从 csv 文件自动确定 2d 矢量大小

Question

我有一个 C++ 代码可以从 csv 文件加载二维向量。在这里，output.csv 文件的数据为 65 x 74496。在此代码中，我将行和列的大小固定为 65 和 74496。但是，将来可以根据数据更改行数和列数。你知道一种从 csv 文件中自动选择行数和列数的方法吗？谢谢！

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

const std::string csvFileName{ "output.csv" };
constexpr size_t NumberOfRows = 65U; // I want to decide this number from csv file automatically
constexpr size_t NumberOfColumns = 74496U; // I want to decide this number from csv file automatically

int main() {
    // writeTestFile();

    // Our data
    std::vector<std::vector<std::string>> data(NumberOfRows, std::vector<std::string>(NumberOfColumns));

    // Open file and check, if it is open
    if (std::ifstream testFileStream{ csvFileName }){//; testFileStream) {

        size_t row{};
        std::string line{};
        line.resize(1'000'000U);

        // Read all lines/rows in a loop
        while (testFileStream >> line) {

            size_t columnIndex{};
            std::vector<std::string> columns(NumberOfColumns);

            // split strings
            std::istringstream iss{ line };
            for (std::string part{}; std::getline(iss, part, ','); columns[columnIndex++] = std::move(part))
                ;
            // Add all column values
            data[row++] = std::move(columns);
        }
    }
    else std::cerr << "\n\nError: Could not open file: " << csvFileName << "\n\n";

    std::cout << "Hello World!\n";
    std::cout << data[1][1] << std::endl;
    return 0;
}

Answer 1

见this answer

这是使用动态数组大小的修改版本。

sam源文件46MB，耗时1399ms。

正如多次写的：请在您的编译器中启用 C++ 17！！！

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

const std::string csvFileName{ "r:\\test.csv" };
constexpr size_t NumberOfRows = 65U;
constexpr size_t NumberOfColumns = 74496U;


void writeTestFile() {
    // Open file and check, if it could be opened
    if (std::ofstream testFileStream{ csvFileName }; testFileStream) {

        // Write all columns
        for (size_t row{}; row < NumberOfRows; ++row) {

            //  Special handling for separator. No comma at the end
            bool writeCommaFirst{ false };

            // For all columns
            for (size_t col{}; col < NumberOfColumns; ++col) {

                // // Create some test string
                std::string temp = "R" + std::to_string(row) + "C" + std::to_string(col);

                // Write data
                testFileStream << (writeCommaFirst ? "," : "") << temp;
                writeCommaFirst = true;
            }
            testFileStream << '\n';
        }
    }
    else std::cerr << "\n\nError. Could not write file" << csvFileName << "\n\n";
}

int main() {
    // writeTestFile();

    // Our data
    std::vector<std::vector<std::string>> data{};

    // Open file and check, if it is open
    if (std::ifstream testFileStream{ csvFileName }; testFileStream) {

        std::string line{};
        line.reserve(1'000'000U);

        // Read all lines/rows in a loop
        while (std::getline(testFileStream, line)) {


            size_t columnIndex{};
            std::vector<std::string> columns{};

            // split strings
            std::istringstream iss{ line };
            for (std::string part{}; std::getline(iss, part, ','); columns.emplace_back(std::move(part)))
                ;
            // Add all column values
            data.emplace_back(std::move(columns));
        }
    }
    else std::cerr << "\n\nError: Could not open file: " << csvFileName << "\n\n";
    return 0;
}