Access the complete project here. Requirements Node.js (used just for building the project with vite)
NPM (node package manager)
Rustup (Rust language toolchain)
Cargo (Rust package manager) Getting started For this example, we will use Vite.js to set up the project and compile the React source code. npm create vite@latest react-rust -- --template react-ts The initialized project is designed for client-side applications exclusively. In the following section, we will explore what is necessary to adapt it for full-stack bundling. React setup React requires two distinct builds tailored for different environments: The client-side build
The server-side build What distinguishes these two outputs? The client build incorporates all the hydration logic, enabling React to connect seamlessly with the HTML generated by the server. In contrast, the server build is a more streamlined version focused solely on rendering HTML based on the props received from the server. Now, let’s create a new file named ./src/server.tsx, which will serve as the entry point for the server build, and insert the following code: import "fast-text-encoding"; // Mandatory for React18
import { renderToString } from "react-dom/server";
import App from "./App";

export const Server = () => {
	const app = renderToString(<App />);
	return `<!doctype html>
<html>
    <head>
		<title>React + Rust = ❤️</title>
        <script type="module" crossorigin src="/index.js"></script>
    </head>
    <body>
		<div id="root">${app}</div>
    </body>
</html>`;
}; If you're working with React 18 or a newer version, it's essential to run npm install fast-text-encoding. This step is necessary because the Rust server lacks the Node.js objects and functions and the Web APIs. As a result, we need to provide a polyfill for TextEncoder, which is required by react-dom/server (in fact it is declared beforehand). We need to modify the vite.config.ts file to include the two outputs: import { defineConfig } from "vite";
import react from "@vitejs/plugin-react";

export default defineConfig({
  build: {
    rollupOptions: {
      output: {
        format: "iife",
        dir: "dist/",
      },
    },
  },
  ssr: {
    target: "webworker",
    noExternal: true,
  },
  plugins: [react()],
}); Next, we should add a new script in the package.json file. -- "build": "tsc && vite build",
++ "build": "tsc && vite build && vite build --ssr src/server.tsx", Prepare the Rust server For this example, we will use axum, a web framework that works on top of tokio. To get started, let's set up the Rust project by creating a Cargo.toml file in the main directory, containing the following details: [package]
name = "react-rust-ssr-example"
version = "0.1.0"
edition = "2021"

[[bin]]
name = "ssr"
path = "src/server/server.rs"

[dependencies]
ssr_rs="0.7.0"
tokio = { version = "1", features = ["full"] }
axum = "0.7.4"
tower-http = {version = "0.6.0", features = ["fs"]} This is the Rust manifest - pretty similar to the JavaScript package.json file. Next, we’ll set up a file named src/server/server.rs, which will serve as the entry point for launching the Rust server. use axum::{response::Html, routing::get, Router};
use ssr_rs::Ssr;
use std::cell::RefCell;
use std::fs::read_to_string;
use std::path::Path;
use tower_http::services::ServeDir;

thread_local! {
    static SSR: RefCell<Ssr<'static, 'static>> = RefCell::new(
            Ssr::from(
                read_to_string(Path::new("./dist/server.js").to_str().unwrap()).unwrap(),
                ""
                ).unwrap()
            )
}

#[tokio::main]
async fn main() {
    Ssr::create_platform();
    // build our application with a single route
    let app = Router::new()
        .route("/", get(root))
        .fallback_service(ServeDir::new("dist"));

    // run our app with hyper, listening globally on port 3000
    let listener = tokio::net::TcpListener::bind("0.0.0.0:3000").await.unwrap();
    axum::serve(listener, app).await.unwrap();
}

