c++ 可以智能判断csv域包裹符 源码

以下是一个例子，展示了如何使用状态机来智能判断CSV域包裹符的C++代码：

#include <iostream>
#include <fstream>
#include <string>
#include <vector>

using namespace std;

enum CSVState {
    UnquotedField,
    QuotedField,
    QuotedQuote
};

vector<string> readCSVRow(const string& row) {
    CSVState state = UnquotedField;
    vector<string> fields {""};
    size_t i = 0;  // index of the current field
    for (char c : row) {
        switch (state) {
            case UnquotedField:
                switch (c) {
                    case ',': // end of field
                        fields.push_back("");
                        i++;
                        break;
                    case '"': state = QuotedField;
                        break;
                    default:  // other character
                        fields[i].push_back(c);
                        break;
                }
                break;
            case QuotedField:
                switch (c) {
                    case '"': state = QuotedQuote;
                        break;
                    default:  // other character
                        fields[i].push_back(c);
                        break;
                }
                break;
            case QuotedQuote:
                switch (c) {
                    case ',': // , after closing quote
                        fields.push_back("");
                        i++;
                        state = UnquotedField;
                        break;
                    case '"': // "" -> "
                        fields[i].push_back('"');
                        state = QuotedField;
                        break;
                    default:  // end of quote
                        state = UnquotedField;
                        break;
                }
                break;
        }
    }
    return fields;
}

vector<vector<string>> readCSV(const string& filename) {
    vector<vector<string>> rows;
    ifstream file(filename);
    string row;
    while (getline(file, row)) {
        rows.push_back(readCSVRow(row));
    }
    return rows;
}

int main() {
    auto rows = readCSV("example.csv");
    for (const auto& row : rows) {
        for (const auto& field : row) {
            cout << field << " ";
        }
        cout << endl;
    }
    return 0;
}

以上代码会将CSV文件中的每一行解析为一个字符串向量，并且自动识别出域包裹符。如果当前字段需要被包裹，则将状态机的状态切换到`QuotedField`，直到遇到另一个域包裹符或行尾。如果当前状态为`QuotedField`，则会将所有字符都作为字段的一部分，直到遇到另一个域包裹符。如果当前状态为`QuotedQuote`，则表示遇到了两个连续的域包裹符，因此会将其视为一个包裹符。