Struct ProcessAwaiter 

pub struct ProcessAwaiter { /* private fields */ }

Core waiter for process terminal state and events (ADR 0016).

The awaiter is the store-only fallback that ProcessWorkDriver uses when no engine-native ProcessAttach owns the wait. It performs narrow point reads (get_process, events_after) and, when constructed with a ProcessChangeHub, wakes promptly on local mutations instead of polling. Callers still bound every wait with tokio::time::timeout.

Implementations

impl ProcessAwaiter

pub fn new(registry: Arc<dyn ProcessRegistry>, hub: ProcessChangeHub) -> Self

Hub-backed awaiter: local mutations published to hub wake waiters without database polling. This is what a [WatchedProcessRegistry] wrapping provides via watch_process_registry.

pub fn polling(registry: Arc<dyn ProcessRegistry>) -> Self

Hubless awaiter: correct without any change signal, using only the bounded backoff point-read loop (25ms floor, doubling, 1s cap). Use when the registry is not wrapped in-process — e.g. a store-only test.

pub async fn await_terminal( &self, process_id: &str, ) -> Result<ProcessAwaitOutput, PluginError>

Resolve once process_id is terminal, returning its outcome. See ProcessWorkDriver::await_terminal for the timeout-bounding contract.

pub async fn await_event( &self, process_id: &str, event_type: &str, after_sequence: u64, ) -> Result<ProcessEvent, PluginError>

Resolve with the first event_type event on process_id past after_sequence. Historical matches resolve immediately.

Trait Implementations

impl Clone for ProcessAwaiter

fn clone(&self) -> ProcessAwaiter

Returns a duplicate of the value.

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

Performs copy-assignment from source.