Gatsby

Created: Getting Started with Gatsby.js

Updated: 03 September 2023

Introduction

Gatsby.js is a Static Site Generator that makes use of React and can plug into a headless CMS to generate a Static Site with SPA support and functionality

I’m using the Gatsby.js Tutorial from the documentation

Prerequisites

  • Node.js
  • Git
  • Gatsby CLI
  • Optional: Yarn

To install the Gatsby CLI you can use npm:

1
npm i -g gatsby-cli

Or if using yarn:

1
yarn global add gatsby-cli

To view the Gatsby help menu once installed:

1
gatsby --help

Create New Site

To create a new Gatsby.js site run:

1
gatsby new gatsby-hello-world https://github.com/gatsbyjs/gatsby-starter-hello-world

Where gatsby-hello-world is the name of the new directory for the site and will clone from the Gatsby GtiHub template

Next cd gatsby-hello-world and run the following command to start development server you can use the gatsby-cli

1
gatsby develop

Or npm:

1
npm run develop

Or

1
yarn develop

You should then be able to launch the site on http://localhost:8000/

Looking at the initial site you should see the following files:

1
gatsby-hellp-world
2
│ .gitignore
3
│ .prettierignore
4
│ .prettierrc
5
│ gatsby-config.js
6
│ LICENSE
7
│ package.json
8
│ README.md
9
│ yarn.lock
10
11
├───src
12
│ └───pages
13
│ index.js
14
15
└───static
16
favicon.ico

In the index.js file you will see a simple React Component that is exported:

1
import React from 'react'
2
3
export default () => <div>Hello world!</div>

Editing this will live update the page as you edit and save the file, this uses HMR in the background and will update you browse live

Create a New Page

Gatsby organises pages similar to the way you would if you were using normal HTML instead. Inside of the pages directory you can create an about.js file with something like:

pages/about.js

1
import React from 'react'
2
3
export default () => <div>About Page</div>

And then we can add a link to this from the home component using the React Link component

pages/index.js

1
import React from 'react'
2
import { Link } from 'gatsby'
3
4
export default () => (
5
<div>
6
<h1>Hello World</h1>
7
<Link to="/about/">About</Link>
8
</div>
9
)

Clicking on the Link on the index.js page will take you to the about.js page

Build the Site

To build the initial site you can just run

1
gatsby build

Or

1
npm run build

Or

1
yarn build

You can then simply deploy the public directory using your preferred method

Adding Styles

To add styles we first need to create a src/styles/global.css file, this will contain all the global CSS for our application - we can add some basic content to it to start off

global.css

1
html,
2
body {
3
margin: 0;
4
padding: 0;
5
}
6
7
h1 {
8
color: lightblue;
9
}

Next in the src project root directory create a file called gatsby-browser.js, this is one of a few standard files that Gatsby uses. In this file import the global.css file we just created with:

gatsby-browser.js

1
import './src/styles/global.css'

After adding this file you will need to restart the Gatsby development server

Now let’s create a component called Container:

src/components/container.js

1
import React from 'react'
2
import containerStyles from './container.module.css'
3
4
export default ({ children }) => (
5
<div className={containerStyles.container}>{children}</div>
6
)

This file imports css file in the same directory called container.module.css which is a CSS Module which means that the styles will be scoped to this component. We also use containerStyles.container to apply the .container class to the main element

container.module.css

1
.container {
2
margin: 3rem auto;
3
max-width: 600px;
4
}

We can then update the index.js page to use this container:

index.js

1
import React from 'react'
2
import { Link } from 'gatsby'
3
import Container from '../components/container'
4
5
export default () => (
6
<Container>
7
<h1>Hello World</h1>
8
<Link to="/about/">About</Link>
9
</Container>
10
)

Plugins

Using plugins in Gatsby involves three steps:

  1. Installing the plugin. For example we can install the typography plugin with:
1
npm install --save gatsby-plugin-typography react-typography typography typography-theme-fairy-gates

