mockflow v1.1.0

Mockflow

Mockflow lets you send mock requests to your Dialogflow fulfillment code, for offline unit testing.

The Basics

Connect to your fulfillment code and send a mock intent request

const Mockflow = require('mockflow');
// Import the Firebase Functions code for your Dialogflow Fulfillment Webhook
// This assumes that the file is in the same directory as index.js - change as necessary
const app = require('.');

// Initialize Mockflow with your Dialogflow project's name and the webhook handler
// Change app.handler if you exported your handler with a custom name
const mock = new Mockflow('my-voice-app', app.handler);

// Send an intent request and check results
it("Enter a description of what should happen here", async function () {
    let result = await mock.intent('My Intent Name').send()
    
    // Provide the result and a string array with keywords/phrases to look for in in the response
    return mock.existsAndMatches(result, ['words','to','look for'])
});

Further customization

The above example creates a simple intent request using sensible defaults for Dialogflow's many parameters. However, you can use the request builder's with...() methods to set the other parameters.

it("Should respond by saying that my favorite color is blue", async function () {
    // Send an intent request
    let result = await mock.intent('Favorite Color Intent')
        .withParameter('Color', 'blue')
        .withQueryText('i like blue')
        .send()

    return mock.existsAndMatches(result, ['favorite','color','blue'])
});

Available Overrides

• withQueryText(queryText)
• withParameter(name, value)
• withAllRequiredParamsPresent(allRequiredParamsPresent)
• withFulfillmentText(fulfillmentText)
• withFulfillmentMessage(fulfillmentMessage)
• withOutputContext(outputContext)
• withIntentDetectionConfidence(intentDetectionConfidence)
• withDiagnosticInfo(name, value)
• withLanguageCode(languageCode)
• withOriginalDetectIntentRequest(originalDetectIntentRequest)

Limitations

• Only Dialogflow v2 fulfillments are supported. v1 has not been tested and probably won't work.

Acknowledgements

• Carter Kwon - Chai integration