Sourcing from Sanity
Sanity is a hosted backend for structured content that comes with an open source editor built in React. It has powerful real-time APIs for both reading and writing data.
You can use Sanity as a headless CMS that lets your authors work in a user friendly environment, or as a pure data backend for your apps. We make it easier for you to reuse content across multiple websites, apps, print, voice assistants, and other channels.
Begin with setting up a Gatsby project. If you want to start from scratch, the Quick Start guide is a good place to begin. Come back to this guide when you’re set up.
You can also check out the company website example we have set up. It contains both a configured Sanity Studio and a Gatsby frontend, which you can get up and running within minutes. It can be a useful reference for how to build a website using structured content. Follow the instructions in its README.md to get up and running.
This guide will cover how configure and use the
At this point you can choose to (and probably should) set up a GraphQL API for your Sanity dataset, if you have not done so already. This will help the plugin in knowing which types and fields exists, so you can query for them even without them being present in any current documents.
http://localhost:8000/___graphql after running
gatsby develop to understand the created data. Create a new query and check available collections and fields by using the autocomplete (
CTRL + SPACE).
|projectId||string||[required] Your Sanity project’s ID|
|dataset||string||[required] The dataset to fetch from|
|token||string||Authentication token for fetching data from private datasets, or when using |
|overlayDrafts||boolean||Set to |
|watchMode||boolean||Set to |
Getting errors such as these?
Cannot query field “allSanityBlogPost” Unknown field
By deploying a GraphQL API for your dataset, we are able to introspect and figure out which schema types and fields are available and make them available to prevent this problem. Once the API is deployed it will be transparently applied. If you have deployed your API and are still seeing similar issues, remember that you have to redeploy the API if your schema changes.
Some background for this problem:
Gatsby cannot know about the types and fields without having documents of the given types that contain the fields you want to query. This is a known problem with Gatsby - luckily there is ongoing work to solve this issue, which will lead to much clearer schemas and less boilerplate.
Image fields will have the image URL available under the
field.asset.url key, but you can also use gatsby-image for a smooth experience. It’s a React component that enables responsive images and advanced image loading techniques. It works great with this source plugin, without requiring any additional build steps.
There are two types of responsive images supported; fixed and fluid. To decide between the two, ask yourself: “do I know the exact size this image will be?” If yes, you’ll want to use fixed. If no and its width and/or height need to vary depending on the size of the screen, then you’ll want to use fluid.
These are the fragments available on image assets, which allows easy lookup of the fields required by gatsby-image in various modes:
Sometimes you might be working on some new content that is not yet published, which you want to make sure looks alright within your Gatsby site. By setting the
overlayDrafts setting to
true, the draft versions will as the option says “overlay” the regular document. In terms of Gatsby nodes, it will replace the published document with the draft.
Keep in mind that drafts do not have to conform to any validation rules, so your frontend will usually want to double-check all nested properties before attempting to use them.
While developing, it can often be beneficial to get updates without having to manually restart the build process. By setting
watchMode to true, this plugin will set up a listener which watches for changes. When it detects a change, the document in question is updated in real-time and will be reflected immediately.
If you add an environment token and set
overlayDrafts to true, each small change to the draft will immediately be applied.
Sanity does not have any concept of a “page”, since it’s built to be totally agnostic to how you want to present your content and in which medium, but since you’re using Gatsby, you’ll probably want some pages!
As with any Gatsby site, you’ll want to create a
gatsby-node.js in the root of your Gatsby site repository (if it doesn’t already exist), and declare a
createPages function. Within it, you’ll use GraphQL to query for the data you need to build the pages.
For instance, if you have a
project document type in Sanity that you want to generate pages for, you could do something along the lines of this:
The above query will fetch all projects that have a
slug.current field set, and generate pages for them, available as
/project/<project-slug>. It will use the template defined in
src/templates/project.js as the basis for these pages.
Most Gatsby starters have some example of building pages, which you should be able to modify to your needs.
Remember to use the GraphiQL interface to help write the queries you need - it’s usually running at
http://localhost:8000/___graphql while running
Arrays and object types at the root of documents will get an additional “raw JSON” representation in a field called
_raw<FieldName>. For instance, a field named
body will be mapped to
_rawBody. It’s important to note that this is only done for top-level nodes (documents).
Rich text in Sanity is usually represented as Portable Text (previously known as “Block Content”).
These data structures can be deep and a chore to query (specifying all the possible fields). As noted above, there is a “raw” alternative available for these fields which is usually what you’ll want to use.
You can install block-content-to-react from npm and use it in your Gatsby project to serialize Portable Text. It lets you use your own React components to override defaults and render custom content types. Learn more about Portable Text in our documentation.
If you don’t want to attach your Sanity project’s ID to the repo, you can easily store it in .env files by doing the following:
This example is based off Gatsby Docs’ implementation.