Chapter 9. Getting Started with Data Transformation
One of the challenges that comes with system and data integration is that the component systems often work with different data formats. You cannot simply send messages from one system to another without translating it into a format (or language) recognized by the receiving system. Data transformation is the term given to this translation.
In this chapter, you learn how to include data transformation in a predefined Camel route. The Camel route directs messages from a source endpoint that produces XML data to a target endpoint that consumes JSON data. You add and define a data transformation component that maps the source’s XML data format to the target’s JSON data format.
9.1. Creating a project for the data transformation example
Create a new Fuse Integration Project (select File
New Fuse Integration Project). Provide the following information in the wizard:
- Project name: starter
- Deployment platform: Standalone
- Runtime environment: Karaf/Fuse on Karaf
- Camel version: Use the default
- Template: Empty - Blueprint DSL
-
Download the prepared data examples from:
https://github.com/FuseByExample/fuse-tooling-tutorials/archive/user-guide-11.1.zip
-
Extract the
data
folder and the three files that it contains from theuser-guide-11.1.zip
archive into the Fuse Integration project’ssrc
directory (starter/src/data
). - In the Project Explorer view, expand the starter project.
-
Double-click Camel Contexts
src/main/resources/OSGI-INF/blueprint/blueprint.xml
to open the route in the route editor’s Design tab. - Click the Source tab to view the underlying XML.
Replace
<route id="_route1"/>
with the following code:<route id="_route1"> <from id="_from1" uri="file:src/data?fileName=abc-order.xml&noop=true"/> <setHeader headerName="approvalID" id="_setHeader1"> <simple>AUTO_OK</simple> </setHeader> <to id="_to1" uri="file:target/messages?fileName=xyz-order.json"/> </route>
Click the Design tab to return to the graphical display of the route:
9.2. Adding a data transformation node to the Camel route
- In the Palette, expand the Transformation drawer.
Click the Data Transformation pattern and then, in the canvas, click the arrow between the
SetHeader _setHeader1
andTo_to1
nodes.The New Transformation wizard opens with the Dozer File Path field auto-filled.
Fill in the remaining fields:
-
In the Transformation ID field, enter
xml2json
. - For Source Type, select XML from the drop-down menu.
- For Target Type, select JSON from the drop-down menu.
-
In the Transformation ID field, enter
Click Next.
The Source Type (XML) definition page opens, where you specify either an XML Schema (default) or an example XML Instance Document to provide the type definition of the source data:
- Leave XML Schema enabled.
For Source file, browse to the location of the XML schema file or the XML instance file to use for the type definition of the source data, and select it (in this case,
abc-order.xsd
).The XML Structure Preview pane displays a preview of the XML structure.
In the Element root field, enter
ABCOrder
.The tooling uses this text to label the pane that displays the source data items to map.
The Source Type (XML) definition page should now look like this:
Click Next to open the Target Type (JSON) definition page. This is where you specify the type definition for the target data.
Click JSON Instance Document.
In the Target File field, enter the path to the
xyz-order.json
instance document, or browse to it. The JSON Structure Preview pane displays a preview of the JSON data structure:- Click Finish.
The transformation editor opens. This is where you can map data items in your XML source to data items in your JSON target.
![Description](https://access.redhat.com/webassets/avalon/d/Red_Hat_Fuse-7.4-Tooling_User_Guide-en-US/images/00614bcb54a3394c126676fe27433357/DTtransfomEditor1.png)
The transformation editor is composed of three panels:
- Source — lists the available data items of the source
- Mappings — displays the mappings between the source and target data items
- Target — lists the available data items of the target
In addition, the editor’s details pane, located just below the editor’s three panels (once the first mapping has been made), graphically displays the hierarchical ancestors for both the mapped source and target data items currently selected. For example:
![Details pane with source property customerNum mapped to target property role="italic">custId](https://access.redhat.com/webassets/avalon/d/Red_Hat_Fuse-7.4-Tooling_User_Guide-en-US/images/87179fa4429322e639662b7e235209a1/DTdetailsView1.png)
Using the details pane, you can customize the mapping for the selected source and target data items:
- Set property — Modify an existing mapping or map a simple data item to one in a collection (see Section 9.8, “Mapping a simple data item to a data item in a collection”).
