Hello Rust community,

I'm pleased to announce the first public release of async-graphql-dataloader , a DataLoader pattern implementation designed for native integration with the async-graphql ecosystem.

Goal: Efficiently solve the N+1 problem in Rust GraphQL servers by providing automatic request batching and request-scoped caching.

Key Features:

• Automatic Batching: Multiple .load() calls are automatically coalesced into a single batch.
• Intelligent Caching: Request-scoped caching using DashMap , preventing duplicate loads.
• Seamless Integration: Works out-of-the-box with async-graphql via context injection.
• Flexible: The Loader trait can be implemented for any data source (SQL DBs, HTTP APIs, etc.).
• Async Native: Built on tokio and async/await .
• Example usage in a resolver:
  use async_graphql_dataloader::{DataLoader, Loader};
  use std::collections::HashMap;

struct UserLoader;
#[async_trait::async_trait]
impl Loader for UserLoader {
type Value = String;
type Error = std::convert::Infallible;

async fn load(&self, keys: &[i32]) -> Result<HashMap<i32, Self::Value>, Self::Error> {
    Ok(keys.iter().map(|&k| (k, format!("User {}", k))).collect())
}

}

// In your GraphQL resolver
async fn get_users(ctx: &async_graphql::Context<'_>, ids: Vec) -> async_graphql::Result<Vec> {
let loader = ctx.data_unchecked::<DataLoader>();
let futures = ids.into_iter().map(|id| loader.load_one(id));
let users = futures::future::join_all(futures).await;
users.into_iter().collect()
}
The crate is stable for use, and community feedback is highly welcomed, especially on:

• The API and Loader trait design.
• Advanced use cases or edge conditions.
• Ideas for future features.

Links:

• Crates.io: async-graphql-dataloader
• Documentation: docs.rs/async-graphql-dataloader
• Source Code & Issues: GitHub Repository

Thank you for your interest. I look forward to constructive discussions and contributions.

Cleiton Augusto Correa Bezerra

1 post - 1 participant

Read full topic