Skip to content

MikeyParton/react-speech-kit

Repository files navigation

react-speech-kit 🎤    CircleCI codecov

React hooks for in-browser Speech Recognition and Speech Synthesis. Demo here

Table of Contents

Install

yarn add react-speech-kit

Examples and Demo

A full example can be found here. The source code is in the examples directory.

useSpeechSynthesis

A react hook for the browser's SpeechSynthesis API. It exposes the options and controls to the underlying SpeechSynthesis in the browser.

Usage

import React, { useState } from 'react';
import { useSpeechSynthesis } from 'react-speech-kit';

function Example() {
  const [value, setValue] = useState('');
  const { speak } = useSpeechSynthesis();

  return (
    <div>
      <textarea
        value={value}
        onChange={(event) => setValue(event.target.value)}
      />
      <button onClick={() => speak({ text: value })}>Speak</button>
    </div>
  );
}

Args

onEnd

function() optional

Called when SpeechSynthesis finishes reading the text or is cancelled. It is called with no argumnents. Very useful for triggering a state change after something has been read.

Returns

useSpeechSynthesis returns an object which contains the following:

speak

function({ text: string, voice: SpeechSynthesisVoice })

Call to make the browser read some text. You can change the voice by passing an available SpeechSynthesisVoice (from the voices array). Note that some browsers require a direct user action initiate SpeechSynthesis. To make sure it works, it is recommended that you call speak for the first time in a click event handler.

cancel

function()

Call to make SpeechSynthesis stop reading.

speaking

boolean

True when SpeechSynthesis is actively speaking.

supported

boolean

Will be true if the browser supports SpeechSynthesis. Keep this in mind and use this as a guard before rendering any control that allow a user to call speak.

voices

[SpeechSynthesisVoice]

An array of available voices which can be passed to the speak function. An example SpeechSynthesisVoice voice has the following properties.

{
  default: true
  lang: "en-AU"
  localService: true
  name: "Karen"
  voiceURI: "Karen"
}

In some browsers voices load asynchronously. In these cases, the array will be empty until they are available.

useSpeechRecognition

A react hook for the browser's SpeechRecognition API.

Usage

import React, { useState } from 'react';
import { useSpeechRecognition } from 'react-speech-kit';

function Example() {
  const [value, setValue] = useState('');
  const { listen, listening, stop } = useSpeechRecognition({
    onResult: (result) => {
      setValue(result);
    },
  });

  return (
    <div>
      <textarea
        value={value}
        onChange={(event) => setValue(event.target.value)}
      />
      <button onMouseDown={listen} onMouseUp={stop}>
        🎤
      </button>
      {listening && <div>Go ahead I'm listening</div>}
    </div>
  );
}

Args

onEnd

function()

Called when SpeechRecognition stops listening.

onResult

function(string)

Called when SpeechRecognition has a result. It is called with a string containing a transcript of the recognized speech.

Returns

useSpeechRecognition returns an object which contains the following:

listen

function({ interimResults: boolean, lang: string })

Call to make the browser start listening for input. Every time it processes a result, it will forward a transcript to the provided onResult function. You can modify behavior by passing the following arguments:

  • lang
    string
    The language the SpeechRecognition will try to interpret the input in. Use the languageCode from this list of languages that Chrome supports (here) e.g: "en-AU". If not specified, this defaults to the HTML lang attribute value, or the user agent's language setting if that isn't set either.

  • interimResults
    boolean (default: true)
    SpeechRecognition can provide realtime results as it's trying to figure out the best match for the input. Set to false if you only want the final result.

stop

function()

Call to make SpeechRecognition stop listening. This will call the provided onEnd function as well.

listening

boolean

True when SpeechRecognition is actively listening.

supported

boolean

Will be true if the browser supports SpeechRecognition. Keep this in mind and use this as a guard before rendering any control that allow a user to call listen.