Example application to get you started can be found in src/safe_queue.cpp.
#include "safe_queue.hpp"Requirements:
Cmake > 3.5
Catch2 > 3.0
Run:
mkdir build
cd build
cmake ..
make
./safe_queue
./tests// some things you'll need
#include <iostream>
// new queue
SafeQueue<int> queue;
// put three in
queue.put(1);
queue.put(2);
queue.put(3);
// get three out
for(int i = 0; i < 3; ++i){
int item = queue.get();
std::cout << "Item: " << item << "\n";
}
// some things you'll need
#include <iostream>
#include <thread>
#include <chrono>
// new queue
std::shared_ptr<SafeQueue<int>> queue = std::make_shared<SafeQueue<int>>();
// listen to the queue in a thread
std::thread test_thread = std::thread([queue](){
while(true){
// get an item from the queue
int item = queue->get();
std::cout << "Item: " << item << "\n";
}
});
test_thread.detach();
// put three in, waiting a bit in between
for(int i = 0; i < 3; ++i){
queue->put(i);
std::this_thread::sleep_for(std::chrono::milliseconds(500));
}
// some things you'll need
#include <iostream>
#include <thread>
#include <chrono>
// new queue
std::shared_ptr<SafeQueue<int>> queue = std::make_shared<SafeQueue<int>>();
// listen to the queue in a thread
std::thread test_thread = std::thread([queue](){
while(true){
// get an item from the queue
int item = queue->get();
std::cout<<"Item number: "<<item<<" queue size: "<<queue->size<<"\n";
// wait 1ms bit for demonstration purposes
std::this_thread::sleep_for(std::chrono::milliseconds(1));
}
});
test_thread.detach();
// If the queue is processed too slow and we want to stop
// adding items if there are too many items already
for(int i = 0; i < 20; ++i){
if(queue->size < 3){
queue->put(i);
}
}
// wait for this all to be over
std::this_thread::sleep_for(std::chrono::milliseconds(500));
If you're working with a lot of data, it's faster to create a unique pointer to the data you want in the queue, move it into the queue, and move it out of the queue again when you need it to gain back ownership again.
Run the tests to see a benchmark showing the performance increase.
for(int i = 0; i < 100; ++i){
// make some big data
auto big = std::make_unique<std::vector<int>>(1e6);
// notice the move here
big_queue->put_move(std::move(big));
}
std::cout << "Added 100 to big_queue\n";
for(int i = 0; i < 100; ++i){
// notice the move here
auto big = big_queue->get_move();
}
std::cout << "Got 100 to big_queue\n";
If you want to do something else with the object after you put it into the queue, you'll need the reference counting of the std::shared_ptr to keep track of the memory lifetime of the object in the queue. Like in this example:
// new queue
auto queue = std::make_shared<SafeQueue<std::shared_ptr<std::vector<int>>>>();
// put 100 in
for(int i = 0; i < 100; ++i){
auto foo = std::make_shared<std::vector<int>>(1e6);
queue->put(foo);
}
// get 100 out
for(int i = 0; i < 100; ++i){
std::shared_ptr<std::vector<int>> foo = queue->get();
}
-------------------------------------------------------------------------------
PutGet
-------------------------------------------------------------------------------
benchmark name samples iterations est run time
mean low mean high mean
std dev low std dev high std dev
-------------------------------------------------------------------------------
PutGet 1 item 100 18 2.133 ms
1.1953 us 1.1843 us 1.23831 us
96.1105 ns 17.3557 ns 221.307 ns
PutGet 3 items 100 9 2.3292 ms
2.69148 us 2.64209 us 2.77551 us
321.14 ns 214.406 ns 463.596 ns
PutGet 1000 items 100 1 67.5246 ms
676.699 us 674.708 us 679.362 us
11.6007 us 9.23667 us 15.628 us
PutGet 1000 items, with 1000
elements each 100 1 3.52502 s
35.2378 ms 35.2049 ms 35.3009 ms
224.446 us 135.443 us 369.356 us
-------------------------------------------------------------------------------
PutGetMove
-------------------------------------------------------------------------------
benchmark name samples iterations est run time
mean low mean high mean
std dev low std dev high std dev
-------------------------------------------------------------------------------
PutGetMove 1 item 100 27 2.1546 ms
787.362 ns 782.206 ns 807.316 ns
46.9645 ns 7.65404 ns 110.72 ns
PutGetMove 3 items 100 15 2.1765 ms
1.45697 us 1.44714 us 1.49258 us
85.606 ns 22.4203 ns 197.675 ns
PutGetMove 1000 items 100 1 30.3233 ms
304.776 us 303.399 us 306.706 us
8.31293 us 6.35508 us 11.3026 us
PutGetMove 1000 items, with 1000
elements each 100 1 1.05036 s
10.4601 ms 10.4415 ms 10.4847 ms
108.646 us 82.545 us 154.92 us