|
|
#include <bits/stdc++.h>
|
|
|
|
|
|
using namespace std;
|
|
|
const int N = 5e5 + 10, M = 4e6 + 10;
|
|
|
|
|
|
int e[M], h[N], idx, w[M], ne[M];
|
|
|
void add(int a, int b, int c = 0) {
|
|
|
e[idx] = b, ne[idx] = h[a], w[idx] = c, h[a] = idx++;
|
|
|
}
|
|
|
|
|
|
int dfn[N], low[N], stk[N], top, ts, root;
|
|
|
int n, m;
|
|
|
|
|
|
// 边双模板
|
|
|
int bcnt;
|
|
|
vector<int> bcc[N];
|
|
|
// 这里与有向图的强连通分量有区别，那个是id[N]+sz[N]静态数组实现，记录的是某个点在哪个强连通分量中，
|
|
|
// 而这里，记录的是某个边双连通分量中有哪些节点，是捋着的顺序不同，实现方式不同，无法追求统一，否则就会在某种场景下造成使用双重循环使得代码TLE掉，
|
|
|
// 不要问我是怎么知道的~
|
|
|
|
|
|
void tarjan(int u, int from) {
|
|
|
dfn[u] = low[u] = ++ts;
|
|
|
stk[++top] = u;
|
|
|
// 边双连通分量模板，对比有向图强连通分量，只用栈记录数据，并没有用状态数组in_stk标识是否某节点在栈中
|
|
|
for (int i = h[u]; ~i; i = ne[i]) {
|
|
|
int v = e[i];
|
|
|
if (!dfn[v]) {
|
|
|
tarjan(v, i); // v:点，i:哪条边
|
|
|
low[u] = min(low[u], low[v]);
|
|
|
} else if (i != (from ^ 1))
|
|
|
low[u] = min(low[u], dfn[v]);
|
|
|
}
|
|
|
|
|
|
if (dfn[u] == low[u]) {
|
|
|
++bcnt; // 边双数量+1
|
|
|
int x;
|
|
|
do {
|
|
|
x = stk[top--];
|
|
|
bcc[bcnt].push_back(x); // 记录边双中有哪些点
|
|
|
} while (x != u);
|
|
|
}
|
|
|
}
|
|
|
|
|
|
int main() {
|
|
|
memset(h, -1, sizeof h);
|
|
|
scanf("%d %d", &n, &m);
|
|
|
|
|
|
while (m--) {
|
|
|
int a, b;
|
|
|
scanf("%d %d", &a, &b);
|
|
|
if (a != b) add(a, b), add(b, a);
|
|
|
}
|
|
|
|
|
|
for (root = 1; root <= n; root++)
|
|
|
if (!dfn[root]) tarjan(root, -1);
|
|
|
|
|
|
// 个数
|
|
|
printf("%d\n", bcnt);
|
|
|
|
|
|
for (int i = 1; i <= bcnt; i++) {
|
|
|
printf("%d ", bcc[i].size());
|
|
|
for (auto j : bcc[i]) printf("%d ", j);
|
|
|
printf("\n");
|
|
|
}
|
|
|
return 0;
|
|
|
} |
