字符串识别-优快云博客

本文链接：https://blog.youkuaiyun.com/ray_kong/article/details/145518105

使用C++在Linux平台下实现的简易的字符串识别，类型包括常见进制、base64编码、URL等格式。

问题

输入：不定长字符串

输出：数据类型

例：

输入：0x1A2B

输出：十六进制

思路

使用正则表达式进行格式判断；

使用libcurl发送HTTP（默认GET）请求，验证URL是否可达

代码

头文件

#include <iostream>
#include <regex>
#include <string>
#include <curl/curl.h>

方法

//Determine if the string is binary
bool isBinaryString(const std::string& str) {
    std::regex BinaryPattern("^[01]+$");
    return std::regex_match(str, BinaryPattern);
}

//Determine if the string is octal
bool isOctalString(const std::string& str){
    std::regex OctalPattern("^0[0-7]+$");
    return std::regex_match(str,OctalPattern);
}

//Determine if the string is hex
bool isHexString(const std::string& str){
    std::regex HexPattern("^(0[xX])?[0-9a-fA-F]+$");
    return std::regex_match(str,HexPattern);
}

//Determine if the string is base64
bool isBase64String(const std::string& str){
    std::regex Base64Pattern("^(?:[A-Za-z0-9+/]{4})*(?:[A-Za-z0-9+/]{3}=|[A-Za-z0-9+/]{2}==)?$");
    return std::regex_match(str,Base64Pattern);
}

//Determine if the string is a valid URL format (using regex)
bool isURLString(const std::string& str){
    std::regex UrlPattern(R"(^(https?|ftp|file)://[-A-Za-z0-9+&@#/%?=~_|!:,.;]*[-A-Za-z0-9+&@#/%=~_|])");
    return std::regex_match(str,UrlPattern);
}

// check if the URL is reachable (using libcurl)
// send HTTP request using libcurl to check if the URL is reachable
bool isURLReachable(const std::string& url){
    CURL* curl = curl_easy_init();
    if(!curl){
        std::cerr<<"Failed to initialize libcurl."<<std::endl;
        return false;
    }

    // set the URL to check
    curl_easy_setopt(curl, CURLOPT_URL, url.c_str());

    // set timeout for the request
    curl_easy_setopt(curl, CURLOPT_TIMEOUT, 5L); // 5 seconds timeout

    // perform the request
    CURLcode res = curl_easy_perform(curl);

    // clean up
    curl_easy_cleanup(curl);

    // check if the request was successful
    return (res == CURLE_OK); 
}

主函数

int main(int argc, char* argv[]) {
    if(argc < 2){
        std::cout<<"enter your parameter!"<<std::endl;
        return 1;
    }

    std::string input = argv[1];

    if(isBinaryString(input)){
        std::cout<<"the string is binary."<<std::endl;
    }else if(isOctalString(input)){
        std::cout<<"the string is octal."<<std::endl;
    }else if(isHexString(input)){
        std::cout<<"the string is hexadecimal."<<std::endl;
    }else if(isBase64String(input)){
        std::cout<<"the string is base64"<<std::endl;
    }else if(isURLString(input)){
        std::cout<<"the string is a valid URL format."<<std::endl;

        // check if the URL is reachable
        if(isURLReachable(input)){
            std::cout<<"the URL is reachable."<<std::endl;
        }else{
            std::cout<<"the URL is unreachable."<<std::endl;
        }
    }else{std::cout<<"the string format is unknown."<<std::endl;
    }
    return 0;
}