Struct SupervisorRuntime 

pub struct SupervisorRuntime { /* private fields */ }

An async runtime for supervising a group of child tasks.

SupervisorRuntime manages a JoinSet of async tasks and applies a RestartStrategy to determine which children to restart when a failure occurs. It handles the concurrency details of spawning, joining, and tracking child tasks.

Examples

use gust_runtime::prelude::*;

let runtime = SupervisorRuntime::with_strategy(RestartStrategy::OneForAll);
let handle = runtime.spawn_named("worker-1", async {
    // ... child machine logic ...
    Ok(())
});

// Wait for the next child to complete
if let Some(result) = runtime.join_next().await {
    match result {
        Ok(()) => println!("child completed successfully"),
        Err(e) => println!("child failed: {e}"),
    }
}

Implementations

impl SupervisorRuntime

pub fn new() -> Self

Creates a new SupervisorRuntime with the default RestartStrategy::OneForOne strategy.

pub fn with_strategy(strategy: RestartStrategy) -> Self

Creates a new SupervisorRuntime with the specified restart strategy.

pub fn spawn_named<F>(&self, id: impl Into<String>, fut: F) -> ChildHandle

where F: Future<Output = Result<(), String>> + Send + 'static,

Spawns a named child task and returns a ChildHandle.

The future fut is spawned onto the Tokio runtime. If the internal task set lock is immediately available the spawn happens synchronously; otherwise a helper task is used to avoid blocking.

The id is stored in the returned ChildHandle for identification purposes.

pub async fn join_next(&self) -> Option<Result<(), String>>

Waits for the next child task to complete and returns its result.

Returns Some(Ok(())) when a child succeeds, Some(Err(..)) when a child fails, or None when all children have completed and no pending spawns remain.

This method will yield cooperatively if there are pending spawns that have not yet been added to the task set.

pub fn strategy(&self) -> RestartStrategy

Returns the RestartStrategy configured for this runtime.

pub fn restart_scope( &self, failed_child_index: usize, child_count: usize, ) -> Range<usize>

Computes the range of child indices that should be restarted when the child at failed_child_index fails.

The returned range depends on the configured RestartStrategy:

• OneForOne: only the failed child (index..index+1)
• OneForAll: all children (0..child_count)
• RestForOne: the failed child and all that follow (index..child_count)

Trait Implementations

impl Clone for SupervisorRuntime

fn clone(&self) -> SupervisorRuntime

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for SupervisorRuntime

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for SupervisorRuntime

fn default() -> Self

Returns the “default value” for a type.