Run a function asynchronously in a background process

Spawns an R background process via callr::r_bg() and polls for completion. When the task finishes, a "task" event is fired with:

event$timer_id: the task name
event$value: the return value of fn (or the error message on failure)
event$widget_id: "__async_ok" on success, "__async_error" on failure

Usage

run_async(app, fn, name, args = list(), poll = 0.5)

Arguments

app: An RtuiApp object (must be running).
fn: A function to execute in the background. Must be self-contained - it cannot access the parent session's variables.
name: A unique name for this task (character string).
args: A named list of arguments passed to fn. Default list().
poll: Polling interval in seconds. Default 0.5.

Value

Invisible app.

Details

Handle results with the on_task callback in tui_app().

The background process runs in a separate R session via the callr package. The function fn must be self-contained: it cannot reference objects from the calling environment. Pass any needed data through args.

Examples

if (FALSE) { # \dontrun{
quick_app(
  title = "Async Demo",
  layout = vstack(
    header(),
    static("Click to start a background task", id = "status"),
    button("Run Task", id = "run_btn"),
    footer()
  ),
  on_click = list(
    run_btn = function(event, state) {
      update(state$app, "status", content = "Working...")
      run_async(state$app, function() {
        Sys.sleep(3)
        42
      }, name = "my_task")
      state
    }
  ),
  on_task = function(event, state) {
    if (event$widget_id == "__async_ok") {
      update(state$app, "status",
             content = paste("Result:", event$value))
    } else {
      update(state$app, "status",
             content = paste("Error:", event$value))
    }
    state
  }
)
} # }