Leptos Server Signals
Server signals are leptos signals kept in sync with the server through websockets.
The signals are read-only on the client side, and can be written to by the server. This is useful if you want real-time updates on the UI controlled by the server.
Changes to a signal are sent through a websocket to the client as json patches.
ssr
: ssr is enabled when rendering the app on the server.actix
: integration with the Actix web framework.axum
: integration with the Axum web framework.
Cargo.toml
[dependencies]
leptos_server_signal = "*"
serde = { version = "*", features = ["derive"] }
[features]
ssr = [
"leptos_server_signal/ssr",
"leptos_server_signal/axum", # or actix
]
Client
use leptos::*;
use leptos_server_signal::create_server_signal;
use serde::{Deserialize, Serialize};
#[derive(Clone, Default, Serialize, Deserialize)]
pub struct Count {
pub value: i32,
}
#[component]
pub fn App() -> impl IntoView {
// Provide websocket connection
leptos_server_signal::provide_websocket("ws://localhost:3000/ws").unwrap();
// Create server signal
let count = create_server_signal::<Count>("counter");
view! {
<h1>"Count: " {move || count.get().value.to_string()}</h1>
}
}
Server (Axum)
#[cfg(feature = "ssr")]
pub async fn websocket(ws: WebSocketUpgrade) -> Response {
ws.on_upgrade(handle_socket)
}
#[cfg(feature = "ssr")]
async fn handle_socket(mut socket: WebSocket) {
let mut count = ServerSignal::<Count>::new("counter").unwrap();
loop {
tokio::time::sleep(Duration::from_millis(10)).await;
let result = count.with(&mut socket, |count| count.value += 1).await;
if result.is_err() {
break;
}
}
}
With the example above, the connection does not get reestablished after a connection lost.
To regularly try to reconnect again, the function provide_websocket_with_retry(...)
can
be used:
#[component]
pub fn App() -> impl IntoView {
// Provide websocket connection
leptos_server_signal::provide_websocket_with_retry(
"ws://localhost:3000/ws",
5000, // retry in 5000 milliseconds
).unwrap();
// ... code from above
}