What is this?

This is a fork of Jesse Stuart amazing gatsby-source-s3-image, updated for GatsbyJS v4.

gatsby-source-s3-image is a GatsbyJS Source plugin for converting images from any S3-compliant API[1] into GatsbyJS nodes.

[1] This includes AWS S3, of course, as well as third-party solutions like Digital Ocean Spaces, or open source / self-hosted products like MinIO.

But I can just query S3 manually client-side...

Sure, knock yourself out. But there are a few benefits you get out-of-the-box with this package:

• Native integration with Gatsby's GraphQL data ontology, of course. You just provide the bucket details (and IAM credentials, if not public, which is recommended).
• Several other benefits come with this tight integration with Gatsby API's, such as intelligent caching (nobody wants to wind up with an unexpected S3 bill as your CI server happily churns out builds, amiright?); automatic image asset optimization thanks to gatsby-image, etc.
• And to top things off — gatsby-source-s3-image will automatically detect and extract image EXIF metadata from your photos, and expose this data at the GraphQL layer as node fields.

Tell me more about this EXIF stuff.

Currently supported EXIF fields that are automatically extracted when available include:

• DateCreated (date)
• DateCreatedISO (string)
• DateTime (date)
• DateTimeOriginal (number)
• ExposureTime (number)
• FNumber (number)
• FocalLength (number)
• ISO (number)
• LensModel (string)
• Model (string)
• ShutterSpeedValue (number)
• UserComment (string)

These fields are properties of the "wrapper" node, S3ImageAsset. This type composes the ImageSharp node, the File node representing the cached image on disk (fetched via the RemoteFileNode API), and lastly the extracted EXIF data. As a result, you can easily retrieve both a set of images as well as any subset of their associated metadata in a single request — or just the metadata by itself, if that's all you need. For example:

export const pageQuery = graphql`
  query PhotographyPostsQuery {
    allS3ImageAsset {
      edges {
        node {
          id
          EXIF {
            DateCreatedISO
            ExposureTime
            FNumber
            ShutterSpeedValue
          }
          childrenFile {
            childImageSharp {
              original {
                height
                width
              }
              thumbnailSizes: fluid(maxWidth: 256) {
                aspectRatio
                src
                srcSet
                sizes
              }
              largeSizes: fluid(maxWidth: 1024) {
                aspectRatio
                src
                srcSet
                sizes
              }
            }
          }
        }
      }
    }
  }
`

Usage

Setup

Add the dependency to your package.json:

$ yarn add @marceasen/gatsby-source-s3-image
$ # Or:
$ npm install --save @marceasen/gatsby-source-s3-image

Next, register the plugin with the GatsbyJS runtime in the plugins field exported from your gatsby-config.js file, filling in the values to point to wherever your bucket is hosted:

const sourceS3 = {
  resolve: 'gatsby-source-s3-image',
  options: {
    bucketName: 'easen.pics',
    domain: null, // [optional] Not necessary to define for AWS S3; defaults to `s3.amazonaws.com`
    protocol: 'https', // [optional] Default to `https`.
  },
}

const plugins = [
  sourceS3,
  // ...
]

module.exports = { plugins }

Querying

As mentioned above, gatsby-source-s3-image exposes nodes of type S3ImageAsset:

interface S3ImageAssetNode {
  id: string
  absolutePath: string
  LastModified: Date
  ETag: string
  Key: string
  EXIF?: ExifData
  internal: {
    content: string
    contentDigest: string
    mediaType: string
    type: string
  }
}

interface ExifData {
  DateCreated?: Date
  DateCreatedISO?: string
  DateTime?: Date
  DateTimeOriginal?: number
  ExposureTime?: number
  Exposure?: string
  FNumber?: number
  FocalLength?: number
  ISO?: number
  LensModel?: string
  Model?: string
  ShutterSpeedFraction?: string
  ShutterSpeedValue?: string
  UserComment?: string
}

Not only can this be used to populate page data, I've found it useful in bootstrapping the pages themselves, e.g., to programmatically create dynamic Photo Gallery pages at build time depending on the contents of a bucket. For example:

// In `gatsby-node.js` -- using a query like this:
const photographyQuery = graphql`
  {
    allS3ImageAsset {
      edges {
        node {
          ETag
          EXIF {
            DateCreatedISO
          }
        }
      }
    }
  }
`

// We can then dynamically generate pages based on EXIF data, like this:
const createPages = ({ actions }) => {
  const { createPage } = actions
  const photographyTemplate = path.resolve(
    './src/templates/photography-post.js'
  )

  const createPhotographyPosts = edges => {
    // Create the photography "album" pages -- these are a collection of photos
    // grouped by ISO date.
    const imagesGroupedByDate = _.groupBy(edges, 'node.EXIF.DateCreatedISO')
    _.each(imagesGroupedByDate, (images, date) => {
      createPage({
        path: `/photography/${date}`,
        component: photographyTemplate,
        context: {
          name: date,
          datetime: DateTime.fromISO(date),
          type: PageType.Photography,
        },
      })
    })
  }
}