NestJS Dataloader

!Node.js CI

NestJS dataloader simplifies adding graphql/dataloader to your NestJS project. DataLoader aims to solve the common N+1 loading problem.

This is a fork of krislefeber/nestjs-dataloader with the following changes:

- Replace generateDataLoader method in interface with getBatchFunction, reducing the need to import anything from the dataloader package and simplifying the code
- Added sortDataLoaderResultsByKey convenience method, since some batching methods you can use (like SELECT * IN (...)) do not guarantee the order of the returned results to match the order of provided keys, which is essential for dataloader to work properly
- Updated all dependencies to latest versions

Installation

``bash
npm i @applifting-io/nestjs-dataloader
`

Usage

$3

We start by implementing the NestDataLoader interface. This tells DataLoader how to load our objects.

`typescript
import { Injectable } from '@nestjs/common';
import { NestDataLoader, DataLoaderBatchFunction } from '@applifting-io/nestjs-dataloader';
...

@Injectable()
export class AccountLoader implements NestDataLoader {
constructor(private readonly accountService: AccountService) { }

getBatchFunction(): DataLoaderBatchFunction {
return (keys: string[]) => {
const results = await this.accountService.findByIds(keys);
return sortDataLoaderResultsByKey(keys, results, "id");
};
}
}
`

The first generic of the interface is the type of ID the datastore uses. The second generic is the type of object that will be returned. In the above instance, we want DataLoader to return instances of the Account class.

$3

For each NestDataLoader we create, we need to provide it to our module.

`typescript
import { Module } from '@nestjs/common';
import { APP_INTERCEPTOR } from '@nestjs/core';
import { DataLoaderInterceptor } from 'nestjs-dataloader'
...

@Module({
providers: [
AccountResolver,
AccountLoader,
{
provide: APP_INTERCEPTOR,
useClass: DataLoaderInterceptor,
},
],

})
export class ResolversModule { }
`

$3

Now that we have a dataloader and our module is aware of it, we need to pass it as a parameter to an endpoint in our graphQL resolver.

`typescript
import * as DataLoader from 'dataloader';
import { Loader } from 'nestjs-dataloader';
...

@Resolver(Account)
export class AccountResolver {

@Query(() => [Account])
public getAccounts(
@Args({ name: 'ids', type: () => [String] }) ids: string[],
@Loader(AccountLoader) accountLoader: DataLoader): Promise {
return accountLoader.loadMany(ids);
}
}
`

The important thing to note is that the parameter of the @Loader decorator is the entity/class of the NestDataLoader` we want to be injected to the method. The DataLoader library will handle bulk retrieval and caching of our requests. Note that the caching is stored on a per-request basis.

Contributing

Pull requests are always welcome. For major changes, please open an issue first to discuss what you would like to change.