Source: https://github.com/janmarques/discussion
Package: https://www.npmjs.com/package/discussion
Discussion: combining SpeechRecognition and SpeechSynthesis
When working with SpeechRecognition and SpeechSynthesis, it is important that you don't synthesise while you are recognizing.
This package helps keep this in check.
It will continually be SpeechRecognizing and passing those results to the onRecognitionResult callback.
But whenever you want to Synthesise something, you can do this using the synthesize method. The Recognition gets disabled and after the synthesis is done, it gets restarted.
How to use
Install with npm i discussion
We need to have a bit of hacky stuff to ensure that in the recognition callback, we can access all the properties and the DOM from your own class.
That is why we pass a factory method to create the SpeechRecognition rather than have the package do it itself.
Example:
// in the header somewhere
public CT = window.SpeechRecognition || webkitSpeechRecognition;
// in the constructor
this.discussion = new Discussion(() => new this.CT(), this.language.code, this.language.code);
this.discussion.onRecognitionResult = (x) => {
// do something with result
};
// in the initializer (eg ngOnInit or whenever your page is loaded)
this.discussion.initialize();
// after user consent, enable the Recognition
this.discussion.enableRecognition();
// anywhere in your code
this.discussion.synthesize("hello world");