Enum DispatchAction 

#[non_exhaustive]
pub enum DispatchAction {
    Handled,
    Reply(Vec<u8>),
    ForwardToAiDaemon(RequestId),
}

Outcome returned to the connection loop after dispatching a frame.

Marked #[non_exhaustive] because new variants will be added as the AI daemon layer is built out. Downstream code must include a wildcard arm when matching on this type.

Examples

use nerve_ipc_core::DispatchAction;

fn handle(action: DispatchAction) -> Option<Vec<u8>> {
    match action {
        DispatchAction::Handled => None,
        DispatchAction::Reply(bytes) => Some(bytes),
        DispatchAction::ForwardToAiDaemon(_) => None,
        // Required: new variants will be added as the AI daemon is built out.
        _ => None,
    }
}

Variants (Non-exhaustive)

Non-exhaustive enums could have additional variants added in future. Therefore, when matching against variants of non-exhaustive enums, an extra wildcard arm must be added to account for any future variants.

Handled

Frame was handled; no response needs to be sent to the client.

Reply(Vec<u8>)

Frame generated a response that the transport must write back to the client. The bytes are a complete, encoded NERVE frame.

ForwardToAiDaemon(RequestId)

A MessageType::SearchQuery was received. The request has been registered in the RequestTable so that a subsequent MessageType::Cancel can mark it cancelled. The AI daemon (a future milestone) is responsible for consuming and responding to this request. The enclosed RequestId is the ID already stored in the request table.