Go语言的OOP,接口,接口的组合,基础库的函数及接口如何抽象设计,
这些东西在Go的Heap源码及演示例子处理中,都有很好的展示.
在"container/heap"中,它的接口是如下定义的:
type Interface interface {
sort.Interface
Push(x interface{}) // add x as element Len()
Pop() interface{} // remove and return element Len() - 1.
} 我不太明白为啥是这样的设计,只好通过下面的方法来尝试了解.1. 通过测试例子,了解使用方式.
2. 试图还原原始场景,即利用源码整合实现了一个原始的堆排序
然后通过这两者的对比,来慢慢体会.
container/heap的测试例子:
package main
import (
"container/heap"
"fmt"
"sort"
)
// An IntHeap is a min-heap of ints.
type IntHeap []int
func (h IntHeap) Len() int { return len(h) }
func (h IntHeap) Less(i, j int) bool { return h[i] < h[j] }
func (h IntHeap) Swap(i, j int) { h[i], h[j] = h[j], h[i] }
func (h *IntHeap) Push(x interface{}) {
// Push and Pop use pointer receivers because they modify the slice's length,
// not just its contents.
*h = append(*h, x.(int))
}
func (h *IntHeap) Pop() interface{} {
old := *h
n := len(old)
x := old[n-1]
*h = old[0 : n-1]
return x
}
// This example inserts several ints into an IntHeap, checks the minimum,
// and removes them in order of priority.
func main() {
h := &IntHeap{100,16,4,8,70,2,36,22,5,12}
fmt.Println("\nHeap:")
heap.Init(h)
fmt.Printf("最小值: %d\n", (*h)[0])
//for(Pop)依次输出最小值,则相当于执行了HeapSort
fmt.Println("\nHeap sort:")
for h.Len() > 0 {
fmt.Printf("%d ", heap.Pop(h))
}
//增加一个新值,然后输出看看
fmt.Println("\nPush(h, 3),然后输出堆看看:")
heap.Push(h, 3)
for h.Len() > 0 {
fmt.Printf("%d ", heap.Pop(h))
}
fmt.Println("\n使用sort.Sort排序:")
h2 := IntHeap{100,16,4,8,70,2,36,22,5,12}
sort.Sort(h2)
for _,v := range h2 {
fmt.Printf("%d ",v)
}
}
/*
输出结果:
-----------------------------------
Heap:
最小值: 2
Heap sort:
2 4 5 8 12 16 22 36 70 100
Push(h, 3),然后输出堆看看:
3
使用sort.Sort排序:
2 4 5 8 12 16 22 36 70 100
*/
自定义的类,实现相关接口后,交由heap.Init()去构建堆.
从堆中Pop()后,数据就被从heap中移除了.
升降序由Less()来决定.
自定义类也可以直接用Sort来排序,因为实现了相关接口.
再看下我手工整合实现的堆排序代码:
package main
//Heap
//author:Xiong Chuan Liang
//date:2015-2-4
import (
"fmt"
)
var(
heap = []int{100,16,4,8,70,2,36,22,5,12}
)
func main(){
fmt.Println("\n数组:")
Print(heap)
MakeHeap()
fmt.Println("\n构建树后:")
Print(heap)
fmt.Println("\n增加 90,30,1 :")
Push(90)
Push(30)
Push(1)
Print(heap)
n := Pop()
fmt.Println("\nPop出最小值(",n,")后:")
Print(heap)
fmt.Println("\nRemove()掉idx为3即值",heap[3-1],"后:")
Remove(3)
Print(heap)
fmt.Println("\nHeapSort()后:")
HeapSort()
Print(heap)
}
func Print(arr []int){
for _,v := range arr {
fmt.Printf("%d ",v)
}
}
//构建堆
func MakeHeap(){
n := len(heap)
for i := n/2-1 ;i >= 0;i--{
down(i, n)
}
}
//由父节点至子节点依次建堆
//parent : i
//left child : 2*i+1
//right child : 2*i+2
func down(i,n int) {
//构建最小堆,父小于两子节点值
for {
j1 := 2*i + 1
if j1 >= n || j1 < 0 { // j1 < 0 after int overflow
break
}
//找出两个节点中最小的(less: a<b)
j := j1 // left child
if j2 := j1 + 1; j2 < n && !Less(j1, j2) {
j = j2 // = 2*i + 2 // right child
}
//然后与父节点i比较,如果父节点小于这个子节点最小值,则break,否则swap
if !Less(j, i) {
break
}
Swap(i, j)
i = j
}
}
//由子节点至父节点依次重建堆
func up(j int) {
for {
i := (j - 1) / 2 // parent
if i == j || !Less(j, i) {
//less(子,父) !Less(9,5) == true
//父节点小于子节点,符合最小堆条件,break
break
}
//子节点比父节点小,互换
Swap(i, j)
j = i
}
}
func Push(x interface{}){
heap = append(heap, x.(int))
up(len(heap)-1)
return
}
func Pop() interface{} {
n := len(heap) - 1
Swap(0, n)
down(0, n)
old :=heap
n = len(old)
x := old[n-1]
heap = old[0 : n-1]
return x
}
func Remove(i int) interface{} {
n := len(heap) - 1
if n != i {
Swap(i, n)
down(i, n)
up(i)
}
return Pop()
}
func Less(a,b int)bool{
return heap[a] < heap[b]
func Swap(a,b int){
heap[a],heap[b] = heap[b],heap[a]
}
func HeapSort(){
//升序 Less(heap[a] > heap[b]) //最大堆
//降序 Less(heap[a] < heap[b]) //最小堆
for i := len(heap)-1 ;i > 0;i--{
//移除顶部元素到数组末尾,然后剩下的重建堆,依次循环
Swap(0, i)
down(0, i)
}
}
/*
输出结果:
-----------------------------------
数组:
100 16 4 8 70 2 36 22 5 12
构建树后:
2 5 4 8 12 100 36 22 16 70
增加 90,30,1 :
1 5 2 8 12 4 36 22 16 70 90 100 30
Pop出最小值( 1 )后:
2 5 4 8 12 30 36 22 16 70 90 100
Remove()掉idx为3即值 4 后:
4 5 8 16 12 30 36 22 100 70 90
HeapSort()后:
100 90 70 36 30 22 16 12 8 5 4
*/
两相对比之后发现,container/heap 竞然,真的,仅且只是封装了一个堆而已.
把那些不确定的,各种需要定制的东西,都交给了用户去实现.
它仅仅只负责最核心的堆这部份的东西.
这样基础库清爽了,使用堆时也不会再感到缚手缚脚了.
MAIL: xcl_168@aliyun.com
BLOG:http://blog.csdn.ent/xcl168
有疑问加站长微信联系（非本文作者）
