Resolvers: Turning Queries into Data

What a resolver is

The schema says what is available; resolvers say how to fetch it. A resolver is a function attached to a field that returns that field's value – from a database, another API, or memory.

A resolver map (JavaScript)

const resolvers = {
  Query: {
    user: (parent, args) => db.users.find(args.id),
    posts: () => db.posts.all(),
  },
  User: {
    posts: (user) => db.posts.byAuthor(user.id),
  },
};

Each resolver receives the parent object, the args passed to the field, and a shared context (holding things like the logged-in user and database handles).

Resolving a query tree

GraphQL executes a query by walking its tree, calling a resolver for each field. For user(id:1) { posts { title } }:

1. Query.user runs, returning a user.
2. That user becomes the parent for User.posts, which fetches the posts.
3. Each post resolves its title (often just reading a field).

You only write resolvers where the default (read the field off the parent object) is not enough.

The N+1 problem

Walking the tree naively can fire one query per item – fetch 10 users, then 10 separate queries for their posts. This is the classic N+1 problem. The standard fix is a DataLoader, which batches and caches those lookups into a single query per level.

Key points

• Resolvers are functions that fetch the value for a schema field.
• Each receives the parent object, field arguments and a shared context.
• GraphQL walks the query tree, resolving fields top-down.
• Batch lookups with a DataLoader to avoid the N+1 query problem.