• hihoCoder #1039 字符消除


    目前除了暴力穷举,还没有想出更好的办法。在我的代码里枚举分为两类:

    • 在字符间添加同一两侧之一相同的字符,如向“BC”中添加B或C;
    • 在字符间添加与一两侧均不相同的字符,如向“BC”中添加A。

    之后,将新字符串按连续字符整理为块,再进行操作模拟。

    #include <stdio.h>
    #include <vector>
    #include <string.h>
    #include <string>
    #include <limits>
    #include <algorithm>
    using namespace std;
    
    const int MAX_LEN = 101;
    
    struct Item
    {
        Item(char rc, int rcnt) : c(rc), cnt(rcnt) {}
        char c;
        int cnt;
    };
    
    vector<Item> build_form(char *str)
    {
        size_t len = strlen(str);
        vector<Item> form;
        int i = 0, i1 = 1;
        while (i1 < len)
        {
            while (str[i1] == str[i]) i1++;
            form.push_back(Item(str[i], i1 - i));
            i = i1;
            i1++;
        }
        if (i<len) form.push_back(Item(str[i], i1 - i));
        return form;
    }
    
    void merge_form(vector<Item> &form)
    {
        if (form.empty()) return;
        bool bCont = true;
        while (bCont)
        {
            int i = 0;
            while (i < form.size() - 1 && form[i].c != form[i + 1].c) i++;
            if (i == form.size() - 1)
            {
                bCont = false;
            }
            else
            {
                form[i].cnt += form[i + 1].cnt;
                form.erase(form.begin() + i + 1);
            }
        }
    }
    
    int calc(vector<Item> form)
    {
        int ret = 0;
    
        bool bCont = true;
        while (bCont)
        {
            vector<int> taken;
            for (int i = 0; i < form.size(); i++)
            {
                if (form[i].cnt > 1)
                {
                    ret += form[i].cnt;
                    taken.push_back(i);
                }
            }
            if (!taken.empty())
            {
                int delcnt = 0;
                for (int k = 0; k < taken.size(); k++)
                {
                    int j = taken[k];
                    int offset = j - delcnt;
                    if (offset < form.size() - 1)
                        form.erase(form.begin() + offset);
                    else
                        form.pop_back();
                    delcnt++;
                }
                merge_form(form);
            }
            else
            {
                bCont = false;
            }
        }
    
        return ret;
    }
    
    int try_add_item_at(char *str, int i, char c)
    {
        string s(str);
        s.insert(s.begin() + i, c);
        vector<Item> form = build_form((char *)s.c_str());
        return calc(form);
    }
    
    int try_item_at(vector<Item> form, int i)
    {
        form[i].cnt++;
        return calc(form);
    }
    
    int del(char *str)
    {
        size_t len = strlen(str);
        if (len == 0) return 0;
        if (len == 1) return 2;
        
        vector<Item> form = build_form(str);
    
        //
        int ret = std::numeric_limits<int>::min();
        for (int i = 0; i < form.size(); i++)
        {
            ret = std::max(ret, try_item_at(form, i));
        }
        for (int i = 0; i < len; i++)
        {
            bool val[] = {true, true, true};
            if (i == 0)
            {
                val[str[0] - 'A'] = false;
            }
            else if (i == len - 1)
            {
                val[str[i] - 'A'] = false;
            }
            else
            {
                val[str[i] - 'A'] = false;
                val[str[i + 1] - 'A'] = false;
            }
            for (int j = 0; j < 3; j++)
            {
                if (val[j])
                    ret = std::max(ret, try_add_item_at(str, i, 'A' + j));
            }
        }
        return ret;
    }
    
    int main()
    {
        int t; scanf("%d", &t);
        while (t--)
        {
            char in[MAX_LEN] = {0};
            scanf("%s", in);
    
            int ret = del(in);
            printf("%d
    ", ret);
        }
        return 0;
    }
    View Code
  • 相关阅读:
    4.定时器
    LSTM分类层的理解
    对机器学习中end-to-end learning(端到端学习)的理解
    对非极大值限制(NMS)的理解
    ROI pooling
    softmax的理解
    长尾理论(long tail )的理解
    对双线性插值的理解
    RPN的理解
    卷积神经网络(CNN)和Faster-RCNN的理解
  • 原文地址:https://www.cnblogs.com/tonix/p/4126901.html
Copyright © 2020-2023  润新知