Skip to content

Examples.md

Latest commit

 

History

History
100 lines (70 loc) · 3.97 KB

Examples.md

File metadata and controls

100 lines (70 loc) · 3.97 KB

Examples

Runnable example apps that exercise the library end-to-end. Useful as starter wiring and as smoke tests when verifying a fresh checkout works against your environment.

Located under src/main/kotlin/com/github/prule/acc/client/example/.

FocusedCarDashboard

Live single-line CLI view of whatever ACC's broadcasting view is currently focused on.

Output (refreshed in place via \r):

[Red Bull Ring       ] focused=#13  Mercedes-AMG GT3 Evo            G:2   120kmh  L12  lap  11.9%

Columns:

Field Source
Track ClientContext.track?.name
Focused car index ClientContext.focusedCarIndex
Car model CarModelRepository.findById(cars[focusedCarIndex].carModelType).name
Gear RealtimeCarUpdate.gear() — translated: 0=R, 1=N, 2+=1st onwards
Speed RealtimeCarUpdate.kmh()
Lap number RealtimeCarUpdate.laps() — completed laps for the focused car
Lap % RealtimeCarUpdate.splinePosition() * 100

Run

Standalone (against the bundled simulator):

# terminal 1
./gradlew runAccSimulator

# terminal 2
./gradlew runFocusedCarDashboard

Against a real ACC server: edit serverIp in FocusedCarDashboard.kt's main() and run the dashboard task on its own.

What it proves

If the dashboard line populates with a sensible track + car name + non-zero speed and lap %, the wiring is healthy:

  • UDP socket reaches the server.
  • Register handshake succeeded (connectionId set inside ClientContext).
  • ContextUpdater requested + received entry list and track data.
  • RealtimeCarUpdate packets are arriving and being parsed.
  • CarModelRepository recognises the car id (or shows ? if the id isn't in the bundled CSV — see CarModels.md).

If the line shows [—] for track, ? for the car model, or stays at 0kmh / 0.0%, something upstream is broken. Drop in a LoggingListener ahead of ContextUpdater to see raw messages and diagnose.

Code walk-through

The class is a single MessageListener<AccBroadcastingInbound>. Two responsibilities:

  1. On REALTIME_CAR_UPDATE matching the focused car, capture per-tick fields (gear, speed, spline) and re-render.
  2. On any preamble or focus-change message, also re-render so the line refreshes promptly when the track/car changes.

State that survives across messages:

@Volatile private var gear: Int = 1
@Volatile private var kmh: Int = 0
@Volatile private var spline: Float = 0f
@Volatile private var laps: Int = 0

Track + car model + focused-car-index come from ClientContext directly — no caching needed in the listener itself.

Wiring (the main() at the bottom of the file):

suspend fun main() {
  val context = ClientContext()
  AccClient(AccClientConfiguration(name = "Dashboard", port = 9000, serverIp = "127.0.0.1"))
    .connect(
      listOf(
        ContextUpdater(context),
        FocusedCarDashboard(context),
      )
    )
}

No LoggingListener (would clutter the dashboard line), no SessionDetector (the dashboard is happy to render outside an active session — useful for showing the focused car in the lobby / pit).

Things to try

  • Add SessionDetector(context, listOf(RecordingSessionListener(Path.of("./recordings")))) to the listener list to record while watching.
  • Replace @Volatile var gear etc. with a RealtimeCarUpdate? reference if you want richer telemetry (delta, position, splits, world coords). See WireProtocol.md for the full body shape.
  • Wrap the FocusedCarDashboard in a FilteredMessageListener<RealtimeCarUpdate> instead of doing type discrimination inside onMessage. See the recipe in ListenerRecipes.md.

See also