Or

1
yarn add gatsby-plugin-typography react-typography typography typography-theme-fairy-gates
  1. Configuring the plugin which is done using the gatsby-config.js` file and a configuration file for the relevant plugin

gatsby-config.js

1
module.exports = {
2
plugins: [
3
{
4
resolve: `gatsby-plugin-typography`,
5
options: {
6
pathToConfigModule: `src/utils/typography`,
7
},
8
},
9
],
10
}

src/utils/typography.js

1
import Typography from 'typography'
2
import fairyGateTheme from 'typography-theme-fairy-gates'
3
4
const typography = new Typography(fairyGateTheme)
5
6
export const { scale, rhythm, options } = typography
7
export default typography

If you inspect the output HTML now after running gatsby develop you should see some styles in the head which are as a result of the typography plugin, the generated styles will look like:

1
<style id="typography.js">
2
...;
3
</style>

Data

The Gatsby Data Layer is a feature of Gatsby that enables you to build sites using a variety of CMSs

For the purpose of Gatsby, Data is anything that lives outside of a React component

Gatsby primarily makes use of GraphQL to load data into components however there are other data sources that can be used as well as custom plugins that can be used or custom written for this purpose

Common Site Metadata

The place for common site data, such as the site title is the gatsby-config.js file, we can put this in the siteMetadata object like so:

gatsby-config.js

1
module.exports = {
2
siteMetadata: {
3
title: 'Site Title from Metadata'
4
},
5
...
6
}

We can then query for the data by using the GraphQL query constant that we export on a Page Component which states the data required for the page itself

index.js

1
import React from 'react'
2
import { graphql } from 'gatsby'
3
import Container from '../components/container'
4
5
export default ({ data }) => (
6
<Container>
7
<h1>Title: {data.site.siteMetadata.title}</h1>
8
</Container>
9
)
10
11
export const query = graphql`
12
query {
13
site {
14
siteMetadata {
15
title
16
}
17
}
18
}
19
`

Other components can make use of the useStaticQuery hook, we can import it from gatsby

Let’s add the a simple static query for the title in the container component

container.js

1
import { useStaticQuery, Link, graphql } from 'gatsby'

We can then use this in our component

1
import React from 'react'
2
import { graphql } from 'gatsby'
3
import containerStyles from './container.module.css'
4
import { useStaticQuery } from 'gatsby'
5
6
export default ({ children }) => {
7
const data = useStaticQuery(
8
graphql`
9
query {
10
site {
11
siteMetadata {
12
title
13
}
14
}
15
}
16
`
17
)
18
19
return (
20
<div className={containerStyles.container}>
21
<p>{data.site.siteMetadata.title}</p>
22
{children}
23
</div>
24
)
25
}

Source Plugins

Source plugins are how we pull data into our site, Gatsby comes with a tool called GraphiQL which can be accessed at http://localhost:8000/___graphql when the development server is running

We can write a query to get the title using the GraphiQL UI:

1
query TitleQuery {
2
site {
3
siteMetadata {
4
title
5
}
6
}
7
}

Filesystem Plugin

We can access data from the File System using the gatsby-source-filesystem

1
yarn add gatsby-source-filesystem

And then in the gatsby-config.js file:

1
...
2
plugins: [
3
{
4
resolve: `gatsby-source-filesystem`,
5
options: {
6
name: `src`,
7
path: `${__dirname}/src/`,
8
},
9
},
10
...

If we restart the dev server we should see the allFile and file in the GraphiQL interface

We can then query for some data from the file system and log it to the console:

1
import React from 'react'
2
import { graphql } from 'gatsby'
3
import Container from '../components/container'
4
5
export default ({ data }) =>
6
console.log(data) || (
7
<Container>
8
<h1>Title: {data.site.siteMetadata.title}</h1>
9
</Container>
10
)
11
12
export const query = graphql`
13
query {
14
__typename
15
allFile {
16
edges {
17
node {
18
relativePath
19
prettySize
20
extension
21
birthTime(fromNow: true)
22
}
23
}
24
}
25
site(siteMetadata: { title: {} }) {
26
siteMetadata {
27
title
28
}
29
}
30
}
31
`

