The shared_singleton
trait provides singleton pattern state management with shared container.
Implement the singleton pattern with very short code are available.
You can choose the better shared container from Rc<T>
, Rc<Cell<T>>
, Rc<RefCell<T>>
, Arc<T>
, Arc<Mutex<T>>
and Arc<RwLock<T>>
.
Rc<T>
use shared_singleton::*;
struct Foo(usize);
impl_singleton_rc!(Foo, Foo(1));
let x1 = Foo::singleton();
let x2 = Foo::singleton();
assert_eq!(1, Foo::singleton().0);
assert!(std::rc::Rc::ptr_eq(&x1, &x2));
Rc<Cell<T>>
use shared_singleton::*;
#[derive(Clone, Copy)]
struct Foo(usize);
impl_singleton_rc_cell!(Foo, Foo(1));
let x1 = Foo::singleton();
let x2 = Foo::singleton();
assert_eq!(1, x1.get().0);
x2.set(Foo(2));
assert_eq!(2, x1.get().0);
assert!(std::rc::Rc::ptr_eq(&x1, &x2));
Rc<RefCell<T>>
use shared_singleton::*;
struct Foo(usize);
impl_singleton_rc_refcell!(Foo, Foo(1));
let x1 = Foo::singleton();
let x2 = Foo::singleton();
assert_eq!(1, x1.borrow().0);
x2.borrow_mut().0 = 2;
assert_eq!(2, x1.borrow().0);
assert!(std::rc::Rc::ptr_eq(&x1, &x2));
Arc<T>
use shared_singleton::*;
struct Foo(usize);
impl_singleton_arc!(Foo, Foo(1));
let x1 = Foo::singleton();
let x2 = Foo::singleton();
assert_eq!(1, x1.0);
assert!(std::sync::Arc::ptr_eq(&x1, &x2));
Arc<Mutex<T>>
use shared_singleton::*;
struct Foo(usize);
impl_singleton_arc_mutex!(Foo, Foo(1));
let x1 = Foo::singleton();
let x2 = Foo::singleton();
assert!(std::sync::Arc::ptr_eq(&x1, &x2));
let t1 = std::thread::spawn(move || {
x1.lock().unwrap().0 = 2;
println!("{}", x1.lock().unwrap().0);
});
let t2 = std::thread::spawn(move || {
x2.lock().unwrap().0 = 3;
println!("{}", x2.lock().unwrap().0);
});
t1.join().unwrap();
t2.join().unwrap();
Arc<RwLock<T>>
use shared_singleton::*;
struct Foo(usize);
impl_singleton_arc_rwlock!(Foo, Foo(1));
let x1 = Foo::singleton();
let x2 = Foo::singleton();
assert!(std::sync::Arc::ptr_eq(&x1, &x2));
let t1 = std::thread::spawn(move || {
x1.write().unwrap().0 = 2;
println!("{}", x1.read().unwrap().0);
});
let t2 = std::thread::spawn(move || {
x2.write().unwrap().0 = 3;
println!("{}", x2.read().unwrap().0);
});
t1.join().unwrap();
t2.join().unwrap();
tokio::sync::Mutex
and tokio::sync::RwLock
are available.
tokio::sync::Mutex
use shared_singleton::*;
struct Foo(usize);
impl_singleton_arc_mutex_tokio!(Foo, Foo(1));
#[tokio::main]
async fn main() {
let x1 = Foo::singleton();
let x2 = Foo::singleton();
assert!(std::sync::Arc::ptr_eq(&x1, &x2));
let t1 = tokio::spawn(async move {
x1.lock().await.0 = 2;
println!("{}", x1.lock().await.0);
});
let t2 = tokio::spawn(async move {
x2.lock().await.0 = 3;
println!("{}", x2.lock().await.0);
});
t1.await.unwrap();
t2.await.unwrap();
}
tokio::sync::RwLock
use shared_singleton::*;
struct Foo(usize);
impl_singleton_arc_rwlock_tokio!(Foo, Foo(1));
#[tokio::main]
async fn main() {
let x1 = Foo::singleton();
let x2 = Foo::singleton();
assert!(std::sync::Arc::ptr_eq(&x1, &x2));
let t1 = tokio::spawn(async move {
x1.write().await.0 = 2;
println!("{}", x1.read().await.0);
});
let t2 = tokio::spawn(async move {
x2.write().await.0 = 3;
println!("{}", x2.read().await.0);
});
t1.await.unwrap();
t2.await.unwrap();
}