nanospeech

Web speech api simplified

Usage

Text to speech

import { speak } from 'nanospeech'
 
speak('hello world')

Speech to text

import { recognition } from 'nanospeech'
 
recognition.start()
 
recognition.on('result', text => console.log(text))

Install

Include the following tags in your html

<script type="module" src="https://unpkg.com/nanospeech?module"></script>
<script nomodule src="https://unpkg.com/nanospeech@1.0.0/dist/nanospeech.umd.js"></script>

This will install the module version when the browser support modules and fallback to the umd version when it don't. You can still import or require this

import { speak } from 'nanospeech'
 
// or
 
var speak = require('nanospeech').speak

API

let speech = speak(text, [opts])

The speak function speak a text passed as the first argument. Optionally the following options can be passed as an object on the second argument:

• lang(DOMString): A BCP47 format string (eg. en-US) to set the language to be spoken. Notice that the selected language must be supported by the availaible voices, otherwise, default voice and language will be used.
• rate(float): The speed of the utterance to be spoken, can be a value from 0.1 to 10, defaults to 1.
• pitch(float): The pitch of the utterance to be spoken, can be a value from 0 to 2, defaults to 1.
• volume(float): The volume of the utterance to be spoken, can be a value from 0 to 1, defaults to 0.5.

The speech object returned is an Event Emitter like object. It has two single methods on and emit (There is no off method because there is actually no event going on, so no need to remove listeners)

speech.on(event, handler)

event is a string with the name of the event to listen to. handler is a function which always get a SpeechSynthesisEvent as its argument. Availaible events are:

• start: Emitted when a speech starts.
• boundary: Emitted when a word or sentence boundary is reched during a speech.
• end: Emitted when a speech finish being spoken.

speech.emit(event)

event is a string with the name of the event to be emited. Also, every string has a function version, making speech.emit('pause') equivalent to speech.pause()

• cancel: Emit this event to cancel all qeued speechs.
• pause: Emit this event to pause an ongoing speak. If there is no speaking taking place, it will do nothing.
• resume: Emit this event to resume an ongoing speak. If there is no paused speaking, it will do nothing.

recognition.on(event, handler)

event is a string with the name of the event to listen to. handler is a function which always get a SpeechRecognitionEvent as its argument, with the exception of the result event which gets the text recognized as a string for the argument. Availaible events are:

• nomatch: Fired when the speech recognition service returns a final result with no significant recognition.
• audiostart: Fired when the user agent has started to capture audio.
• audioend: Fired when the user agent has finished capturing audio.
• soudnstart: Fired when any sound — recognisable speech or not — has been detected.
• soundend: Fired when any sound — recognisable speech or not — has stopped being detected.
• speechstart: Fired when sound that is recognised by the speech recognition service as speech has been detected.
• speechend: Fired when speech recognised by the speech recognition service has stopped being detected.
• end: Fired when the speech recognition service has disconnected.
• result: Fired when the speech recognition service returns a result.

recognition.emit(event)

event is a string with the name of the event to be emited. Also, every string has a function version, making speech.emit('start') equivalent to speech.start()

• start: Fired when the speech recognition service has begun listening to incoming audio.
• stop: Emit this event to stops the speech recognition service from listening to incoming audio, and attempts to return a SpeechRecognitionResult using the audio captured so far.
• abort: Stops the speech recognition service from listening to incoming audio.

LICENSE

MIT