GQLGen: Mandatory Field Handling & Error Management

GQLGen is a powerful tool for building GraphQL servers in Go. One common challenge developers encounter revolves around handling mandatory fields within their GraphQL schemas. This article dives into the intricacies of this problem, exploring the scenarios where mandatory fields are missing in your resolvers, and offering solutions to ensure data integrity and prevent unexpected errors. We will discuss the current behavior of GQLGen when faced with missing mandatory fields, and then explore the recommended approaches for handling these situations.

The Problem: Null Values and the GraphQL Specification

When working with GraphQL, you define a schema that specifies the types of data your API exposes, as well as the relationships between those types. Within this schema, you designate some fields as mandatory, meaning they must have a value. The absence of a value for a mandatory field can lead to runtime errors, and can potentially crash your application. In this article, we're talking about how to handle scenarios where your resolver, the code that fetches the data for a GraphQL field, fails to provide a value for a mandatory field, leading to errors in the execution of your GraphQL queries.

Specifically, the GraphQL spec does not allow null values for non-null fields. GQLGen, by default, might not always catch these violations during development, leading to runtime exceptions. The default behavior is that the GraphQL server will return an error when a mandatory field is missing. This might not be the most user-friendly experience, as it doesn't clearly indicate why the error occurred. The developer needs to carefully trace back the root cause. GQLGen should provide a better developer experience when it comes to missing mandatory fields.

The ApolloError Scenario

Let's consider the scenario from the original issue. You create a new query stub, and you notice that most of the data is missing. You then receive an ApolloError indicating that a requested element is null, despite the schema disallowing null values. The problem isn't always immediately obvious, and you may end up tracing through your code to find the reason for the error. The error occurs because a mandatory field, as defined in your GraphQL schema, is not being populated by your resolver. This is the core problem and the focus of our discussion.

Solutions and Best Practices

When a mandatory field is missing, there are a couple of approaches to consider. First, you need to determine the best approach for the given context. This approach must address the root problem of missing data and also enhance the developer experience.

Option 1: Auto-Filling with Zero Values (Considerations)

One potential solution is for GQLGen to automatically fill missing mandatory fields with zero values. While this may seem convenient, it may not be suitable in all cases. This approach involves GQLGen supplying a non-null zero value or a pointer for the missing fields, preventing the immediate error. This behavior could be implemented by adding a feature to GQLGen. There are, however, some concerns with this approach. First, you need to define what a