soundoftext-js

A NodeJS library for the SoundOfText API.

Install

Install with npm:

npm install soundoftext-js

or yarn:

yarn add soundoftext-js

Usage

This library supports four operations: create, location, request, and status.

You will most likely want to use create, as it has the simplest interface.

sounds.create

This function requests for a sound to be created, and polls the API until the sound is finished being created, eventually returning a URL that links to an MP3 file.

This function takes an object with two properties:

text - the text to be spoken
voice - the voice (language + accent) to use.

and returns the url for an MP3 file that you could then download.

const client = require('soundoftext-js');

client.sounds.create({ text:'Hello, world!', voice: 'en-US' })
  .then(soundUrl => {
    console.log(soundUrl); // https://soundoftext.nyc3.digitaloceanspaces.com/<sound-id>.mp3
  })
  .catch(e => {
    /* Reasons that the Promise might get rejected:
     * - after 60 seconds, it automatically times out
     * - the API might fail to create the sound or reject it
     * - other miscellaneous network issues
     */
  });

sounds.request

This function requests for a sound to be created, and returns an object containing the sound id.

This function takes an object with two properties:

text - the text to be spoken
voice - the voice (language + accent) to use

and returns an object containing the sound id.

const client = require('soundoftext-js');

client.sounds.request({ text: 'Hello, world!', voice: 'en-US' })
  .then(response => {
    console.log(response);
    /* One of:
     * { success: true, id: '<sound-id>' }
     * { success: false, message: '<error-message>' }
     */
  })
  .catch(e => {
    /* Reasons that the Promise might get rejected:
     * - API rejects the request
     * - other miscellaneous network issues
     */
  });

sounds.status

This function takes a sound id and returns the current status.

This function takes an object with one property:

id - the id for the sound

and returns an object containing the status of the sound.

const client = require('soundoftext-js');

client.sounds.request({ text: 'Hello, world!', voice: 'en-US' })
  .then(response => {
    return client.sounds.status({ id: response.id });
  })
  .then(status => {
    console.log(status);
    /* One of:
     * { status: 'Error', message: '<error-message>' }
     * { status: 'Pending' }
     * { status: 'Done', location: '<url-for-mp3-file>' }
     */
  })
  .catch(e => {
    /* Reasons that the Promise might get rejected:
     * - API rejects the request
     * - other miscellaneous network issues
     */
  });

sounds.location

This is a convenience wrapper for sounds.status, which starts polling regularly for the status to be 'Done', before returning the url for the mp3 file. It rejects the promise if it times out (~60 seconds) or if the API returns an 'Error' status.

This function takes an object with one property:

id - the id for the sound

and returns the url for the MP3 file.

const client = require('soundoftext-js');

client.sounds.request({ text: 'Hello, world!', voice: 'en-US' })
  .then(response => {
    return client.sounds.location({ id: response.id });
  })
  .then(location => {
    console.log(location); // https://soundoftext.nyc3.digitaloceanspaces.com/<sound-id>.mp3
  })
  .catch(e => {
    /* Reasons that the Promise might get rejected:
     * - after 60 seconds, it automatically times out
     * - the API might fail to create the sound or reject it
     * - other miscellaneous network issues
     */
  });

FAQ

What voices does this support?

You can find a list of language codes in the documentation for Sound of Text.

You could also use another package of mine called google-tts-langauges that is another JS library that exports all the language codes.

florabtw/soundoftext-js