Mehrdad Afshari的回答可以解决这个问题，但我觉得对于这个简单的任务来说，它有点太啰嗦了。查询表有时可以创造奇迹:

#include <ctime>
#include <iostream>
#include <unistd.h>

std::string gen_random(const int len) {
    static const char alphanum[] =
        "0123456789"
        "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
        "abcdefghijklmnopqrstuvwxyz";
    std::string tmp_s;
    tmp_s.reserve(len);

    for (int i = 0; i < len; ++i) {
        tmp_s += alphanum[rand() % (sizeof(alphanum) - 1)];
    }
    
    return tmp_s;
}

int main(int argc, char *argv[]) {
    srand((unsigned)time(NULL) * getpid());     
    std::cout << gen_random(12) << "\n";        
    return 0;
}

注意rand生成的随机数质量很差。

 void gen_random(char *s, size_t len) {
     for (size_t i = 0; i < len; ++i) {
         int randomChar = rand()%(26+26+10);
         if (randomChar < 26)
             s[i] = 'a' + randomChar;
         else if (randomChar < 26+26)
             s[i] = 'A' + randomChar - 26;
         else
             s[i] = '0' + randomChar - 26 - 26;
     }
     s[len] = 0;
 }

随机字符串，每个运行文件=不同的字符串

        auto randchar = []() -> char
    {
        const char charset[] =
            "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
            "abcdefghijklmnopqrstuvwxyz";

        const size_t max_index = (sizeof(charset) - 1);

        return charset[randomGenerator(0, max_index)];
    };
            std::string custom_string;
            size_t LENGTH_NAME = 6 // length of name
    generate_n(custom_string.begin(), LENGTH_NAME, randchar);

下面是我用c++ 11改编的Ates Goral的答案。我在这里添加了lambda，但原理是你可以传入它，从而控制你的字符串包含什么字符:

std::string random_string( size_t length )
{
    auto randchar = []() -> char
    {
        const char charset[] =
        "0123456789"
        "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
        "abcdefghijklmnopqrstuvwxyz";
        const size_t max_index = (sizeof(charset) - 1);
        return charset[ rand() % max_index ];
    };
    std::string str(length,0);
    std::generate_n( str.begin(), length, randchar );
    return str;
}

下面是一个传递lambda到随机字符串函数的例子:http://ideone.com/Ya8EKf

为什么要用c++ 11呢?

因为您可以为感兴趣的字符集生成遵循特定概率分布(或分布组合)的字符串。 因为它内置了对非确定性随机数的支持 因为它支持unicode，所以你可以把它改成国际化的版本。

例如:

#include <iostream>
#include <vector>
#include <random>
#include <functional> //for std::function
#include <algorithm>  //for std::generate_n

typedef std::vector<char> char_array;

char_array charset()
{
    //Change this to suit
    return char_array( 
    {'0','1','2','3','4',
    '5','6','7','8','9',
    'A','B','C','D','E','F',
    'G','H','I','J','K',
    'L','M','N','O','P',
    'Q','R','S','T','U',
    'V','W','X','Y','Z',
    'a','b','c','d','e','f',
    'g','h','i','j','k',
    'l','m','n','o','p',
    'q','r','s','t','u',
    'v','w','x','y','z'
    });
};    

// given a function that generates a random character,
// return a string of the requested length
std::string random_string( size_t length, std::function<char(void)> rand_char )
{
    std::string str(length,0);
    std::generate_n( str.begin(), length, rand_char );
    return str;
}

int main()
{
    //0) create the character set.
    //   yes, you can use an array here, 
    //   but a function is cleaner and more flexible
    const auto ch_set = charset();

    //1) create a non-deterministic random number generator      
    std::default_random_engine rng(std::random_device{}());

    //2) create a random number "shaper" that will give
    //   us uniformly distributed indices into the character set
    std::uniform_int_distribution<> dist(0, ch_set.size()-1);

    //3) create a function that ties them together, to get:
    //   a non-deterministic uniform distribution from the 
    //   character set of your choice.
    auto randchar = [ ch_set,&dist,&rng ](){return ch_set[ dist(rng) ];};

    //4) set the length of the string you want and profit!        
    auto length = 5;
    std::cout<<random_string(length,randchar)<<std::endl;
    return 0;
}

样例输出。

标准库中最合适的函数是std::sample:

#include <algorithm>
#include <iterator>
#include <random>
#include <string>
#include <iostream>

static const char charset[] = 
    "0123456789"
    "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
    "abcdefghijklmnopqrstuvwxyz";

template<typename URBG>
std::string gen_string(std::size_t length, URBG&& g) {
    std::string result;
    result.resize(length);
    std::sample(std::cbegin(charset), 
                std::cend(charset),
                std::begin(result),
                std::intptr_t(length),
                std::forward<URBG>(g));
    return result;
}

int main() {
    std::mt19937 g;
    std::cout << gen_string(10, g) << std::endl;
    std::cout << gen_string(10, g) << std::endl;
}

随机数生成器的状态应该在调用之间保持在函数之外。

而不是手动循环，更喜欢使用适当的c++算法，在这种情况下std::generate_n，具有适当的随机数生成器:

auto generate_random_alphanumeric_string(std::size_t len) -> std::string {
    static constexpr auto chars =
        "0123456789"
        "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
        "abcdefghijklmnopqrstuvwxyz";
    thread_local auto rng = random_generator<>();
    auto dist = std::uniform_int_distribution{{}, std::strlen(chars) - 1};
    auto result = std::string(len, '\0');
    std::generate_n(begin(result), len, [&]() { return chars[dist(rng)]; });
    return result;
}

这接近于我所说的这个问题的“规范”解决方案。

不幸的是，正确地播种一个通用的c++随机数生成器(例如MT19937)是非常困难的。因此上面的代码使用了一个辅助函数模板random_generator:

template <typename T = std::mt19937>
auto random_generator() -> T {
    auto constexpr seed_bytes = sizeof(typename T::result_type) * T::state_size;
    auto constexpr seed_len = seed_bytes / sizeof(std::seed_seq::result_type);
    auto seed = std::array<std::seed_seq::result_type, seed_len>();
    auto dev = std::random_device();
    std::generate_n(begin(seed), seed_len, std::ref(dev));
    auto seed_seq = std::seed_seq(begin(seed), end(seed));
    return T{seed_seq};
}

这很复杂，而且效率相对较低。幸运的是，它用于初始化thread_local变量，因此每个线程只调用一次。

最后，上述的必要包括:

#include <algorithm>
#include <array>
#include <cstring>
#include <functional>
#include <random>
#include <string>

上面的代码使用类模板参数演绎，因此需要c++ 17。通过添加所需的模板参数，可以对早期版本进行简单的修改。