Mocking data in Azure Logic Apps

In any distributed system development, various pieces of an application gets developed by multiple teams from different parts of the geography. This is a OK till you hit a roadblock when you cannot proceed until another team finishes their work and provide the data you need. In today's microservices design model this is even more relevant where you need a dependent service to return the data you need so that you can complete your work.

This type is scenario is inevitable and you cannot help but rather have to live with it. So what to do? 😒 

Mocking!! Yes you are right. Why not mock the data you need to finish your work and later just do the plumbing work with the actual service when it's up and ready to throw back actual data. This is a great idea where teams can work on their own schedule without waiting and running after each other.

In Azure, when it comes to mocking, developers often fall back on API Management to implement this with the help of APIM polices. But why going for such costly idea where the need is only to mock data when you can use your other friend Logic Apps to achieve the same. Let's check it out 👍

Scenario:

We need to get the employee details from the database service (developing by the back-end team) by passing the employee code. But the back-end team is not yet there and hence we need to mock the employee details.

Let's get started.

Define an http endpoint where we will send the employee code as below

Add a new action and search for the HTTP action

Fill the details as per your requirement. In my case it's a GET call with the following details.

As mentioned the URI is the service that's getting developed by the backend team which is not yet ready. So, in this action we have to inject out mocking logic. 

Click the "..." eclipse at the top right section of the action and select "Static result (Preview)" from the context menu. Yeah it's preview but works just fine. Go ahead!

Enable the Static result to get the following window. The default selections are good to go for our mocking purpose

But strangely there's no place to provide the static json response content here. Don't worry! It's there but we need to tweak a bit 😉

Hit the JSON mode icon to switch to the JSON view mode as below

Now here in the outputs section paste your return JSON in a body tag as below. I just need the employee first_name, last_name , "dept" and "rank" from the database service. So, I have mocked that piece as the back-end service is not yet ready and I can continue with my development 😃

Hit Done as return to the workflow. Now you will find a new icon added to the action which denotes that this action has a static result configured

As a last step let's add an "Response" action in the workflow and return back the employee details to the caller. 

Save your logic app.

Done! let take a bird's eye view of our logic app workflow. This will help you to match if you are doing it in parallel along with this article

Testing time! Copy the http endpoint from the trigger (first action) and paste it in any rest client to make the call. I have used postman rest client as below. Hit Send and get your static result back.

That's it. You have successfully implemented mocking and what a cool way to do so. 

Isn't that great! 

Congratulations! for coming this far. Hope this article will help you to further explore more on this feature.

Do share with me about your experience and what you have built upon this foundation. You can take it upto any level and integrate. I would love to hear from you.

Inject Inline javascript code in your Azure Logic App workflow

JavaScript is web developers all time favourite. Writing small snippets of JS code into your application to implement few cool functionality is very tempting and developers; I know one cannot think of any web application which do not have any JS flavor in some form or the other 😀 

Wouldn't it be great if we can use our JavaScript knowledge and inject a JS snippet into an Azure Logic App workflow and pull a magic out of it. I know you are thinking "Yessss! Why Not", so lets see how it an be done. 

[Please note that is article expects you to have the basic understanding of Azure Logic Apps and Functions. You can check on Microsoft Azure documentation for such] 

Our scenario for this article:

We need an Http endpoint which will accept a temperature and will do the Celsius or Fahrenheit conversion accordingly. If we send the temperature as Celsius then it will give us its Fahrenheit value and vice-versa.

In most cases I have seen developers quickly jumps into writing Azure Function to do this kind of simple stuff but after going through this article I hope you would think twice.

Our JSON payload will be as below which we will send while calling the Logic App endpoint. It's just sending the temperature in Celsius and expecting its respective Fahrenheit value.

{
   "temperature":"40",
   "metric":"C"
}

Now the JavaScript snippet to do the conversion is as below. It's just applying the formula and doing the respective Celsius and Fahrenheit conversion after reading the temperature and metric from the http request. If metric is coming as "C" then the result will be it's respective Fahrenheit and vice-versa.

var temp = workflowContext.trigger.outputs.body.temperature; 
var metric = workflowContext.trigger.outputs.body.metric;
var cf; 
if (metric == "C")
    cf = (temp * 9/5) + 32 + " degree Fahrenheit";
else if(metric == "F")
    cf = (temp - 32) * 5/9 + " degree Celsius";
else
    cf = "Invalid input"
return cf;

That's it! really 😲

Now lets create a Logic App and inject this piece of JavaScript snippet. There is an "Inline Code" Operation with an Action "Execute JavaScript Code (preview)" in Logic App workflow which we will use to achieve our result. Yes it's in preview but works just fine. Don't worry... you can safely use it 👍

Below is the full workflow and by looking at it you can pretty well understand what's it doing. So simple! 

Save your Logic App and copy the Http Post Url to test it. I will use Postman rest client to test the endpoint but you can use any rest client of your choice.

Test 1: Sending 20 degree Celsius and getting it's respective Fahrenheit value

Test 2: Sending 68 degree Fahrenheit and getting it's respective Celsius value

Cool! the outputs are as expected. With few line of JS snippet you are able to pull out the magic of temperature conversion which would otherwise need an Azure Function.

Isn't that great! 

Great Job and Congratulations! for coming this far. Hope this article will help you to further explore more on this feature.

Do share with me about your experience and what you have built upon this foundation. You can take it upto any level and integrate. I would love to hear from you.