async fn root() -> Html<String> {
    let result = SSR.with(|ssr| ssr.borrow_mut().render_to_string(None));
    Html(result.unwrap())
} This is where the magic unfolds. At the start of the program, we kick things off by initializing the JavaScript V8 engine with Ssr::create_platform(); Next, we create a V8 context in each thread using thread_local!. Finally, we render the HTML with SSR.with(|ssr| ssr.borrow_mut().render_to_string(None)); and send it to the client when the route http://localhost:3000/ is requested. Run the server To start the server, simply compile the assets with Vite and then launch the Rust server. npm run build && cargo run 🎊 You are running a full-stack React application using a Rust server. Finally, React runs within a multithreaded server (you can find some benchmarks here). Final words Managing a full-stack React application is not easy with Node.js that plenty of tools have been built overtime to support it, and as you could see with Rust it is even harder. Tuono is an experimental full-stack framework aimed at simplifying the development of high-performance Rust applications, with a strong focus on both usability and speed. Access the complete project here. Access the complete project here . here Requirements Node.js (used just for building the project with vite) NPM (node package manager) Rustup (Rust language toolchain) Cargo (Rust package manager) Node.js (used just for building the project with vite) NPM (node package manager) Rustup (Rust language toolchain) Cargo (Rust package manager) Getting started For this example, we will use Vite.js to set up the project and compile the React source code. npm create vite@latest react-rust -- --template react-ts npm create vite@latest react-rust -- --template react-ts The initialized project is designed for client-side applications exclusively. In the following section, we will explore what is necessary to adapt it for full-stack bundling. React setup React requires two distinct builds tailored for different environments: The client-side build The server-side build The client-side build The server-side build What distinguishes these two outputs? The client build incorporates all the hydration logic, enabling React to connect seamlessly with the HTML generated by the server. In contrast, the server build is a more streamlined version focused solely on rendering HTML based on the props received from the server. Now, let’s create a new file named ./src/server.tsx , which will serve as the entry point for the server build, and insert the following code: ./src/server.tsx import "fast-text-encoding"; // Mandatory for React18
import { renderToString } from "react-dom/server";
import App from "./App";

export const Server = () => {
	const app = renderToString(<App />);
	return `<!doctype html>
<html>
    <head>
		<title>React + Rust = ❤️</title>
        <script type="module" crossorigin src="/index.js"></script>
    </head>
    <body>
		<div id="root">${app}</div>
    </body>
</html>`;
}; import "fast-text-encoding"; // Mandatory for React18
import { renderToString } from "react-dom/server";
import App from "./App";

export const Server = () => {
	const app = renderToString(<App />);
	return `<!doctype html>
<html>
    <head>
		<title>React + Rust = ❤️</title>
        <script type="module" crossorigin src="/index.js"></script>
    </head>
    <body>
		<div id="root">${app}</div>
    </body>
</html>`;
}; If you're working with React 18 or a newer version, it's essential to run npm install fast-text-encoding. This step is necessary because the Rust server lacks the Node.js objects and functions and the Web APIs. As a result, we need to provide a polyfill for TextEncoder, which is required by react-dom/server (in fact it is declared beforehand). If you're working with React 18 or a newer version, it's essential to run npm install fast-text-encoding . This step is necessary because the Rust server lacks the Node.js objects and functions and the Web APIs. As a result, we need to provide a polyfill for TextEncoder , which is required by react-dom/server (in fact it is declared beforehand). npm install fast-text-encoding TextEncoder react-dom/server We need to modify the vite.config.ts file to include the two outputs: vite.config.ts import { defineConfig } from "vite";
import react from "@vitejs/plugin-react";

export default defineConfig({
  build: {
    rollupOptions: {
      output: {
        format: "iife",
        dir: "dist/",
      },
    },
  },
  ssr: {
    target: "webworker",
    noExternal: true,
  },
  plugins: [react()],
}); import { defineConfig } from "vite";
import react from "@vitejs/plugin-react";

export default defineConfig({
  build: {
    rollupOptions: {
      output: {
        format: "iife",
        dir: "dist/",
      },
    },
  },
  ssr: {
    target: "webworker",
    noExternal: true,
  },
  plugins: [react()],
}); Next, we should add a new script in the package.json file. package.json -- "build": "tsc && vite build",
++ "build": "tsc && vite build && vite build --ssr src/server.tsx", -- "build": "tsc && vite build",
++ "build": "tsc && vite build && vite build --ssr src/server.tsx", Prepare the Rust server For this example, we will use axum , a web framework that works on top of tokio . axum tokio To get started, let's set up the Rust project by creating a Cargo.toml file in the main directory, containing the following details: Cargo.toml [package]
name = "react-rust-ssr-example"
version = "0.1.0"
edition = "2021"

[[bin]]
name = "ssr"
path = "src/server/server.rs"

[dependencies]
ssr_rs="0.7.0"
tokio = { version = "1", features = ["full"] }
axum = "0.7.4"
tower-http = {version = "0.6.0", features = ["fs"]} [package]
name = "react-rust-ssr-example"
version = "0.1.0"
edition = "2021"

[[bin]]
name = "ssr"
path = "src/server/server.rs"

