Packaged Task | Advanced C++ (Multithreading & Multiprocessing)

The  std::packaged_task class wraps any Callable objects (function, lambda expression, bind expression, or another function object) so that they can be invoked asynchronously. A packaged_task won’t start on its own, you have to invoke it, As its return value is stored in a shared state that can be called/accessed by  std::future objects.

Need of packaged_task

The main advantage of a packaged task is that it can link a callable object to a future and that is very important in a flooding environment. For example, if we have an existing function that fetches the data from Database (DB) and returns it. Now there is a need to execute this function in a separate thread. This can be done using:

std::packaged_task<>

 Otherwise, we’ll have to use:

std::promise<>

and have to change code but with the help of  std::packaged_task<> its simple and we don’t need to do that.

Member Functions

Some of the member functions in packaged_task are:

• Operator=- it moves packaged tasks and it’s a public member function.
• Swap- It just swaps to the packaged task or you can say exchange two packaged tasks with each other.
• get_future- It returns a std::future associated with the promised result.
• reset- This public member function just resets the task.
• (constructor)- As the name suggests this public member function constructs the packaged task.
• (destructor)- Similarly, (destructor) destructs the task object.

Non-Member Functions

One of the non-member functions is:

• swap(packaged_task)- It specializes the std::swap algorithm.

Below is the C++ program to implement the above functions-

C++

// C++ program to implement
// the functions
#include <bits/stdc++.h>

using namespace std;

// Factorial function
int factorial(int N)
{
    int res = 1;
    for (int i = N; i > 1; i--) {
        res *= i;
    }

    cout << "Result is = " << res << "\n";
    return res;
}

// packaged task
deque<packaged_task<int()> > task_q;
mutex mu;
condition_variable cond;

void thread1()
{
    // packaged task
    packaged_task<int()> t;
    {
        unique_lock<mutex> locker(mu);
        cond.wait(locker, []() { return !task_q.empty(); });
        t = move(task_q.front());
        task_q.pop_front();
    }
    t();
}

// Driver Code
int main()
{
    thread t1(thread1);

    // Create a packaged_task<> that
    // encapsulated the callback i.e. a function
    packaged_task<int()> t(bind(factorial, 6));

    // Fetch the associated future<>
    // from packaged_task<>
    future<int> fu = t.get_future();
    {
        lock_guard<mutex> locker(mu);
        task_q.push_back(move(t));
    }
    cond.notify_one();

    // Fetch the result of packaged_task<>
    cout << fu.get();

    // Join the thread. Its blocking and
    // returns when thread is finished.
    t1.join();

    return 0;
}

Output:

Result is = 720
720