Let us consider this example, use of EventEmitter:
import{EventEmitter}from"events";consteventEmitter=newEventEmitter();eventEmitter.on("text",text=>console.log(text));eventEmitter.once("time",time=>console.log(time));eventEmitter.emit("text","hi!");//Prints "hi!"eventEmitter.emit("time",123);//Prints "123"eventEmitter.emit("time",1234);//Prints nothing ( once )
In EVT the recommended approach is to give every event it's Evt instance. Translation of the example:
import{Evt}from"evt";//Or import { Evt } from "https://evt.land/x/evt/mod.ts" on denoconstevtText=Evt.create<string>();constevtTime=Evt.create<number>();evtText.attach(text=>console.log(text));evtTime.attachOnce(time=>console.log(time));evtText.post("hi!");evtTime.post(123);evtTime.post(1234);
However, the traditional approach that consists of gathering all the events in a single bus is also an option.
Note: Due to a current TypeScript limitation the .attach()methods need to be prefixed with $ when used with fλ ( to in this case) operators but evt.$attach*() are actually just aliases to the corresponding evt.attach*() methods.
Unlike RxJS operators that return Observable EVT operators are function build using native language features, no by composing other pre-existing operators or instantiating any particular class.
Consider that we have an emitter for this data type:
We want to get a Promise<string> that resolves with the next text event.