Have ideas to improve npm?Join in the discussion! »

    witspeech

    0.0.1 • Public • Published

    Wit speech

    Makes network request to Wit.ai's speech API. For more information about the API see the official documentation.

    Installation

    npm install --save witspeech
    

    Usage

    Constructor

    // Import module.
    const WitSpeech = require('witspeech');
    // Create an instance.
    const witSpeech = new WitSpeech('WIT_TOKEN', '20171207');

    Replace WIT_TOKEN with your token retrieved from the wit.ai website.

    Methods

    // The type of content that you will stream.
    let contentType = 'audio/wav';
    // Additional and optional web request parameters.
    let queryParameters = {};
    // Callback function with the response.
    let callback = function(error, response) {
      if (error) {
        console.error('ERROR', error);
        return;
      }
      console.log(JSON.parse(response));
    };
     
    // Retrieves the web request to pipe information to.
    witSpeech.request(
      contentType,
      queryParameters,
      callback
    );

    To see what content types you can send over see the official documentation.

    Example

    From file

    // Import node module.
    const fs = require('fs');
     
    // Initialize module, see constructor section for more information.
    const WitSpeech = require('witspeech');
    const witSpeech = new WitSpeech('WIT_TOKEN', '20171207');
     
    // Create request, see methods section for more information.
    let request = witSpeech.request('audio/wav', {}, function(error, response) {
      if (error) {
        console.error('ERROR', error);
        return;
      }
      console.log(JSON.parse(response));
    });
     
    // Create read file stream.
    let stream = fs.createReadStream('audio.wav');
    // Pipe the stream to the request.
    stream.pipe(request);

    From microphone

    // Initialize module, see constructor section for more information.
    const WitSpeech = require('witspeech');
    const witSpeech = new WitSpeech('WIT_TOKEN', '20171207');
     
    // Import audio recorder
    const AudioRecorder = require('node-audiorecorder');
    const audioRecorder = new AudioRecorder();
     
    // Create request, see methods section for more information.
    let request = witSpeech.request('audio/wav', {}, function(error, response) {
      if (error) {
        console.error('ERROR', error);
        return;
      }
      console.log(JSON.parse(response));
    });
     
    // Start audio recorder.
    audioRecorder.start();
    // Get microphone stream.
    let stream = audioRecorder.stream();
    // Pipe the stream to the request.
    stream.pipe(request);

    For the audio recorder see the package or repository for it.

    For another example see the Electron-VoiceInterfacePrototype's input.js.

    Install

    npm i witspeech

    DownloadsWeekly Downloads

    2

    Version

    0.0.1

    License

    MIT

    Last publish

    Collaborators

    • avatar