[프로그래머스] 뉴스 클러스터링

#include <string>
#include <map>
#include <vector>
using namespace std;

int solution(string str1, string str2)
{
    int answer = 0;

    string clean_str1 = "";
    string clean_str2 = "";

    //둘 다 소문자로 바꾸고
    for (char c : str1)
        clean_str1 += tolower(c);

    for (char c : str2)
        clean_str2 += tolower(c);

    //비교
    map<string, int> m1;
    map<string, int> m2;
    vector<string> v1;
    vector<string> v2;

    for (int i = 0; i < clean_str1.length() - 1; i++)
    {
        string tmp = "";
        
        if(isalpha(clean_str1[i]) && isalpha(clean_str1[i+1])){
            tmp += clean_str1[i];
            tmp += clean_str1[i + 1];
            v1.push_back(tmp);
            m1[tmp]++;
        }
    }
    for (int i = 0; i < clean_str2.length() - 1; i++)
    {
        string tmp = "";
        if(isalpha(clean_str2[i]) && isalpha(clean_str2[i+1])){
            tmp += clean_str2[i];
            tmp += clean_str2[i + 1];
            v2.push_back(tmp);
            m2[tmp]++;
        }
    }


    double cnt_inter = 0; //교집합
    double cnt_union = 0; //합집합
    int total_size = v1.size() + v2.size();
    if(total_size == 0){
        return 65536;
    }
    //합집합, 교집합 찾기
    if (v1.size() < v2.size())
    {
        for (int i = 0; i < v2.size(); i++)
        {
            string cmp = v2[i];
            if(m1[cmp]){
                if(m1[cmp] < m2[cmp]){
                    cnt_inter += m1[cmp];
                    m1[cmp] = 0;
                }else{
                    cnt_inter += m2[cmp];
                    m2[cmp] = 0;
                }
            }
        }
    }else{
        for (int i = 0; i < v1.size(); i++)
        {
            string cmp = v1[i];
            if(m2[cmp]){
                if(m1[cmp] < m2[cmp]){
                    cnt_inter += m1[cmp];
                    m1[cmp] = 0;
                }else{
                    cnt_inter += m2[cmp];
                    m2[cmp] = 0;
                }
            }
        }
    }
    cnt_union = total_size - cnt_inter;

    double jacard = cnt_inter / cnt_union;
    return jacard * 65536;
}