We can then build a simple table with the data:

1
export default ({ data }) =>
2
console.log(data) || (
3
<Container>
4
<h1>Title: {data.site.siteMetadata.title}</h1>
5
<table>
6
<thead>
7
<tr>
8
<th>relativePath</th>
9
<th>prettySize</th>
10
<th>extension</th>
11
<th>birthTime</th>
12
</tr>
13
</thead>
14
<tbody>
15
{data.allFile.edges.map(({ node }, index) => (
16
<tr key={index}>
17
<td>{node.relativePath}</td>
18
<td>{node.prettySize}</td>
19
<td>{node.extension}</td>
20
<td>{node.birthTime}</td>
21
</tr>
22
))}
23
</tbody>
24
</table>
25
</Container>
26
)

Transformers

Transformers are used by Gatsby to transform the data that is read in, we can use the following transformer to transform markdown

1
yarn add gatsby-transformer-remark

gatsby-config.js

1
...
2
plugins: [
3
'gatsby-transformer-remark',
4
...

We can then use the remark plugin combined with the GraphQL query to get markdown content from files in our application

1
query AllMarkdownQuery {
2
__typename
3
allMarkdownRemark {
4
edges {
5
node {
6
fileAbsolutePath
7
frontmatter {
8
title
9
date
10
}
11
excerpt
12
html
13
}
14
}
15
}
16
}

In the above query the result will be the rendered html node along with any metadata, for example in the file below:

src/pages/article-1.md

1
---
2
title: 'Sweet Pandas Eating Sweets'
3
date: '2017-08-10'
4
---
5
6
Pandas are really sweet.
7
8
Here's a video of a panda eating sweets.
9
10
<iframe width="560" height="315" src="https://www.youtube.com/embed/4n0xNbfJLR8" frameborder="0" allowfullscreen></iframe>

We can then use the query from above to create a page that lists all the markdown content we have in the site:

src/pages/blog.js

1
import React from 'react'
2
import { graphql } from 'gatsby'
3
import Container from '../components/container'
4
5
export default ({ data }) => {
6
console.log(data)
7
return (
8
<Container>
9
<div>
10
<h1>Amazing Pandas Eating Things</h1>
11
<h4>{data.allMarkdownRemark.totalCount} Posts</h4>
12
{data.allMarkdownRemark.edges.map(({ node }) => (
13
<div key={node.id}>
14
<h3>
15
{node.frontmatter.title} <span>{node.frontmatter.date}</span>
16
</h3>
17
<p>{node.excerpt}</p>
18
</div>
19
))}
20
</div>
21
</Container>
22
)
23
}
24
25
export const query = graphql`
26
query {
27
allMarkdownRemark {
28
totalCount
29
edges {
30
node {
31
id
32
frontmatter {
33
title
34
date(formatString: "DD MMMM, YYYY")
35
}
36
excerpt
37
}
38
}
39
}
40
}
41
`

Create Pages Programatically

Using Gatsby we can create pages using the data output from a query

Generate Page Slugs

We can make use of the onCreateNode and createPages API’s that Gatsby exposes. To implement an API we need to export the function in the gatsby-node.js file

The onCreateNode function is run every time a new node is created or updated

We can add the following into the gatsby-node.js file and can see each node that has been created

gatsby-node.js

1
exports.onCreateNode = ({ node }) => {
2
console.log(node.internal.type)
3
}

We can then check when a node is the MarkdownRemark and use the gatsby-source-filesystem plugin to generate a slug for the file

1
const { createFilePath } = require(`gatsby-source-filesystem`)
2
3
exports.onCreateNode = ({ node, getNode }) => {
4
if (node.internal.type === `MarkdownRemark`) {
5
slug = createFilePath({ node, getNode, basePath: `pages` })
6
console.log(slug)
7
}
8
}

Using the above, we can update a node with the createNodeField function which is part of the actions object that’s passed into the onCreateNode field

