This is a plugin for GatsbyJS that allows you to seamlessly integrate ImageEngine into your Gatsby workflow.
It includes functionality to make it easy to use external CMS (e.g.: Contentful, Sanity.io, Storyblok), static File assets created through gatsby-plugin-filesystem, and others, allowing them to be used directly with Gatsby components such as GatsbyImage and ImageEngine engines.
Helpers to build your own urls and ImageEngine related functionality along with components are also exposed in order to provide flexibility in how you organise your assets and how you deal with including them in your final web pages/content.
- Global Configuration
- Query Directives
- Formats
- Graphql Resolvers
- url Resolver
- gatsbyImageData Resolver
- responsive_details Resolver
Keep your ImageEngine delivery address at hand. If you don’t have one, get one here.
You need to install the package on your npm project.
npm install @imageengine/gatsby-plugin-imageengineTo use the component you define on your gatsby-config.js the plugin, under the plugins key an object with the resolve key pointing to the package name and an options key containing at least a sources key with an array of sources you want to use.
plugins: [
// ... other plugins
{
resolve: "@imageengine/gatsby-plugin-imageengine",
options: {
sources: [
{
source: "contentful",
ie_delivery_address: "https://some-ie-url.cdn.imgeng.in/"
},
{
source: "sanityio",
ie_delivery_address: "https://another-ie-url.cdn.imgeng.in/"
},
{
source: "storyblok",
ie_delivery_address: "https://another-ie-url.cdn.imgeng.in/"
},
{source: "file"}
],
ie_delivery_address: "https://yet-another-ie-url.cdn.imgeng.in/"
}
}
]Most of these plugins on their turn require using additional plugins, including gatsby-plugin-image.
How does it work? This plugin automatically creates child ImageEngineAsset nodes when using supported sources. This means you can use graphql to access these nodes and those contain all you need to use GatsbyImage or get an ImageEngine url. E.g.:
import { GatsbyImage } from "gatsby-plugin-image";
import { useStaticQuery, graphql } from "gatsby";
const Lightbox = () => {
const data = useStaticQuery(graphql`
query {
allContentfulAsset {
nodes {
childImageEngineAsset {
gatsbyImageData(width: 500, height: 300, compression: 10)
}
}
}
}`);
return (
<div class="light-box">
{data.allContentfulAsset.childImageEngineAsset.map(({ gatsbyImageData }, index) => {
return <GatsbyImage key={`image-${index}`} image={gatsbyImageData} />;
})}
</div>
);
}Built-in sources supported by the plugin automatically include Contentful, Sanity.io and File assets. In these cases you just need to define an object inside the sources key with the source and ie_delivery_address (e.g.: {source: "contentful", ie_delivery_address: "imageengine-delivery-address}).
You can also set a global ImageEngine delivery address at the top level of the options object, with the key ie_delivery_address and in this case, sources that don't specify their own will use that. Between both, at least one has always to be set. It might be useful to use a global one if different sources use the same cdn address for some reason.
In the example above, contentful will use the ie_delivery_address address of https://some-ie-url.cdn.imgeng.in/, sanityio in turn will use https://another-ie-url.cdn.imgeng.in/ and the File assets will default to https://yet-another-ie-url.cdn.imgeng.in/. This means you can create multiple engines in ImageEngine and use them easily in the same Gatsby project.
For contentful functionality to work you'll need to use gatsby-source-contentful.
With that in place Gatsby will create Graphql Nodes for your Contentful elements. When an element is of the type ContentfulAsset we'll create a child node of ImageEngineAsset under it, that you can access through graphql.
You need to have an ImageEngine delivery address using Contentful's CDN as origin and use that address as your ie_delivery_address.
For sanityio you'll need to use gatsby-source-sanity
The same as with contentful, an ImageEngineAsset node is created as a child node.
You need to have an ImageEngine delivery address using Sanity's CDN as origin and use that address as your ie_delivery_address.
For storyblok functionality to work you'll need to use gatsby-source-storyblok.
With that in place Gatsby will create Graphql Nodes for your Storyblok elements. When an element is of the type StoryblokEntry we'll create a child node of ImageEngineAsset under it, that you can access through graphql.
You need to have an ImageEngine delivery address using Storyblok's CDN as origin and use that address as your ie_delivery_address.
For File assets you'll need to use gatsby-source-filesystem
In this particular case, since it relies on static files, it will follow the same logic of using File assets with that plugin - you cannot use StaticImage with this plugin (read below).
