Introduction to GraphQL

This post followed by this article. It’s a series so it contians example from the previous one.

GraphQL

GraphQL provides an API interface that users can submit queries to (like SQL) and receive data from.

query {
    allQuestions {
        questionText,
        pubDate
    }
}

query {
    resolver {
        fields
    }
}

The first example shows a practical example and the following is grammar to show what is matching. It means the ‘allQuestions’ is called a resolver. The two inside of the bracket is called fields. The syntax looks like a JSON. The ‘query’ is a reserved keyword. This command will return a JSON result with the ‘data’ attribute.

{
    "data": {
        "allQuestions": [
            {
                "questionText": "Do you like Django?",
                "pubDate": "2020-03-22T07:22:19+00:00"
            }
        ]
    }
}

REST vs GraphQL

GraphQL	REST
Single API End-Point	Little setup
No Over/Under Fetching Issues	More “magic” abstraction
Automatically Cascades Relationships	Error handling built in (generics)
Extremely Flexible API	Much more modular
Building momentum	Sane and Stable generic API

REST has multiple API end-point. GraphQL does not have multiple end-point. GraphQL does support cascaded relationship for fetching and this bring over or under fetching issues for REST.

GraphQL is released at 2015, 5 years ago but REST API has a much longer history. Building a momentum means the GraphQL has a relatively short history rather than the REST.

To more deep-dive inspection between two, this article is a wonderful article full of practical examples.

Graphene

To use graphQL in Django, we will use the graphene library. It provides a GraphiQL interface that shows us how to submit a command to fetch and a correspondence result.