- Set variable — Specify a constant value for a data item (see Section 9.5, “Mapping a constant variable to a data item”).
- Set expression — Map a data item to the dynamic evaluation of a specified expression (see Section 9.6, “Mapping an expression to a data item”).
- Add transformation — Modify the value of a mapped data item using a built-in function (see Section 9.9, “Adding a built-in function to a mapped data item”).
- Add custom transformation — Modify the value of a mapped data item using the Java method you create or one you previously created (see Section 9.7, “Adding a custom transformation to a mapped data item”).
9.3. Mapping source data items to target data items
Expand all items in the Source and Target panels located on left and right sides of the Mappings panel.
Drag a data item from the Source panel and drop it on its corresponding data item in the Target panel.
For example, drag the
customerNum
data item from the Source panel and drop it on thecustId
data item in the Target panel.The mapping appears in the Mappings panel, and the details of both the Source and Target data items appear below in the details pane.
Continue dragging and dropping source data items onto their corresponding target data items until you have completed all basic mappings.
In the
starter
example, the remaining data items to map are:Source Target orderNum
orderId
status
priority
id
itemId
price
cost
quantity
amount
NoteYou can map collections (data items containing lists or sets) to non-collection data items and vice versa, but you cannot map collections to other collections.
Click
on both the Source and Target panels to quickly determine whether all data items have been mapped.
Only data items that have not been mapped are listed in the Source and Target panels.
In the
starter
example, the remaining unmapped Target attributes areapprovalCode
andorigin
.Click the blueprint.xml tab to return to the graphical display of the route:
-
Click File
Save.
You can run a JUnit test on your transformation file after you create the transformation test. For details, see Section 9.4, “Creating the transformation test file and running the JUnit test”. If you do so at this point, you will see this output in the Console view:
For the source XML data:
<?xml version="1.0" encoding="UTF-8"?> <ABCOrder xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:java="http://java.sun.com"> <header> <status>GOLD</status> <customer-num>ACME-123</customer-num> <order-num>ORDER1</order-num> </header> <order-items> <item id="PICKLE"> <price>2.25</price> <quantity>1000</quantity> </item> <item id="BANANA"> <price>1.25</price> <quantity>400</quantity> </item> </order-items> </ABCOrder>
For the target JSON data:
{"custId":"ACME-123","priority":"GOLD","orderId":"ORDER1","lineItems":[{"itemId":"PICKLE", "amount":1000,"cost":2.25},{"itemId":"BANANA","amount":400,"cost":1.25
9.4. Creating the transformation test file and running the JUnit test
-
Right-click the
starter
project in the Project Explorer view, and select NewOther Fuse Tooling Fuse Transformation Test. - Select Next to open the New Transformation Test wizard.
In the New Transformation Test wizard, set the following values:
Field Value Package
example
Camel File Path
OSGI-INF/blueprint/blueprint.xml
Transformation ID
xml2json
- Click Finish.
-
In the Project Explorer view, navigate to
starter/src/test/java/example
, and open theTransformationTest.java
file. Add the following code to the
transform
method:startEndpoint.sendBodyAndHeader(readFile("src/data/abc-order.xml"), "approvalID", "AUTO_OK");
Click File
Save. You can now run a JUnit test on your transformation file at any point in these tutorials.
-
In the Project Explorer view, expand the
starter
project to expose the/src/test/java/example/TransformationTest.java
file. Right click it to open the context menu, and select Run as JUnit Test.
The JUnit Test pane opens to display the status of the test. To avoid cluttering your workspace, drag and drop the pane in the bottom panel near the Console view.
