npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2024 – Pkg Stats / Ryan Hefner

@bradleygriffith/gatsby-source-sanity

v6.0.4-patch2

Published

Gatsby source plugin for building websites using Sanity.io as a backend.

Downloads

3

Readme

gatsby-source-sanity

Source plugin for pulling data from Sanity.io into Gatsby websites. Develop with real-time preview of all content changes. Compatible with gatsby-image.

Get up and running in minutes with a fully configured starter project:

Watch a video about the company website built with Gatsby using Sanity.io as a headless CMS

Table of contents

See the getting started video

Install

From the command line, use npm (node package manager) to install the plugin:

npm install gatsby-source-sanity

In the gatsby-config.js file in the Gatsby project's root directory, add the plugin configuration inside of the plugins section:

module.exports = {
  // ...
  plugins: [
    {
      resolve: `gatsby-source-sanity`,
      options: {
        projectId: `abc123`,
        dataset: `blog`,
        // a token with read permissions is required
        // if you have a private dataset
        token: process.env.SANITY_TOKEN,

        // If the Sanity GraphQL API was deployed using `--tag <name>`,
        // use `graphqlTag` to specify the tag name. Defaults to `default`.
        graphqlTag: 'default',
      },
    },
  // ...
],
  // ...
}

You can access projectId and dataset by executing sanity debug --secrets in the Sanity studio folder. Note that the token printed may be used for development, but is tied to your Sanity CLI login session using your personal Sanity account - make sure to keep it secure and not include it in version control! For production, you'll want to make sure you use a read token generate in the Sanity management interface.

Basic usage

At this point you 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.

You should redeploy the GraphQL API everytime you make changes to the schema that you want to use in Gatsby by running sanity graphql deploy from within your Sanity project directory

Explore http://localhost:8000/___graphql after running gatsby develop to understand the created data and create a new query and checking available collections and fields by typing CTRL + SPACE.

Options

| Options | Type | Default | Description | | ------------- | ------- | --------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | 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 Learn more | | graphqlTag | string | default | If the Sanity GraphQL API was deployed using --tag <name>, use this to specify the tag name. | | overlayDrafts | boolean | false | Set to true in order for drafts to replace their published version. By default, drafts will be skipped. | | watchMode | boolean | false | Set to true to keep a listener open and update with the latest changes in realtime. If you add a token you will get all content updates down to each keypress. |

Preview of unpublished content

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.

Real-time content preview with watch mode

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 a token with read rights and set overlayDrafts to true, each small change to the draft will immediately be applied.

GraphQL API

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 informed choices based on this information.

Previous versions did not require this, but often lead to very confusing and unpredictable behavior, which is why we have now made it a requirement.

Using images

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.

Fluid

import React from 'react'
import Img from 'gatsby-image'

const Person = ({data}) => (
  <article>
    <h2>{data.sanityPerson.name}</h2>
    <Img fluid={data.sanityPerson.profileImage.asset.fluid} />
  </article>
)

export default Person

export const query = graphql`
  query PersonQuery {
    sanityPerson {
      name
      profileImage {
        asset {
          fluid(maxWidth: 700) {
            ...GatsbySanityImageFluid
          }
        }
      }
    }
  }
`

Fixed

import React from 'react'
import Img from 'gatsby-image'

const Person = ({data}) => (
  <article>
    <h2>{data.sanityPerson.name}</h2>
    <Img fixed={data.sanityPerson.profileImage.asset.fixed} />
  </article>
)

export default Person

export const query = graphql`
  query PersonQuery {
    sanityPerson {
      name
      profileImage {
        asset {
          fixed(width: 400) {
            ...GatsbySanityImageFixed
          }
        }
      }
    }
  }
`

Available fragments

These are the fragments available on image assets, which allows easy lookup of the fields required by gatsby-image in various modes:

  • GatsbySanityImageFixed
  • GatsbySanityImageFixed_noBase64
  • GatsbySanityImageFluid
  • GatsbySanityImageFluid_noBase64

Usage outside of GraphQL

If you are using the raw fields, or simply have an image asset ID you would like to use gatsby-image for, you can import and call the utility functions getFluidGatsbyImage and getFixedGatsbyImage:

import Img from 'gatsby-image'
import {getFluidGatsbyImage, getFixedGatsbyImage} from 'gatsby-source-sanity'

const sanityConfig = {projectId: 'abc123', dataset: 'blog'}
const imageAssetId = 'image-488e172a7283400a57e57ffa5762ac3bd837b2ee-4240x2832-jpg'

const fluidProps = getFluidGatsbyImage(imageAssetId, {maxWidth: 1024}, sanityConfig)

<Img fluid={fluidProps} />

Generating pages

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:

exports.createPages = async ({graphql, actions}) => {
  const {createPage} = actions

  const result = await graphql(`
    {
      allSanityProject(filter: {slug: {current: {ne: null}}}) {
        edges {
          node {
            title
            description
            tags
            launchDate(format: "DD.MM.YYYY")
            slug {
              current
            }
            image {
              asset {
                url
              }
            }
          }
        }
      }
    }
  `)

  if (result.errors) {
    throw result.errors
  }

  const projects = result.data.allSanityProject.edges || []
  projects.forEach((edge, index) => {
    const path = `/project/${edge.node.slug.current}`

    createPage({
      path,
      component: require.resolve('./src/templates/project.js'),
      context: {slug: edge.node.slug.current},
    })
  })
}

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 gatsby develop.

"Raw" fields

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).

Quite often, you'll want to replace reference fields (eg _ref: '<documentId>'), with the actual document that is referenced. This is done automatically for regular fields, but within raw fields, you have to explicitly enable this behavior, by using the field-level resolveReferences argument:

{
  allSanityProject {
    edges {
      node {
        _rawTasks(resolveReferences: {maxDepth: 5})
      }
    }
  }
}

Portable Text / Block Content

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.

Using .env variables

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:

// In your .env file
SANITY_PROJECT_ID = abc123
SANITY_DATASET = production
SANITY_TOKEN = my-super-secret-token

// In your gatsby-config.js file
require('dotenv').config({
  path: `.env.${process.env.NODE_ENV}`
})

module.exports = {
  // ...
  plugins: [
    {
      resolve: 'gatsby-source-sanity',
      options: {
        projectId: process.env.SANITY_PROJECT_ID,
        dataset: process.env.SANITY_DATASET,
        token: process.env.SANITY_TOKEN
        // ...
      }
    }
  ]
  // ...
}

This example is based off Gatsby Docs' implementation.

Credits

Huge thanks to Henrique Doro for doing the initial implementation of this plugin, and for donating it to the Sanity team. Mad props!

Big thanks to the good people backing Gatsby for bringing such a joy to our developer days!