mirror of
https://github.com/ThePhD/sol2.git
synced 2024-03-22 13:10:44 +08:00
57d9a05f88
- 🛠 Make sure the tutorials compile across platforms! - ✍ Redo quite a bit of the documentation
69 lines
1.7 KiB
C++
69 lines
1.7 KiB
C++
#define SOL_ALL_SAFETIES_ON 1
|
|
#include <sol/sol.hpp>
|
|
|
|
#include <string>
|
|
#include <iostream>
|
|
|
|
int main() {
|
|
std::cout << "=== coroutine ===" << std::endl;
|
|
|
|
sol::state lua;
|
|
std::vector<sol::coroutine> tasks;
|
|
|
|
lua.open_libraries(sol::lib::base, sol::lib::coroutine);
|
|
|
|
sol::thread runner_thread = sol::thread::create(lua);
|
|
|
|
lua.set_function("start_task",
|
|
[&runner_thread, &tasks](
|
|
sol::function f, sol::variadic_args va) {
|
|
// You must ALWAYS get the current state
|
|
sol::state_view runner_thread_state
|
|
= runner_thread.state();
|
|
// Put the task in our task list to keep it alive
|
|
// and track it
|
|
std::size_t task_index = tasks.size();
|
|
tasks.emplace_back(runner_thread_state, f);
|
|
sol::coroutine& f_on_runner_thread
|
|
= tasks[task_index];
|
|
// call coroutine with arguments that came
|
|
// from main thread / other thread
|
|
// pusher for `variadic_args` and other sol types
|
|
// will transfer the arguments from the calling
|
|
// thread to the runner thread automatically for
|
|
// you using `lua_xmove` internally
|
|
int wait = f_on_runner_thread(va);
|
|
std::cout << "First return: " << wait
|
|
<< std::endl;
|
|
// When you call it again, you don't need new
|
|
// arguments (they remain the same from the first
|
|
// call)
|
|
f_on_runner_thread();
|
|
std::cout << "Second run complete: " << wait
|
|
<< std::endl;
|
|
});
|
|
|
|
lua.script(
|
|
R"(
|
|
function main(x, y, z)
|
|
-- do something
|
|
coroutine.yield(20)
|
|
-- do something else
|
|
-- do ...
|
|
print(x, y, z)
|
|
end
|
|
|
|
function main2(x, y)
|
|
coroutine.yield(10)
|
|
print(x, y)
|
|
end
|
|
|
|
start_task(main, 10, 12, 8)
|
|
start_task(main2, 1, 2)
|
|
)");
|
|
|
|
std::cout << std::endl;
|
|
|
|
return 0;
|
|
}
|