js神秘的电报密码 哈弗曼编码实现
muamaker 人气:0这篇文章主要介绍了js神秘的电报密码 哈弗曼编码,文中通过示例代码介绍的非常详细，对大家的学习或者工作具有一定的参考学习价值,需要的朋友可以参考下
哈夫曼编码，根据每个单词在文本中出现的次数频率为权值，频率高的权值大。然后每次取两个频率最小的生成树，最后生成一颗大树。从根节点到该单词的路径，左边为0，右边为1,
function HFM(){
var souce = [];
function createNode(node){
var obj = {
weight:0,
parent:-1,
lchild:-1,
rchild:-1,
value:''
};
return Object.assign(obj,node);
}
this.addNode = function(node){
//添加单词和频率（权值）
souce.push(createNode(node));
}
this.createTree = function(){
//哈夫曼树
var HuffNode = JSON.parse(JSON.stringify(souce));
var n = HuffNode.length;
var x1,x2; //两个权值最小的索引
var m1,m2; //两个权值最小的值
for(var i = 0; i < n ; i++){
m1 = m2 = Infinity; //初始化为最大值
x1 = x2 = -1;
for(var j = 0; j < n+i; j++){ //寻找两个权值最小，且父节点为-1的
var item = HuffNode[j];
if(item.weight < m1 && item.parent == -1){
m2 = m1;
x2 = x1;
m1 = item.weight;
x1 = j;
}else if(item.weight < m2 && item.parent == -1){
m2 = item.weight;;
x2 = j;
}
}
if(x1 != -1 && x2 != -1){
HuffNode[x1].parent = n + i; //更新父节点
HuffNode[x2].parent = n + i;
//创建一个新的节点
HuffNode[n+i] = createNode({
weight:m1+m2,
lchild:x1,
rchild:x2
});
}
}
return HuffNode;
};
this.getCode = function(){
//哈夫曼编码
var n = souce.length;
var tree = this.createTree();
var codes = {};
for(var i = 0; i < n; i++){
var p = tree[i].parent;
var code = '';
var c = i;
while(p != -1){ //迭代前溯
if(tree[p].lchild == c){
code = 0 + code;
}else{
code = 1 + code;
}
c = p;
p = tree[p].parent;
}
codes[ tree[i].value ] = code;
console.log(tree[i].value , code);
}
return codes;
}
}
var hfm = new HFM();
hfm.addNode({
weight:5,
value:"a"
});
hfm.addNode({
weight:32,
value:"b"
});
hfm.addNode({
weight:18,
value:"c"
});
hfm.addNode({
weight:7,
value:"d"
});
hfm.addNode({
weight:25,
value:"e"
});
hfm.addNode({
weight:13,
value:"f"
});
console.log(hfm.getCode())
加载全部内容