- Open the Console view to see the log output.
9.5. Mapping a constant variable to a data item
When a source/target data item has no corresponding target/source data item, you can map a constant variable to the existing data item.
In the starter
example, the target data item origin
does not have a corresponding source data item. To map the origin
attribute to a constant variable:
In the Source panel, click the Variables view.
In the Variables view, click
to open the Enter a new variable name dialog.
Enter a name for the variable you want to create.
For the
starter
example, enterORIGIN
.Click OK.
The newly created variable
ORIGIN
appears in the Variables view in the Name column and the default valueORIGIN
in the Value column.-
Click the default value to edit it, and change the value to
Web
. - Press Enter.
Drag and drop the new variable
ORIGIN
onto theorigin
data item in the Target panel.The new mapping of the variable
$(ORIGIN)
appears in the Mappings panel and in the details pane.Run a JUnit test on your
TransformationTest.java
file. For details, see Section 9.4, “Creating the transformation test file and running the JUnit test”.The Console view displays the JSON-formatted output data:
{"custId":"ACME-123","priority":"GOLD","orderId":"ORDER1","origin":"Web", "approvalCode":"AUTO_OK","lineItems":[{"itemId":"PICKLE","amount":1000,"cost":2.25}, {"itemId":"BANANA","amount":400,"cost":1.25}]}
9.6. Mapping an expression to a data item
This feature enables you, for example, to map a target data item to the dynamic evaluation of a Camel language expression.
Use the target approvalCode
data item, which lacks a corresponding source data item:
Click
to add an empty transformation map to the Mappings panel.
From the Target panel, drag and drop the
approvalCode
data item to the target field of the newly created mapping in the Mappings panel.The
approvalCode
data item also appears in the details pane’s target box.In the details pane, click
on the
ABCOrder
source box to open the drop-down menu.Menu options depend on the selected data item’s data type. The available options are bolded.
Select Set expression to open the Expression dialog.
In Language, select the expression language to use from the list of those available. Available options depend on the data item’s data type.
For the
starter
example, select Header.In the details pane, select the source of the expression to use.
The options are Value and Script.
For the
starter
example, click Value, and then enterApprovalID
.Click OK.
Both the Mappings panel and the details pane display the new mapping for the target data item
approvalCode
.Run a JUnit test on your
TransformationTest.java
file. For details, see Section 9.4, “Creating the transformation test file and running the JUnit test”.The Console view displays the JSON-formatted output data:
{"custId":"ACME-123","priority":"GOLD","orderId":"ORDER1","origin":"Web", "approvalCode":"AUTO_OK","lineItems":[{"itemId":"PICKLE","amount":1000,"cost":2.25}, {"itemId":"BANANA","amount":400,"cost":1.25}]}
9.7. Adding a custom transformation to a mapped data item
You may need to modify the formatting of source data items when they do not satisfy the requirements of the target system.
For example, to satisfy the target system’s requirement that all customer IDs be enclosed in brackets:
In the Mappings panel, select the
customerNum
mapping to populate the details pane.In the details pane, click
on the
ABCOrder
source box to open the drop-down menu.Select Add custom transformation to open the Add Custom Transformation page.
Click
next to the Class field to open the Create a New Java Class wizard.
Modify the following fields:
-
Package — Enter
example
. -
Name — Enter
MyCustomMapper
. Method Name — Change
map
tobrackets
.Leave all other fields as is.
-
Package — Enter
Click Finish.
The Add Custom Transformation page opens with the Class and Method fields auto filled:
Click OK to open the
MyCustomMapper.java
file in the Java editor:Edit the
brackets
method to change the last linereturn null;
to this:return "[" + input + "]";
Click the transformation.xml tab to switch back to the transformation editor.
