#include <iostream> #include <string> #include <map> using namespace std; enum State { START, HAS_A, ONLY_B, END }; bool DFA(string s) { int state = START; map<State, map<char, State>> transition = { {START, {{'a', HAS_A}}}, {HAS_A, {{'a', END}, {'b', ONLY_B}}}, {ONLY_B, {{'a', HAS_A}, {'b', ONLY_B}}}, {END, {}} }; for (int i = 0; i < s.size(); i++) { if (transition[state].find(s[i]) == transition[state].end()) return false; state = transition[state][s[i]]; } return state == END;} int main() { string s; cin >> s; if (DFA(s)) { cout << "Accept" << endl; } else { cout << "Error" << endl; } return 0; }优化这段代码

这段代码的 DFA 实现比较简洁，已经使用了状态转移表，但是可以在细节上进行一些优化，比如使用 const、引用等：

#include <iostream>
#include <string>
#include <unordered_map>

enum class State {
    START,
    HAS_A,
    ONLY_B,
    END
};

bool DFA(const std::string& s) {
    State state = State::START;
    const std::unordered_map<State, std::unordered_map<char, State>> transition = {
        {State::START, {{'a', State::HAS_A}}},
        {State::HAS_A, {{'a', State::END}, {'b', State::ONLY_B}}},
        {State::ONLY_B, {{'a', State::HAS_A}, {'b', State::ONLY_B}}},
        {State::END, {}}
    };
    for (const char& c : s) {
        if (transition.find(state) == transition.end() || transition.at(state).find(c) == transition.at(state).end()) {
            return false;
        }
        state = transition.at(state).at(c);
    }
    return state == State::END;
}

int main() {
    std::string s;
    std::cin >> s;
    if (DFA(s)) {
        std::cout << "Accept" << std::endl;
    } else {
        std::cout << "Error" << std::endl;
    }
    return 0;
}

主要的改进有：

1. 使用 `enum class` 代替 `enum`，避免命名空间污染
2. 使用 `std::unordered_map` 代替 `std::map`，提高查找效率
3. 使用 const、引用等来提高代码的安全性和效率