Next.js cache-effective pages

What it does

Let's say you want to re-generate a static file (e.g. public/sitemap.xml) every 15 minutes. The first solution that comes to mind is doing this at build time and it's great and simple, but.... it wouldn't work for mid and big-scale applications (considering that you're rebuilding your app every time there's a change in CMS). And this is where next-cache-effective-pages comes into the picture.
It makes it easier to change your static file into a regeneratable page without you worrying about effective caching and bandwidth attacks.

Features

• 🙉 Effective caching
• 🚚 Bandwidth attack proofed
• 🤠 Simple and flexible API
• 🐄 No dependencies

Installation

$ npm i --save next-cache-effective-pages

# or

$ yarn add next-cache-effective-pages

Example use cases

Sitemap

// pages/sitemap.xml.js

export default function Sitemap() {}

export async function getServerSideProps(ctxt) {
  return withCacheEffectivePage(async ({ res }) => {
    res.setHeader('Content-Type', 'text/xml')
    res.write(await getAllPosts())
    res.end()
  })({ ...ctxt, options: { secondsBeforeRevalidation: 60 * 15 } }) // Re-generate the page every 15 minutes
}

Sitemap with pagination

// pages/sitemap.xml.js

export default function Sitemap() {}

export async function getServerSideProps(ctxt) {
  return withCacheEffectivePage(async ({ res, query }) => {
    const maxPages = await getMaxPages()

    if (query.page > maxPages) {
      // redirect to last
    }

    res.setHeader('Content-Type', 'text/xml')
    res.write(await getPostsByPage(query.page))
    res.end()
  })({ ...ctxt, options: { secondsBeforeRevalidation: 60 * 15, allowedQueryParams: ['page'] } }) // You can whitelist a query parameter
}

Options

{
  secondsBeforeRevalidation?: number; # Self-descriptive
  allowedQueryParams?: string[]; # These won't be removed from the url while redirecting
}

FAQ

How does it prevent bandwidth attacks?

The easiest way to attack an app's bandwidth quota is by adding the current timestamp to a request, like so:

$ curl -s -I -X GET "https://bstefanski.com/sitemap.xml?$(date +%s)"

If your site is server-side rendered it will probably miss the cached entry and create a new one. This library prevents from returning an uncached big chunk of data by redirecting to a query-less url (https://bstefanski.com/sitemap.xml?43534543=0 -> https://bstefanski.com/sitemap.xml)

How are you caching this?

By setting Cache-Control header to s-maxage=${secondsBeforeRevalidation}, stale-while-revalidate.

stale-while-revalidate - Indicates the client will accept a stale response, while asynchronously checking in the background for a fresh one.