1
const { createFilePath } = require(`gatsby-source-filesystem`)
2
3
exports.onCreateNode = ({ node, getNode, actions }) => {
4
const { createNodeField } = actions
5
if (node.internal.type === `MarkdownRemark`) {
6
const slug = createFilePath({ node, getNode, basePath: `pages` })
7
createNodeField({
8
node,
9
name: `slug`,
10
value: slug,
11
})
12
}
13
}

We can then run the following query in the GraphiQL editor to see the slugs that were generated

1
query {
2
allMarkdownRemark {
3
edges {
4
node {
5
fields {
6
slug
7
}
8
}
9
}
10
}
11
}

Gatsby uses the createPages API from plugins to create pages, we can additionally export the createPages function from our gatsby-node.js file. To create a page programatically we need to:

  1. Query the data

gatsby-node.js

1
...
2
exports.createPages = async ({ graphql, actions }) => {
3
// **Note:** The graphql function call returns a Promise
4
// see: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Promise for more info
5
const result = await graphql(`
6
query {
7
allMarkdownRemark {
8
edges {
9
node {
10
fields {
11
slug
12
}
13
}
14
}
15
}
16
}
17
`)
18
console.log(JSON.stringify(result, null, 4))
19
}
  1. Map the query resilt to a page

We can first update the createPages function to set the slug route to resolve to a specific component, in this case the src/templates/blog-post.js

1
const path = require(`path`)
2
3
...
4
5
exports.createPages = async ({ graphql, actions }) => {
6
const { createPage } = actions
7
const result = await graphql(`
8
query {
9
allMarkdownRemark {
10
edges {
11
node {
12
fields {
13
slug
14
}
15
}
16
}
17
}
18
}
19
`)
20
result.data.allMarkdownRemark.edges.forEach(({ node }) => {
21
createPage({
22
path: node.fields.slug,
23
component: path.resolve(`./src/templates/blog-post.js`),
24
context: {
25
// Data passed to context is available
26
// in page queries as GraphQL variables.
27
slug: node.fields.slug,
28
},
29
})
30
})
31
}

We can then create the src/templates/blog-post.js file to render the new data:

1
import React from 'react'
2
import { graphql } from 'gatsby'
3
import Container from '../components/container'
4
5
export default ({ data }) => {
6
const post = data.markdownRemark
7
return (
8
<Container>
9
<div>
10
<h1>{post.frontmatter.title}</h1>
11
<div dangerouslySetInnerHTML={{ __html: post.html }} />
12
</div>
13
</Container>
14
)
15
}
16
17
export const query = graphql`
18
query ($slug: String!) {
19
markdownRemark(fields: { slug: { eq: $slug } }) {
20
html
21
frontmatter {
22
title
23
}
24
}
25
}
26
`

You should be able to view any created pages by navigating to a random route on your site which should open the development server’s 404 page which has a listing of the available pages

We can then also update the blog.js file to query for the slug and create a Link to the new page based on the slug

blog.js

1
import React from 'react'
2
import { graphql, Link } from 'gatsby'
3
import Container from '../components/container'
4
5
export default ({ data }) => {
6
console.log(data)
7
return (
8
<Container>
9
<div>
10
<h1>Amazing Pandas Eating Things</h1>
11
<h4>{data.allMarkdownRemark.totalCount} Posts</h4>
12
{data.allMarkdownRemark.edges.map(({ node }) => (
13
<div key={node.id}>
14
<h3>
15
{node.frontmatter.title} <span>{node.frontmatter.date}</span>
16
</h3>
17
<p>{node.excerpt}</p>
18
<Link to={node.fields.slug}>Read More</Link>
19
</div>
20
))}
21
</div>
22
</Container>
23
)
24
}
25
26
export const query = graphql`
27
query {
28
allMarkdownRemark {
29
totalCount
30
edges {
31
node {
32
id
33
frontmatter {
34
title
35
date(formatString: "DD MMMM, YYYY")
36
}
37
fields {
38
slug
39
}
40
excerpt
41
}
42
}
43
}
44
}
45
`