graphql-advanced-projection

Fully customizable Mongoose/MongoDB projection generator.

Why

We already have graphql-projection, graphql-mongodb-projection, and graphql-db-projection.But graphql-advanced-projection is different from all of them above in the following ways:

• Separete graphql schema and mongodb projection config. This helps you decouple schema and mongodb into two parts, each of them may change independently. Write graphql in .graphql, write config in javascript or .json.
• Easy customization. No more gqlField: { type: new GraphQLNonNull(GraphQLInt), projection: 'mongoField' }. Simply gqlField: 'mongoField'.
• We create resolvers. gqlField: (parent) => parent.mongoField can be automatically generated, even complicated ones like first: (parent) => parent.items.data[0].value.
• Fully supports interfaces, fragments, and inline fragments. Write typeProj: 'type' and switch (parent.type) in __resolveType.

Installation

$ yarn add graphql-advanced-projection

Usage

For a complete working demo, see the examples folder.

Setup mongoose

const UserSchema = new mongoose.Schema({
  _id: String,
  mongoA: String,
});
const User = mongoose.model('users', UserSchema);

Setup graphql

type Query {
  user(id: ID!): User
}
type User {
  userId: ID
  field1: String
  field2: String
}

Setup graphql-advanced-projection

const { project, resolvers } = gqlProjection({
  User: {
    proj: {
      userId: '_id',
      field1: 'mongoA',
      field2: null,
    },
  },
});

Combine everything together

module.exports = makeExecutableSchema({
  typeDefs,
  resolvers: _.merge(resolvers, {
    Query: {
      async user(parent, args, context, info) {
        const proj = project(info);
        const result = await User.findById(args.id, proj);
        return result.toObject();
      },
    },
    User: {
      field2: () => 'Hello World',
    },
  }),
  resolverValidationOptions: { requireResolversForResolveType: false },
});

Run

query {
  user(id: $id) {
    field1
    field2
  }
}

proj = {
  _id: 0,
  mongoA: 1,
}

License

MIT