Sunday, January 10, 2016

Handling ADF BC 12.2.1 REST Validation in Oracle JET

CRUD use case would not be complete without validation logic implementation. Oracle JET allows to implement standard and custom validators on the client. Probably most of the simple to average complexity logic will be implemented in Oracle JET. Complex use cases can be handled in ADF BC validation rules, data will be validated through REST calls. When validation fails in ADF BC, error message is propagated through REST back to Oracle JET client, where we can parse it and attach to the UI field.

You can watch demo video, it shows how it works. I'm trying to update a row with invalid data, rules are executed in ADF BC and validation messages are displayed next to the UI fields in JET:

There are three rules defined in ADF BC. It might be tricky to map UI field with error message. This is easy in case of PATCH method, ADF BC REST returns failed attribute name together with the message. This doesn't work in the same way with POST method, when new row is created. For this reason, I have introduced custom indicator into each validation message *AttributeName*. I'm parsing this value in JET, to understand which UI field should be marked as incorrect.

Rule 1. Unique Key validation rule for Email attribute.

Rule 2. Hire Date validation rule, it should not be in the future.

Rule 3. Salary range validation rule.

Mandatory check is performed in JET, if mandatory UI field is not set - JET will prevent further call to REST service. I have achieved this with JET invalidComponentTracker property defined for JET UI input fields:

Tracker variable is defined and accessed in JavaScript, here we are going to evaluate if there are any issues on UI components, before calling REST service. There are three variables defined to report valdiation messages from ADF BC REST for Hire Date, Email and Salary fields:

In update function, before calling REST - I check if there are no issues with UI components by obtaining tracker:

Method showMessages is invoked for tracker, if there are issues - focus is set on the first one. If this doesn't happen, it means there are no issues and further REST call is allowed:

This is how it looks on UI. I'm trying to update empty row, there are mandatory fields and built-in functionality prevents me from submitting data:

Let's move to the main topic of this post - ADF BC REST call validation error handling. JET allows to define callbacks for REST call action - success/error. If ADF BC REST call fails with validation issue, error callback is invoked. Here we can get response text, parse it and assign to the UI fields. Multiple validation rules can fail at once in ADF BC, REST returns multiple messages in one response. This is good, we can display all errors at once:

Messages are attached to UI field through special property (and variable defined above) - messageCustom. This must be set on JET UI field:

In JavaScript callback, we are parsing response, constructing error message and assigning it to the messageCustom of specific UI field. This is how message is displayed on UI, next to the field:

Here is example of all three validation rules failing and messages are displayed on JET UI:

All three are reported in REST call response by ADF BC validation. We just need to parse returned message in Oracle JET and display it:

Download sample application -

No comments: