek
,为此你必须使用一个内部类,我的方法不能声明为静态。因此,我实例化了Aux a = new Aux();
以便能够调用aux
。
请告诉我有一种方法可以让它更优雅,并告诉我我忽略了什么......就像蜂鸟能够以某种方式使aux
静态或不需要实例化Aux
。
答案 0 :(得分:3)
不需要内部类,也不需要使用静态类:
public class FastPot {
//Static, use only from within FastPot
private static double aux(double x, int n, int c) {
if (n == 0) {
System.out.print("It took "+c+" recursive calls to compute ");
return 1;
} else {
return (n % 2 == 0) ? aux(x*x, n/2, c+1)
: x * aux(x, n-1, c+1);
}
}
}
//Your outward interface
public static double fastPotRek(double x, int n) {
return aux(x, n, 0);
}
}
或者如果你坚持使用内部类:
public class FastPot {
//Static, use only from within FastPot
private static class Aux {
private static double aux(double x, int n, int c) {
if (n == 0) {
System.out.print("It took "+c+" recursive calls to compute ");
return 1;
} else {
return (n % 2 == 0) ? aux(x*x, n/2, c+1)
: x * aux(x, n-1, c+1);
}
}
}
//Your outward interface
public static double fastPotRek(double x, int n) {
return Aux.aux(x, n, 0);
}
}
答案 1 :(得分:2)
我会发布这个答案,虽然很多人(包括我)对此不满意。请不要使用此类代码:
public static double fastPotRek(double x, int n) {
return new Cloneable() {
private double aux(double x, int n, int c) {
if (n == 0) {System.out.print("It took "+c+" recursive calls to compute "); return 1;}
else return (n % 2 == 0) ? aux(x*x, n/2, c+1) : x * aux(x, n-1, c+1);
}
}.aux(x, n, 0);
}
同样,我强烈建议使用私有静态方法,如:
public static double fastPotRek(double x, int n) {
return aux(x,n,0);
}
private static double aux(double x, int n, int c) {
...
}
答案 2 :(得分:0)
你有这个:
Aux a = new Aux();
return a.aux(x, n, 0);
我会这样写(也一样):
return new Aux().aux(x, n, 0);
编辑:我已完成StackOverflow。你们不想要帮忙吗?我不会放弃它。从现在开始,我只会询问我的问题,并且不会回答。