Open sl1673495 opened 4 years ago
data.json
{
"code": 1,
"children": [
{
"code": 2,
"children": [
{
"code": 5,
"children": [
{
"code": 7
},
{
"code": 8
}
]
}
]
},
{
"code": 3,
"children": [
{
"code": 12,
"children": [
{
"code": 13
},
{
"code": 15
}
]
},
{
"code": 90,
"children": [
{
"code": 99
},
{
"code": 91
},
{
"code": 78
},
{
"code": 66
},
{
"code": 16,
"children": [
{
"code": 74,
"children": [
{
"code": 73,
"children": [
{
"code": 70,
"children": [
{
"code": 61
},
{
"code": 62
},
{
"code": 63,
"children": [
{
"code": 888
},
{
"code": 999,
"children": [
{
"code": 777
},
{
"code": 222
},
{
"code": 111
}
]
},
{
"code": 555
}
]
}
]
}
]
}
]
}
]
}
]
}
]
},
{
"code": 4,
"children": [
{
"code": 21,
"children": [
{
"code": 26
},
{
"code": 95
},
{
"code": 75
}
]
},
{
"code": 36,
"children": [
{
"code": 38,
"children": [
{
"code": 39
}
]
}
]
}
]
}
]
}
index.js
const data = require('./data.json')
function isChildren(node){
return node.children?true:false
}
var pathCode = function(node, code,key) {
const pathes = [];
_pathCode(node, code, pathes, [],key);
return pathes;
};
function _pathCode(node, code, pathes, path,key) {
path = [...path, node[key]];
if (node[key]===code){
pathes.push(...path)
return
}
if (isChildren(node)){
for (let i=0;i<node.children.length;i++){
_pathCode(node.children[i],code,pathes,path,key)
}
}
}
const paths = pathCode(data,555,'code')
大佬的思路还是很明确的,一看就懂,但是 sumVals 这个函数对于节点较多时候还是比较耗时的,可以优化一下:
var pathSum = function (root, targetSum) {
const res = [];
helper(root, [], 0, targetSum, res);
return res;
};
var helper = function (node, paths, sum, targetSum, res) {
if (!node) return;
paths.push(node.val);
sum += node.val;
if (node.left) helper(node.left, paths, sum, targetSum, res);
if (node.right) helper(node.right, paths, sum, targetSum, res);
if (!node.left && !node.right) {
if (sum === targetSum) {
res.push(paths.slice());
}
}
// 退出当前递归,一定要将这次递归的节点去掉,因为要回到父节点重新开启另一条路径。
paths.pop();
sum -= node.val;
}
给定一个二叉树和一个目标和,找到所有从根节点到叶子节点路径总和等于给定目标和的路径。
说明: 叶子节点是指没有子节点的节点。
示例:
来源:力扣(LeetCode) 链接:https://leetcode-cn.com/problems/path-sum-ii 著作权归领扣网络所有。商业转载请联系官方授权,非商业转载请注明出处。
思路
典型的可以用 DFS 来解决的问题,定义一个 search 方法并且参数里带一个用来收集路径的 paths 数组,每当到达叶子节点(没有 left 也没有 right),就计算一把路径的总和,如果等于目标值就 push 到结果数组里。(注意这里要浅拷贝一下,防止下面的计算污染这个数组)
任何一个节点处理完成时,都要把当前节点 pop 出 paths 数组。