Edge Computing with Cloudflare Workers & Deno · レッスン

ルーティングとミドルウェア

ルーティングライブラリとミドルウェアパターンを活用して、Worker APIを整理・拡張します

「ルーティングとミドルウェア」はCoddyKit上の無料Edge Computing with Cloudflare Workers & Denoレッスンです。これはレッスン2/3です。下記で完全なレッスンを無料で読むことができます。その後、ブラウザ内の組み込みコードエディタと24時間対応のAIチューターでハンズオン演習できます。これはEdge Computing with Cloudflare Workers & Deno学習パスの一部であり、ウェブとCoddyKitアプリ全体で進捗が同期されます。 Edge Computing with Cloudflare Workers & Denoコースには全3レッスンが含まれています。

このレッスンの一部はまだ翻訳されておらず、英語で表示されています。

API Routing Essentials

When building APIs, we need a way to direct incoming requests to the correct functions based on their URL path and HTTP method (GET, POST, etc.). This process is called routing.

Think of it like a receptionist for your API. When a request comes in, the router checks its destination and sends it to the right department (your code handler).

Manual Worker Routing

In a Cloudflare Worker, all incoming requests are handled by the fetch event listener. You can manually inspect the request's URL and method to decide what to do.

While possible for simple cases, this approach quickly becomes complex and hard to manage as your API grows.

export default {
  async fetch(request) {
    const url = new URL(request.url);
    if (url.pathname === '/hello' && request.method === 'GET') {
      return new Response('Hello from manual route!');
    }
    return new Response('Not Found', { status: 404 });
  }
}

Introducing itty-router

To simplify routing, we use lightweight libraries. For Cloudflare Workers, itty-router is a popular choice. It's tiny, fast, and designed for the edge.

It provides a clean, expressive way to define routes and handlers, making your Worker code much more organized.

Basic GET Routes

With itty-router, you can define routes using methods like router.get() for GET requests. Each route takes a path and a handler function.

Let's create a simple Worker that responds to / and /greet.

import { Router } from 'itty-router';

const router = Router();

router.get('/', () => new Response('Welcome to the API!'));
router.get('/greet', () => new Response('Hello there!'));

export default {
  fetch: router.handle
};

Handling All HTTP Methods

itty-router allows you to define routes for specific HTTP methods like .post(), .put(), .delete(), and more.

You can also use .all() to match any HTTP method for a given path, useful for middleware or generic handlers.

import { Router } from 'itty-router';

const router = Router();

router.get('/data', () => new Response('GET data'));
router.post('/data', () => new Response('POST data', { status: 201 }));
router.all('*', () => new Response('Method Not Allowed', { status: 405 }));

export default {
  fetch: router.handle
};

Dynamic Route Parameters

Often, you need to extract dynamic values from the URL, like an item ID or a username. These are called route parameters.

itty-router uses a colon (:) to define parameters in a route path. The values are then available in the handler's request.params object.

import { Router } from 'itty-router';

const router = Router();

router.get('/users/:id', ({ params }) => {
  return new Response(`Fetching user ${params.id}`);
});

export default {
  fetch: router.handle
};

What is Middleware?

Middleware functions are code snippets that run before or after your main route handler. They can modify the request, perform logging, check authentication, or add headers to the response.

They act as a pipeline, allowing you to add common functionalities across multiple routes without duplicating code.

Implementing Simple Middleware

With itty-router, middleware can be added to specific routes or globally. A middleware function receives the request and can return a response (ending the chain) or continue to the next handler.

Let's add a simple logging middleware that runs for every request.

import { Router } from 'itty-router';

const router = Router();

const loggerMiddleware = async (request, event) => {
  console.log(`Request: ${request.method} ${request.url}`);
  // To proceed to the next handler, don't return a Response.
  // If you return a Response, the chain stops.
};

router.all('*', loggerMiddleware);
router.get('/hello', () => new Response('Hello from route!'));

export default {
  fetch: router.handle
};

Chaining Middleware

You can chain multiple middleware functions. Each middleware executes in order. If a middleware doesn't return a response, the next one in the chain (or the final route handler) is called.

This allows for powerful, modular processing of requests, like authentication, data parsing, and logging.

import { Router } from 'itty-router';

const router = Router();

const authMiddleware = async (request) => {
  if (request.headers.get('Authorization') !== 'Bearer token123') {
    return new Response('Unauthorized', { status: 401 });
  }
};

const headerMiddleware = async (request, event) => {
  // Add a custom header to the response later
  event.response = new Response('', { headers: { 'X-Powered-By': 'CoddyKit' } });
};

router.get('/secure', authMiddleware, (request) => {
  return new Response('Access granted!');
});
router.get('/info', headerMiddleware, (request) => {
  return new Response('Info page');
});

export default {
  fetch: async (request, env, ctx) => {
    const response = await router.handle(request, env, ctx);
    // Merge headers from middleware if present
    if (ctx.response && response) {
      for (let [key, value] of ctx.response.headers.entries()) {
        response.headers.set(key, value);
      }
    }
    return response;
  }
};

Routing & Middleware Check

Consider a Cloudflare Worker using itty-router. You want to ensure that all requests to /admin/* paths require an X-Admin-Key header, and if it's missing, return a 403 Forbidden response. Other paths should not be affected.

Recap: Routing & Middleware

You've mastered the fundamentals of organizing your edge API with routing and middleware!

Routing directs requests to the right handlers.
Libraries like itty-router simplify route definition, including dynamic parameters.
Middleware functions enhance requests/responses, adding common logic like logging or authentication before (or after) the main handler.

These patterns are crucial for building scalable and maintainable serverless APIs at the edge.

よくある質問

「ルーティングとミドルウェア」レッスンは無料ですか？

はい。「ルーティングとミドルウェア」の完全なテキストはこのウェブで無料で読めます。インタラクティブに演習し（組み込みコードエディタと24時間対応のAIチューター）、Edge Computing with Cloudflare Workers & Denoコースの残りをアンロックするには、CoddyKit PROにアップグレードしてください。 Edge Computing with Cloudflare Workers & Denoコースには全3レッスンが含まれています。

「ルーティングとミドルウェア」で何を学びますか？

ルーティングライブラリとミドルウェアパターンを活用して、Worker APIを整理・拡張しますブラウザで直接実行するハンズオンコードでEdge Computing with Cloudflare Workers & Denoを演習し、24時間対応のAIチューターがレッスンを進める中での質問に答えます。

Edge Computing with Cloudflare Workers & Denoを始めるのに経験は必要ですか？

事前経験は必要ありません。CoddyKitのEdge Computing with Cloudflare Workers & Denoは初級者から上級者向けに構成されているため、ここから始めるか最初から始めて、自分のペースで進むことができます。これはレッスン2/3です。

「ルーティングとミドルウェア」レッスンにはどのくらい時間がかかりますか？

ほとんどのCoddyKitレッスンは約5～10分かかります。各レッスンはコンパクトでインタラクティブなので、着実に進歩し、ウェブとアプリ全体で正確に前回の場所から再開できます。

このEdge Computing with Cloudflare Workers & Denoレッスンでコードを書いて実行できますか？

はい。すべてのEdge Computing with Cloudflare Workers & Denoレッスンに組み込みコードエディタが含まれているため、ブラウザでリアルコードを書いて実行し、即座のAIフィードバックを取得できます。ローカル設定は不要です。

このコースのすべてのレッスン

RESTful APIの設計
ルーティングとミドルウェア
バリデーションとエラー処理

← Edge Computing with Cloudflare Workers & Denoに戻る