Mule ESB Tutorial: Introduction To DataWeave

Introduction

MuleSoft's DataWeave available only from Anypoint Studio 3.7 since May 2015. It is a new powerful tool for data transformation. Earlier, we use DataMapper, or customer transformer to perform data transformation in addition to the out-of-box transformers, like Byte Array to String, object to JSON, XML to JSON, etc., just to name a few.

With DataWeave, the transformation become much less painful. In particular, if you don't want to write customer transform in java or scripts. Still, if you are good at writing java code, like me, customer transformer still be the best from development point of view. But it may not be the best for the long term maintenance. Once you get to the habits to write write customer transformer, you tend to use it a lot. I think it is time to change with this kind of powerful tool.

In this tutorial, I am going to demonstrate the basics about the DataWeave. Later, I will introduce more complicated cases. I will put this tutorial to Youtube.com. If you prefer to watch the usage in action.

Using Using Data Weave

The flow is like this:

With this configuration, we will use the input of JSON like the follwoing:

{
 "abc":
     {
     "cba":"ddd",
     "ccc":"eee",
     "aaa":["fgh","ghf","hgf"]
     }
}

Put the above file into your local somewhere name it as transform1.json. I strongly recommend a tool named: subline test 2 for edition JSON.

Now create HTTP connector. Everything is the same as simple configuration. Only extra procedure is add Metadata as the following:

Now, we can add Transform Message to the flow. Note that the "Transform Message" which is DataWeave component, it is not a transformer!

Simple Transform Doing Nothing

For the doing nothing case, the transformer will be the following:

%dw 1.0
%input payload application/json
%output application/xml
---
payload

Here is the snapshot of the transform panel:

Now we can test using PostMan with url: http://lcoalhost:8082/transform1. You should see the xml output like the following:

    ddd
    eee
    
        fgh
        ghf
        hgf
    

All above is nothing fancy. Note, the payload in the transform panel. It means doing nothing. The power comes from that panel. All the logics will be input there. We will see in the next section.

Transform From JSON To XML

Now let's change the transform scripts to the following:

%dw 1.0
%input payload application/json
%output application/xml
---
items: {
 name: payload.abc.aaa[1],
 address: payload.abc.cba,
 ID: payload.abc.ccc
}

Now if you run the POST operation, you will get the following result:

    ghf
    
ddd
    eee

This is the transformation a bit interesting. Of course, you can do this with data mapper transformer. But I hope you see the power behind it. It opens up the huge possibility in manipulate data for the transformation purpose. I will cover more cases in my later blogs.

Conclusion

I think DataWeave is a powerful tool. It takes time to learn the syntaxes, but I think it worths the time spending if you want to be a good integration architect! You can visit my youtube channel to get dynamics of the DataWeave!