[dependencies]
ssr_rs="0.7.0"
tokio = { version = "1", features = ["full"] }
axum = "0.7.4"
tower-http = {version = "0.6.0", features = ["fs"]} This is the Rust manifest - pretty similar to the JavaScript package.json file. package.json Next, we’ll set up a file named src/server/server.rs , which will serve as the entry point for launching the Rust server. src/server/server.rs use axum::{response::Html, routing::get, Router};
use ssr_rs::Ssr;
use std::cell::RefCell;
use std::fs::read_to_string;
use std::path::Path;
use tower_http::services::ServeDir;

thread_local! {
    static SSR: RefCell<Ssr<'static, 'static>> = RefCell::new(
            Ssr::from(
                read_to_string(Path::new("./dist/server.js").to_str().unwrap()).unwrap(),
                ""
                ).unwrap()
            )
}

#[tokio::main]
async fn main() {
    Ssr::create_platform();
    // build our application with a single route
    let app = Router::new()
        .route("/", get(root))
        .fallback_service(ServeDir::new("dist"));

    // run our app with hyper, listening globally on port 3000
    let listener = tokio::net::TcpListener::bind("0.0.0.0:3000").await.unwrap();
    axum::serve(listener, app).await.unwrap();
}

async fn root() -> Html<String> {
    let result = SSR.with(|ssr| ssr.borrow_mut().render_to_string(None));
    Html(result.unwrap())
} use axum::{response::Html, routing::get, Router};
use ssr_rs::Ssr;
use std::cell::RefCell;
use std::fs::read_to_string;
use std::path::Path;
use tower_http::services::ServeDir;

thread_local! {
    static SSR: RefCell<Ssr<'static, 'static>> = RefCell::new(
            Ssr::from(
                read_to_string(Path::new("./dist/server.js").to_str().unwrap()).unwrap(),
                ""
                ).unwrap()
            )
}

#[tokio::main]
async fn main() {
    Ssr::create_platform();
    // build our application with a single route
    let app = Router::new()
        .route("/", get(root))
        .fallback_service(ServeDir::new("dist"));

    // run our app with hyper, listening globally on port 3000
    let listener = tokio::net::TcpListener::bind("0.0.0.0:3000").await.unwrap();
    axum::serve(listener, app).await.unwrap();
}

async fn root() -> Html<String> {
    let result = SSR.with(|ssr| ssr.borrow_mut().render_to_string(None));
    Html(result.unwrap())
} This is where the magic unfolds. At the start of the program, we kick things off by initializing the JavaScript V8 engine with Ssr::create_platform(); Next, we create a V8 context in each thread using thread_local! . Finally, we render the HTML with SSR.with(|ssr| ssr.borrow_mut().render_to_string(None)); and send it to the client when the route http://localhost:3000/ is requested. Ssr::create_platform(); thread_local! SSR.with(|ssr| ssr.borrow_mut().render_to_string(None)); http://localhost:3000/ Run the server To start the server, simply compile the assets with Vite and then launch the Rust server. npm run build && cargo run npm run build && cargo run 🎊 You are running a full-stack React application using a Rust server. Finally, React runs within a multithreaded server (you can find some benchmarks here ). here Final words Managing a full-stack React application is not easy with Node.js that plenty of tools have been built overtime to support it, and as you could see with Rust it is even harder. Tuono is an experimental full-stack framework aimed at simplifying the development of high-performance Rust applications, with a strong focus on both usability and speed. Tuono

The code in this story is for educational purposes. The readers are solely responsible for whatever they build with it.

New Template Offers Minimal Setup to Get React Working

About Author

Comments

TOPICS

THIS ARTICLE WAS FEATURED IN

Related Stories

Untitled Story

30 Next.js Interview Questions: Get Ready for Your Dream Job

5 reasons to consider Next.js for your next projects

5 Reasons to Convert from React to Next.js

50+ GitHub Repositories Every Front-End Developer Needs

6 Easy Steps to Use Next Auth in Next.js 13 Using Route Handler

30 Next.js Interview Questions: Get Ready for Your Dream Job

5 reasons to consider Next.js for your next projects

5 Reasons to Convert from React to Next.js

50+ GitHub Repositories Every Front-End Developer Needs

6 Easy Steps to Use Next Auth in Next.js 13 Using Route Handler

Light-Mode

Classic

Newspaper

Dark-Mode

Neon Noir

Minty

HN StartUps