Tag: nextjs

  • Protected Routes in Next.js

    Protected Routes in Next.js

    If you are building a SaaS website that has awesome features or a simple website with minimal user functionality, you know Authentication and Authorization are crucial (difference between authentication and authorization). Protected Routes in Next.js help us ensure that unauthenticated users are not able to see routes/pages intended for logged in (authenticated) users. There are a few approaches to to implement Protected Routes in Next.js, i.e., enforce authentication for a page/route.

    But, first of all – why do we love Next.js? Next.js is arguably the most popular and go-to React framework. It packs some cool stuff including file-based routing, incremental static regeneration, and internationalization (i18n). With Next.js 13, we have got even more power – layouts and Turbopack!

    You might be wondering – why bother protecting routes? We are building a SaaS product with a Next.js frontend and Nest.js backend. We have implemented authentication in the backend but we also need to ensure that forced browsing* is prevented and User Experience is enriched. Actual authentication logic should reside inside our back-end logic. All the API calls must be appropriately authenticated. In our app, whenever there is an unauthenticated request it returns 401 Unauthorized. An ACL is also in place so whenever user requests a resource they do not have access to, the backend returns 403 Forbidden.

    Now, let’s create a route protection flow in Next.js.:
    If a user requests a protected route (something that requires authentication), we redirect them to the login page.
    We should not prevent access if a route is public (supposed to be viewed regardless of the users’ authentication state) like a login page.

    At the end, the goals are simple: safety and security.

    Jodi Rell

    Using RouteGuard

    The concept of a RouteGuard is simple. It is a wrapper component that checks whether the user has access to the requested page on every route change. To track the access, we use one states: authorized. If authorized is true, then the user may see the page or else user is redirected to the login page. To update the state, we have a function authCheck() which prevents access (sets authorized to false) if the user does not have access and the page is not public (e.g. landing page, login page, sign-up page).

    import { Flex, Spinner } from '@chakra-ui/react';
    import { useRouter } from 'next/router';
    import publicPaths from '../data/publicPaths';
    import { useAppDispatch, useAppSelector } from '../hooks/storeHooks';
    import { setRedirectLink } from '../redux/AuthSlice';
    import {
      JSXElementConstructor,
      ReactElement,
      useEffect,
      useState,
    } from 'react';
    
    const RouteGuard = (props: {
      children: ReactElement<unknown, string | JSXElementConstructor<unknown>>;
    }) => {
      const { children } = props;
    
      const router = useRouter();
      const [authorized, setAuthorized] = useState(false);
      const user = useAppSelector((state) => state.auth);
    
      const dispatch = useAppDispatch();
    
      useEffect(() => {
        const authCheck = () => {
          if (
            !user.isLoggedIn &&
            !publicPaths.includes(router.asPath.split('?')[0])
          ) {
            setAuthorized(false);
            dispatch(setRedirectLink({ goto: router.asPath }));
            void router.push({
              pathname: '/login',
            });
          } else {
            setAuthorized(true);
          }
        };
    
        authCheck();
    
        const preventAccess = () => setAuthorized(false);
    
        router.events.on('routeChangeStart', preventAccess);
        router.events.on('routeChangeComplete', authCheck);
    
        return () => {
          router.events.off('routeChangeStart', preventAccess);
          router.events.off('routeChangeComplete', authCheck);
        };
      }, [dispatch, router, router.events, user]);
    
      return authorized ? (
        children
      ) : (
        <Flex h="100vh" w="100vw" justifyContent="center" alignItems="center">
          <Spinner size="xl" />
        </Flex>
      );
    };
    
    export default RouteGuard;

    Note: we are using Redux to store the user’s data; authentication is out of the scope of this blog post.

    Implementing the Middleware

    In a scenario where the users’ session expires while they are on a protected page, they will not be able to fetch newer resources (or perform any actions for that matter). That’s, once again, really bad UX. We cannot expect a user to refresh, so we need a way to let them know that their session is no longer valid.

    To implement the same, we will use another awesome Next.js feature – Middlewares! In few words, a middleware sits between your server and the frontend. Middleware allows you to run code before a request is completed, then based on the incoming request, you can modify the response by rewriting, redirecting, modifying the request or response headers, or responding directly.

    After session expiration, whenever the user makes a request, it will result in 401 Unauthorized. We have implemented a middleware which listens to the response for each request that is being made from the frontend; if the request results in 401 Unauthorized, we dispatch the same action, i.e. log out the user and redirect to the login page.

    import {
      MiddlewareAPI,
      isRejectedWithValue,
      Middleware,
    } from '@reduxjs/toolkit';
    import { logout } from '../redux/AuthSlice';
    import { store } from '../redux/store';
    
    interface ActionType {
      type: string;
      payload: { status: number };
      meta: {};
      error: {};
    }
    
    const unauthenticatedInterceptor: Middleware =
      (_api: MiddlewareAPI) =>
      (next: (action: ActionType) => unknown) =>
      (action: ActionType) => {
        if (isRejectedWithValue(action)) {
          if (action.payload.status === 401 || action.payload.status === 403) {
            console.error('MIDDLEWARE: Unauthorized/Unauthenticated [Invalid token]');
            store.dispatch(logout());
          }
        }
    
        return next(action);
      };
    
    export default unauthenticatedInterceptor;

    Suggested Readings

  • Getting Started with DatoCMS – Creating a simple blog

    Getting Started with DatoCMS – Creating a simple blog

    DatoCMS is a relative newcomer in the CMS industry. There are plenty of Content Management Systems (CMS). A CMS is a powerful piece of software that helps users to create, manage, and modify content on a website without requiring special technical expertise and without the need to code. WordPress is the most popular CMS with over 43% of the market share. Some other examples of CMS are Drupal, Joomla.

