UVA 12219 Common Subexpression Elimination （dfs瞎搞）

大体题意：

给你一个字符串按照二叉树的形式，用消除公共表达式的方法可以减少表达式树上的的结点，输出最少的结点的图，详细见原题。

思路：

写的比较乱，感觉时间还行吧。借鉴一下吧。

先写dfs 建树，在写个dfs2 从叶子结点向上更新父结点 重新标号，使得相同类的结点归为一类。

在写个dfs3，重新从根节点标号，变得有序。

最后print 函数  要么输出数字 要么输出字符串 讨论一下即可。

#include <cstdio>
#include <cstring>
#include <algorithm>
#include <unordered_map>
#include <string>
#include <map>
#include <iostream>
#define mr make_pair
#define fi first
#define se second
using namespace std;

const int maxn = 50000 + 7;
char s[maxn*20];
struct Node{
    string t;
    int fii,see;
    Node(string t= "", int fii = -1,int see = -1):t(t),fii(fii),see(see){}
    bool operator < (const Node& rhs) const {
        return fii < rhs.fii || (fii==rhs.fii && see < rhs.see) || (fii==rhs.fii && see == rhs.see && t < rhs.t);
    }
    bool operator == (const Node& rhs) const {
        return t == rhs.t && fii == rhs.fii && see == rhs.see;
    }
};

unordered_map<int,bool> mp;
map< Node ,int >Hash;

string node[maxn];
int lch[maxn], rch[maxn],hash_id[maxn],tab[maxn],ord[maxn];
int id, pos,cur,len;
string t;


void dfs(int c){
    if (cur >= len) return;
    t = s[cur];
    while(cur + 1 < len && s[cur+1] != '(' && s[cur+1] != ')' && s[cur+1] != ','){
        ++cur;
        t+= s[cur];
    }
    node[c] = t;
    cur++;
    char ch = s[cur];
    if (ch == ',') return;
    if (ch == ')' || cur >= len ) {
        ++cur;
        return ;
    }
    if (cur < len && ch == '('){
        id++;
        lch[c] = id;
        ++cur;
        dfs(id);
    }
    ch = s[cur];
    if (cur < len && ch == ','){
        id++;
        rch[c] = id;
        ++cur;
        dfs(id);
    }
    ch = s[cur];

    if (cur < len && ch == ')'){
        ++cur;
        return;
    }
}

void print(int c){
    if (c == -1) return ;

    if (mp.count(tab[c])){
        printf("%d",tab[c]);
        return;
    }

    printf("%s",node[c].c_str());
    mp[tab[c]] = 1;

    bool ok = !(lch[c] == -1 && rch[c] == -1);
    bool ok2 = !(lch[c] == -1 || rch[c] == -1);


    if (ok)printf("(");
    print(lch[c]);
    if (ok2)printf(",");
    print(rch[c]);
    if (ok)printf(")");
}
int fii,see;
string tmp;
void dfs2(int cur){
    if (cur == -1) return;
    dfs2(lch[cur]);
    dfs2(rch[cur]);


    if (lch[cur] == -1 && rch[cur] == -1){
        Node u = Node(node[cur],-1,-1);
        if (!Hash.count(u) ){
            hash_id[cur] = id++;
            Hash[u] = id-1;
        }
        else {
            hash_id[cur] = Hash[u];
        }


        return;
    }
    if (lch[cur] == -1) fii = -1;
    else fii = hash_id[lch[cur]];


    if (rch[cur] == -1) see = -1;
    else see = hash_id[rch[cur]];


    tmp = node[cur];
    Node u = Node(tmp,fii,see);


    if (!Hash.count(u) ){
        hash_id[cur] = id++;
        Hash[u] = id-1;
    }
    else {
        hash_id[cur] = Hash[u];
    }

}

void dfs3(int c){
    if (c == -1) return;
    if (ord[hash_id[c] ] == -1){
        tab[c] = id++;
        ord[hash_id[c]] = id-1;
    }
    else {
        tab[c] = ord[hash_id[c]];
    }
//    printf("^^ %d\n",tab[c]);
    dfs3(lch[c]);
    dfs3(rch[c]);
}


int main(){

    int T;
    scanf("%d%*c",&T);

    while(T--){
        memset(lch,-1,sizeof lch);
        memset(rch,-1,sizeof rch);
        memset(ord,-1,sizeof ord);
        len = 0;
        scanf("%s",s);
        len = strlen(s);
        id = pos = cur = 0;
        mp.clear();
        Hash.clear();
        dfs(0);
        id = 1;
        dfs2(0);

        id = 1;
        dfs3(0);

        print(0);
        puts("");
//        printf("&&& %d\n",mp[3]);

    }

    return 0;
}
/**
3
this(is(a,tiny),tree)
a(b(f(a,a),b(f(a,a),f)),f(b(f(a,a),b(f(a,a),f)),f))
z(zz(zzzz(zz,z),zzzz(zz,z)),zzzz(zz(zzzz(zz,z),zzzz(zz,z)),z))


1
a(b(f(a,a),b(f(a,a),f)),f(b(f(a,a),b(f(a,a),f)),f))

**/