useSpeechRecognition

CHANGELOG.md

@@ -1036,3 +1036,9 @@ Errored release
 ### Fixes
 
 - package.json specifiers (exports)
+
+## [4.3.0] - 2023-03-19
+
+### Adds
+
+- `useSpeechRecognition` hook

docs/useSpeechRecognition.md

@@ -0,0 +1,40 @@
+# useSpeechSynthesis
+
+A hook that provides an interface for using the [Web_Speech_API](https://developer.mozilla.org/en-US/docs/Web/API/Web_Speech_API) to
+recognize and transcribe speech in a user's browser.
+
+### Why? 💡
+
+- Abstracts the implementation details of the Web Speech API into a single reusable function.
+
+### Basic Usage:
+
+```jsx harmony
+import { Button, Space, Tag, Typography, Input } from 'antd';
+import useSpeechRecognition from 'beautiful-react-hooks/useSpeechRecognition';
+
+const SpeechSynthesisDemo = () => {
+  const [name, setName] = React.useState('Antonio');
+  const { startRecording, transcript, stopRecording, isRecording, isSupported } = useSpeechRecognition();
+
+  return (
+    <DisplayDemo title="useSpeechSynthesis">
+      <Space direction="vertical">
+        <Typography.Paragraph>
+          Supported: <Tag color={isSupported ? 'green' : 'red'}>{isSupported ? 'Yes' : 'No'}</Tag>
+        </Typography.Paragraph>
+        <Button onClick={!isRecording ? startRecording : stopRecording} type="primary">
+          {isRecording ? 'Stop' : 'Start'} recording
+        </Button>
+        <Typography.Paragraph>
+          {transcript}
+        </Typography.Paragraph>
+      </Space>
+    </DisplayDemo>
+  );
+};
+
+<SpeechSynthesisDemo />
+```
+
+<!-- Types -->

src/useSpeechRecognition.ts

@@ -0,0 +1,85 @@
+import { useCallback, useEffect, useMemo, useState } from 'react'
+
+declare global {
+  interface SpeechRecognitionEvent extends Event {
+    results: SpeechRecognitionResultList
+  }
+
+  interface SpeechRecognitionPolyfill {
+    start: () => void
+    stop: () => void
+    abort: () => void
+    addEventListener: (event: string, callback: (event: SpeechRecognitionEvent) => void) => void
+    removeEventListener: (event: string, callback: (event: SpeechRecognitionEvent) => void) => void
+
+    // eslint-disable-next-line @typescript-eslint/no-misused-new
+    new(): SpeechRecognitionPolyfill
+  }
+
+  interface Window {
+    SpeechRecognition?: SpeechRecognitionPolyfill
+    webkitSpeechRecognition?: SpeechRecognitionPolyfill
+  }
+}
+
+const SpeechRecognition = window.SpeechRecognition ?? window.webkitSpeechRecognition
+
+/**
+ * A hook that provides an interface for using the Web Speech API to recognize and transcribe speech in a user's browser.
+ */
+const useSpeechRecognition = () => {
+  const spInstance = useMemo(() => SpeechRecognition ? new SpeechRecognition() : null, [])
+  const [isRecording, setIsRecording] = useState(false)
+  const [transcript, setTranscript] = useState('')
+  const isSupported = !!spInstance
+
+  useEffect(() => {
+    const getResults = (event: SpeechRecognitionEvent) => {
+      const nextTranscript = event.results[0][0].transcript
+      setTranscript(nextTranscript)
+    }
+
+    if (spInstance && isSupported) {
+      spInstance.addEventListener('result', getResults)
+    }
+    return () => {
+      if (spInstance && isSupported) {
+        spInstance.stop()
+        spInstance.abort()
+        spInstance.removeEventListener('result', getResults)
+      }
+    }
+  }, [spInstance])
+
+  const startRecording = useCallback(() => {
+    if (spInstance && isSupported) {
+      spInstance.start()
+      setIsRecording(true)
+    }
+  }, [spInstance])
+
+  const stopRecording = useCallback(() => {
+    if (spInstance && isSupported) {
+      spInstance.stop()
+      setIsRecording(false)
+    }
+  }, [spInstance])
+
+  return Object.freeze<UseSpeechRecognitionResult>({
+    isSupported,
+    transcript,
+    isRecording,
+    startRecording,
+    stopRecording
+  })
+}
+
+export interface UseSpeechRecognitionResult {
+  isSupported: boolean
+  transcript: string
+  isRecording: boolean
+  startRecording: () => void
+  stopRecording: () => void
+}
+
+export default useSpeechRecognition

test/useSpeechRecognition.spec.js

@@ -0,0 +1,35 @@
+import { cleanup, renderHook } from '@testing-library/react-hooks'
+import useSpeechRecognition from '../dist/useSpeechRecognition'
+import SpeechSynthesisUtteranceMock from './mocks/SpeechSynthesisUtterance.mock'
+import SpeechSynthesisMock from './mocks/SpeechSynthesis.mock'
+import assertHook from './utils/assertHook'
+
+describe('useSpeechRecognition', () => {
+  const originalSpeechSynth = global.speechSynthesis
+  const originalSpeechSynthesisUtterance = global.SpeechSynthesisUtterance
+
+  before(() => {
+    global.speechSynthesis = SpeechSynthesisMock
+    global.SpeechSynthesisUtterance = SpeechSynthesisUtteranceMock
+  })
+
+  beforeEach(() => cleanup())
+
+  after(() => {
+    global.SpeechSynthesisUtterance = originalSpeechSynthesisUtterance
+    global.speechSynthesis = originalSpeechSynth
+  })
+
+  assertHook(useSpeechRecognition)
+
+  it('should return an object containing the speak function and the utter', () => {
+    const { result } = renderHook(() => useSpeechRecognition())
+
+    expect(result.current).to.be.an('object')
+    expect(result.current.startRecording).to.be.a('function')
+    expect(result.current.stopRecording).to.be.a('function')
+    expect(result.current.transcript).to.be.a('string')
+    expect(result.current.isRecording).to.be.a('boolean')
+    expect(result.current.isSupported).to.be.a('boolean')
+  })
+})