Effective RESTful Error Handling in Go | alesr
Go’s approach to error handling tends to spark strong opinions. Whether you love it or hate it, many developers struggle with propagating and processing errors while ensuring meaningful responses are returned to clients, all while keeping application details secure.
In this post, I’ll explain the approach I typically use to turn application errors into JSON responses, along with the benefits of this method.
For example, if an API request fails because a record doesn’t exist, we should convey this with an HTTP Not Found status to the client. On the other hand, unexpected issues like connection errors should result in an Internal Server Error (500) to avoid exposing sensitive information.
You can access the implementation of this error handler here. A complete demonstration of an application using this error handler can be found here.
To achieve this, the idea is to implement an error handler that can be attached to the transport layer. Each time the transport layer receives a request, it calls the service (or domain) layer to process it. When a result is returned, the handler logs the original error and sends a JSON payload mapped to the error returned by the service.