    DatoCMS rolled out for public use in 2019, is a cloud-based headless CMS for mobile apps, static websites, and server-side applications. The concept of headless CMS is relatively newer. It was born out of the modern need to serve content across multiple channels, like web apps, mobile apps, IoT devices, and wearable software. A headless CMS attempts to reduce the modules attached to the software since there is no presentation layer attached; the developer has the flexibility of serving the content across a wide range of channels.

    Why DatoCMS?

    DatoCMS is API-first. Every software powered by DatoCMS makes use of two APIs to work with the content – Content Delivery API, and Content Management API. To perform operations on the content programmatically, we make use of the Content Management API. To retrieve the content for displaying purposes, we make use of the Content Delivery API.

    DatoCMS generates a static website, that results in a faster and more secure website. It fits in very well with frameworks, technologies and generators like React, Next.js, Jekyll, Nuxt, Vue, PHP, Middleman, and Ruby on Rails. It is flexible, offers granular permissions, and comes with a GraphQL API.

    DatoCMS does not build and deploy your website, but delegates it to an external CI/CD services. DatoCMS offers integration with the following services out of the box:

    To read more about the general concepts of DatoCMS, please refer the official docs.

    Creating a DatoCMS Account & Starting a New Project

    Well, signing up for any service should be a piece of cake, really! Just head to https://dashboard.datocms.com/signup and you know the drill!

    Once you sign up, you will be guided to the dashboard. DatoCMS is free for upto 3 projects and that should suit our use-case. DatoCMS Dashboard is easy to navigate, considering that the target audience is non-technical clients. We will start by hitting the ‘New Project’ button.

    Click on ‘Demo project’ and you will be lead to the page of starter projects. We will be using Next.js Blog starter project.

    Now, click on ‘Start free project’. Name your project and choosing a hosting solution. We will use Vercel to host your app.

    Configuring Vercel for DatoCMS Blog

    Clicking on ‘Click project’ should open a new tab for Vercel configuration. First, we need to create a GitHub repository to enable CI/CD. I will create a private repo but it is absolutely all right to keep your repo public.

    Now, we must integrate DatoCMS with the Vercel app (for obvious reasons).

    The next and the final step is to deploy our Git repo to Vercel. It should happen automatically and you should see a nice-looking message on your DatoCMS Dashboard.

    Editing the DatoCMS Blog

    We have successfully deployed our blog. Now, we will modify global settings like Blog Description, SEO, and Favicon from the Dato Editor. Just click on your project in the dashboard and hit ‘Enter Project’. Once you are inside the editor, you will see a lot of settings in the left sidebar. Those are all the customizable options and essentially the content we can edit without having to tweak the codebase. To modify the favicon and SEO settings, we will click on ‘Settings’ option.

    Now, let’s configure the homepage settings like Blog Title using the ‘Homepage’ option from the left sidebar.

    Exploring DatoCMS Blog Models

    DatoCMS handles editable data as a scheme called ‘Models’, which you can think of as Tables. Each model has some set of user-editable fields, like Post Title, Post Content, Featured Image. Models are modular and provide a user-friendly editing experience.

    To see our models, we will click on ‘Settings’ in the top bar. Once we are there, we will click on ‘Models’ in the left sidebar. Our starter project has four pre-defined models – Author, Blog, Category, and Post. Each model can have a collection of records/instances, like we can have multiple posts on a single website.

    Our Post model has a number of fields – it has a Title, Author (which is linked to another model ‘Author’), and a structured text field to store the post Content. Then, it has some fieldsets, which are models inside models (sort of). Those handle our previews and post metadata.

    To know more about DatoCMS Models, you can check out DatoCMS’ documentation.

    Adding Content

    Now that we have an overview of what a model is, we will add a few posts. To add content (posts) to our blog, we will click on ‘Content’ in the top menu. Inside the left sidebar, we will have all our models. Before we do that, let’s add an author so that we can link it with our posts.

    Let’s create a post. I will quickly copy my own post about how you can setup Grafana and Loki to monitor your server logs ?. DatoCMS Editor supports rich-text, and we can add things like lists, images, headings, and quotes by pressing the '/' key (just like WordPress). We will add stuff like quotes, and code blocks (DatoCMS Editor has some awesome codeblock editing experience).

    Similarly, we can add the other two fieldsets – Preview and Metadata. Now, we are ready to publish!

    Building the Blog

    Once you are done with all the editing, you need to deploy the blog by ‘building’. In the top-right corner, click on ‘Build Status’ and hit ‘Build Now’. This will generate a static site and deploy it on Vercel. The logs can be checked in Settings > Activity Logs.

    Bravo! Our website is live now ?. Click on ‘Visit Site’ in the ‘Build Status’ dropdown to see your site. You can check mine here.

    What’s Next?

    In the next blog posts, I plan to go through how we can test our DatoCMS app/website locally, edit the code, and make it more customizable. Also, maybe we can check out how it compares to WordPress ?!?

    I would like to thank the contributors of the DatoCMS Blog Starter Project. Do check out the repo.