51 lines
1.5 KiB
C++
51 lines
1.5 KiB
C++
|
#include <bits/stdc++.h>
|
|||
|
|
|
|||
|
|
using namespace std;
|
|||
|
|
|
|||
|
|
/**
|
|||
|
|
测试用例： 注意这里结点的序号不是连续的
|
|||
|
|
5 5
|
|||
|
|
1 4
|
|||
|
|
4 7
|
|||
|
|
7 18
|
|||
|
|
18 20
|
|||
|
|
20 18
|
|||
|
|
*/
|
|||
|
|
const int N = 1010;
|
|||
|
|
|
|||
|
|
multiset<int> v[N]; //邻接表，这就是存图用的multiset,
|
|||
|
|
int ind[N];
|
|||
|
|
// https://www.cnblogs.com/ucprer/p/12005710.html
|
|||
|
|
// https://www.luogu.com.cn/blog/asuldb/ou-la-hui-lu-yu-ou-la-lu-jing
|
|||
|
|
|
|||
|
|
int n, m;
|
|||
|
|
int path[N], idx;
|
|||
|
|
|
|||
|
|
//递归函数Hierholzer(x)
|
|||
|
|
//插入回路法
|
|||
|
|
void dfs(int x) {
|
|||
|
|
//用一个迭代器依次遍历每一条边，因为multiset内部有序，所以这样先遍历到的边一定较小，就可以保证字典序较小
|
|||
|
|
for (auto it = v[x].begin(); it != v[x].end(); it = v[x].begin()) {
|
|||
|
|
v[x].erase(it);//这里删除一定要删除迭代器，如果直接erase变量的话，会将两点之间的多条路径一起删掉
|
|||
|
|
//v[*it].erase(v[*it].find(x));//如果是无向图，也需要反向删除
|
|||
|
|
dfs(*it);
|
|||
|
|
}
|
|||
|
|
//维护路径
|
|||
|
|
path[++idx] = x;
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
int main() {
|
|||
|
|
//采用邻接表建图
|
|||
|
|
cin >> n >> m;
|
|||
|
|
//m条边
|
|||
|
|
for (int i = 1; i <= m; i++) {
|
|||
|
|
int a, b;
|
|||
|
|
cin >> a >> b;
|
|||
|
|
ind[b]++; //入度
|
|||
|
|
v[a].insert(b);
|
|||
|
|
}
|
|||
|
|
//假设给定的有向图是欧拉图，不需要检查，本代码作用是输出欧拉路径，把检查的代码省略掉了。
|
|||
|
|
dfs(1);//开始递归
|
|||
|
|
for (int i = idx; i; i--) printf("%d\n", path[i]);//倒着输出路径
|
|||
|
|
return 0;
|
|||
|
|
}
|