您应该创建一个实现Comparable 接口的包装器对象,将数组中的值映射到此包装器类的实例,排序并使用这些值。
例子:
static class X implements Comparable<X> {
final int a;
final int b;
final int c;
public X(int a, int b, int c) {
this.a = a;
this.b = b;
this.c = c;
}
public int getA() {return a;}
public int getB() {return b;}
public int getC() {return c;}
@Override
public int compareTo(X that) {
int cmp = Integer.compare(this.a, that.a);
if (cmp == 0) {
cmp = Integer.compare(this.b, that.b);
if (cmp == 0) {
cmp = Integer.compare(this.c, that.c);
}
}
return cmp;
}
}
public static void main(String[] args){
int[] a = {1, 2, 1, 4, 1};
int[] b = {6, 1, 2, 4, 3};
int[] c = {9, 6, 4, 7, 8};
IntStream.range(0, a.length)
.mapToObj(i -> new X(a[i], b[i], c[i]))
.sorted()
.forEach(x -> {
System.out.printf("%d\t%d\t%d\t\n", x.getA(), x.getB(), x.getC());
});
}
此代码打印预期输出:
1 2 4
1 3 8
1 6 9
2 1 6
4 4 7
如果需要,您可以将排序后的值设置回初始数组。
更新
您也可以使用 solution offered by Gerold Boser 的改进版本,尽管它对输入整数的范围和/或数组的计数有一定的限制:
static int m = 0;
public static void main(String[] args) {
// range of integer values for 3 arrays is -2^20 ... 2^20 -1
// or 0..2^21 -1
final int[] a = { Integer.MAX_VALUE>>11, 1, 1, -4, Integer.MIN_VALUE/(1<<11) };
final int[] b = { 6, 1, 1, 4, 3 };
final int[] c = { 9, -6, -24, 7, 8 };
IntSummaryStatistics stats =
Stream.of(Arrays.stream(a), Arrays.stream(b), Arrays.stream(c))
.flatMapToInt(s -> s)
.summaryStatistics();
long min = stats.getMin();
long max = stats.getMax() - min + 1;
IntStream.range( 0, a.length )
// use long to fit multiplication results into 63 bits
.mapToLong( i -> (a[i] - min) * max * max + (b[i]-min) * max + (c[i]-min) )
.sorted()
.forEach( n -> {
// re-calculate long back to ints
a[m] = (int) (n / (max * max) % max + min);
b[m] = (int) (n / max % max + min);
c[m] = (int) (n % max + min);
System.out.printf("% 9d\t% d\t% d\t\n", a[m], b[m], c[m]);
m++;
});
}
输出:
-1048576 3 8
-4 4 7
1 1 -24
1 1 -6
1048575 6 9
注意可以通过使用BigDecimal 而不是long 来克服上述限制,以避免在执行乘法时溢出