The details pane shows that the
brackets
method has been associated with thecustomerNum
data item.The
brackets
method is executed on the source input before it is sent to the target system.Run a JUnit test on your
TransformationTest.java
file. For details, see Section 9.4, “Creating the transformation test file and running the JUnit test”.The Console view displays the JSON-formatted output data:
{"custId":"[ACME-123]","priority":"GOLD","orderId":"ORDER1","origin":"Web", "approvalCode":"AUTO_OK","lineItems":[{"itemId":"PICKLE","amount":1000,"cost":2.25}, {"itemId":"BANANA","amount":400,"cost":1.25}]}
9.8. Mapping a simple data item to a data item in a collection
In this tutorial, you will modify an existing mapping that maps all id
s in the Source to the itemId
s in the Target. The new mapping will map the customerNum
data item in the Source to the itemId
of the second item in the lineItems
collection in the Target.
With this change, no id
s in the Source will be mapped to itemId
s in the Target.
In the Mappings panel, select the mapping
id
— >itemId
to display the mapping in the details pane.On the Source box, click
to open the drop-down menu, and select Set property.
In the Select a property page, expand the
header
node and selectcustomerNum
. Click OK to save the changes.The details pane now shows that
XyzOrder
has alineItems
field. Click the toggle button next tolineItems
to increase its value to1
.NoteIndexes are zero-based, so a value of
1
selects the second instance ofitemId
in the collection.Notice that the details pane shows
customerNum
mapped to theitemId
of the second item in thelineItems
collection.Run a JUnit test on your
TransformationTest.java
file. For details, see Section 9.4, “Creating the transformation test file and running the JUnit test”.The Console view displays the JSON-formatted output data:
{"custId":"[ACME-123]","priority":"GOLD","orderId":"ORDER1","origin":"Web", "approvalCode":"AUTO_OK","lineItems":[{"amount":1000,"cost":2.25}, {"itemId":"ACME-123","amount":400,"cost":1.25}]}
9.9. Adding a built-in function to a mapped data item
You can use the built-in string-related functions to apply transformations to mapped data items.
In the Transformations panel, select the
status
topriority
mapping to populate the details pane.In the Source box, click
to open the drop-down menu, and select Add transformation.
In the Transformations pane, select
append
, and in the Arguments pane, enter-level
for the value of suffix.This
append
function adds the specified suffix to the end of thestatus
string before mapping it to the targetpriority
data item.Click OK.
By default, the details pane displays the results of adding the
append
function to thestatus
data item in a user-friendly format. You can change this formatting by clicking the right-moston the Source box, and selecting Show standard formatting.
Run a JUnit test on your
TransformationTest.java
file. For details, see Section 9.4, “Creating the transformation test file and running the JUnit test”.The Console view displays the JSON-formatted output data:
{"custId":"[ACME-123]","priority":"GOLD-level","orderId":"ORDER1","origin":"Web", "approvalCode":"AUTO_OK","lineItems":[{"amount":1000,"cost":2.25},{"itemId":"ACME-123", "amount":400,"cost":1.25}]}
9.10. Publishing a Fuse Integration project with data transformation to a Red Hat Fuse server
Before you publish your data transformation project to a Fuse server (see Chapter 29, Publishing Fuse Integration Projects to a Server), you need to install the following features in the Fuse runtime:
-
camel-dozer
-
camel-jackson
-
camel-jaxb
To install the required features on the Fuse runtime:
- If not already there, switch to the Fuse Integration perspective.
- If necessary, add the Fuse server to the Servers list (see Section 28.1, “Adding a Server”).
- Start the Fuse Server (see Section 28.2, “Starting a Server”), and wait for the JBoss Fuse shell to appear in the Terminal view.
For each of the required
camel-
features, at theJBossFuse:admin@root>
prompt type:features:install camel-<featureName>
Where featureName is one of
dozer
,jackson
, orjaxb
.To verify that each of the features was successfully installed, at the
JBossFuse:admin@root>
prompt type:features:list --ordered --installed
You should see the camel features you just installed in the output listing: