Struct QueueExecutor 

pub struct QueueExecutor { /* private fields */ }

The background job executor.

Polls the database for pending jobs and processes them using the registered JobHandler implementation. Supports pause/resume, consecutive job limits with cooldown, graceful shutdown, and cancellation.

Implementations

impl QueueExecutor

pub fn new(config: QueueConfig, db: Arc<Mutex<Connection>>) -> QueueExecutor

pub fn spawn<H>( self: Arc<QueueExecutor>, event_emitter: Arc<dyn QueueEventEmitter>, )

where H: JobHandler + 'static,

Spawn the executor loop as a background tokio task.

The executor will poll for pending jobs at the configured interval and process them using the provided JobHandler implementation.

If the current thread is inside a tokio runtime, the loop is spawned directly via tokio::spawn. Otherwise (e.g., during Tauri’s synchronous setup() phase), a dedicated background thread with its own single-threaded tokio runtime is created automatically.

pub fn spawn_on<H>( self: Arc<QueueExecutor>, event_emitter: Arc<dyn QueueEventEmitter>, handle: &Handle, )

where H: JobHandler + 'static,

Spawn the executor loop on a specific tokio runtime handle.

Use this instead of spawn() when you have an explicit runtime handle (e.g., from tauri::async_runtime::handle()).

pub fn pause(&self)

Pause the executor. The current job (if any) will finish, but no new jobs will be started until resume() is called.

pub fn resume(&self)

Resume the executor after a pause.

pub fn is_paused(&self) -> bool

Check if the executor is currently paused.

pub fn shutdown(&self)

Signal the executor to shut down gracefully.

The currently running job (if any) will finish, then the loop exits.

pub fn is_shutdown(&self) -> bool

Check if a shutdown has been requested.

pub async fn process_one<H>( &self, event_emitter: &Arc<dyn QueueEventEmitter>, ) -> Result<Option<ProcessedJob>, QueueError>

where H: JobHandler,

Process the next pending job and return the result.

Unlike spawn(), this method processes exactly one job in the foreground and returns. Returns Ok(None) if no pending jobs are available.

This is useful for CLI tools that want to drive the execution loop manually (e.g., to run cascade logic between jobs).