In the filesystem plugin a file is only made available in the final build, if somewhere you query for the publicURL field of that file node. This plugin mimics that so for the final assets to be copied over to your build static folder, somewhere you'll need to query at least once for those ImageEngineAssets either the gatsbyImageData or url fields so that those files are copied over to static. This is important, because if you don't, then these files won't be available on your final build, and as such, when using the ImageEngine delivery address address it won't be able to retrieve them.
An example would be:
allFile {
nodes {
childImageEngineAsset {
url(width: 500, height: 300, compression: 10)
}
}
}Even if you know your image is static at build time, you need to use GatsbyImage instead of StaticImage. StaticImage will try to download the asset at build time to process and create its sharp variants, but the URL for "production" (when building) will be the URL of ImageEngine's CDN, so it won't exist until it's deployed, but because the build will fail to download the image, the CDN will not be able to mirror the file.
If you use the gatsby-source-filesystem plugin with a config of:
{
resolve: `gatsby-source-filesystem`,
options: {
name: `images`,
path: `${__dirname}/src/images/`,
},
}And say you have a main-page-header.jpg, you can use it on a GatsbyImage by doing:
import * as React from "react";
import { GatsbyImage } from "gatsby-plugin-image";
import { useStaticQuery, graphql } from "gatsby";
const IndexPage = () => {
const data = useStaticQuery(graphql`
query {
file(base: {eq: "main-page-header.jpg"}) {
childImageEngineAsset {
gatsbyImageData(width: 500, height: 300, compression: 10)
}
}
}`);
return (
<main>
<h1>Some header</h1>
<GatsbyImage image={data.file.childImageEngineAsset.gatsbyImageData} />
</main>
);
};While developing the image will be retrieved from your static path, and when building for production the url will be pointing to your ImageEngine distribution path with the directives passed on the gatsbyImageData field query. In this case width, height (also used by the GatsbyImage component) and compression which is unique to ImageEngine.
For types of sources for which you have no plugin available you can write your own bridge functions that enable creating ImageEngineAsset nodes.
In order to do that your source object has to specify both a checker function and a transform function.
export type IECheckFunction = (node: any, options?: IEPluginOption, global_options?: IEPluginOptions) => boolean;
export type IETransformFunction = (node: any, options?: IEPluginOption, global_options?: IEPluginOptions) => void;Support for a custom source could then be implemented with:
{
resolve: "@imageengine/gatsby-plugin-imageengine",
options: {
sources: [
{
source: "my-own-source",
ie_delivery_address: "https://some-url.cdn.imgeng.in/",
checker: my_own_source_checker,
transform: my_own_transform
}
]
}
}And then your checker function would be for instance:
import { IECheckFunction } from "@imageengine/gatsby-plugin-imageengine";
const checker: IECheckFunction = (node_object) => {
let internal = node_object?.node?.internal;
return internal && internal.type === "SomeNodeType"
};
export checker;And your transform:
import { IETransformFunction, node_create, REPLACEMENT_TOKEN } from "@imageengine/gatsby-plugin-imageengine";
const transformer: IETransformFunction = (node_object, options, global_options) => {
let dist_url = options?.ie_delivery_address || global_options?.ie_delivery_address;
let directives = options?.directives || global_options?.directives;
let replace_url = "https://some-cms-address/";
let { url, contentType } = node_object.node.my_node_details;
let tokenized_url = url.replace(replace_url, REPLACEMENT_TOKEN);
return node_create(node_object, url, tokenized_url, dist_url, directives, contentType);
};
export transformer;
}node_object is the parent object for which the checker function returns true.
options are the specific options of the source object that specify this type of asset.
global_options are the full options passed to the gatsby-plugin-imageengine
When you're inside your transformer function you know this particular node returned true for your checker function.
Ultimately you need to call node_create with this node (which will be the parent), the default original url, a tokenized url (which is a url with a specific token for replacement that you need to use), the ImageEngine Engine distribution address, any default directives and optionally a mimeType.
function node_create(node_object: any, base_url: string, tokenized_url: string, distribution_url: MaybeString, directives: IEDirectives, mime_type?: string | undefined)In this particular case, let's imagine your node is an asset in some CMS and has a url field that contains the address to that asset. We replace the portion of the CMS address from the URL with the given token (imported from gatsby-plugin-imageengine) and use that as the tokenized_url and the original url as the base_url.
This way the plugin will know how to retrieve the final correct url for your asset in the ImageEngine Engine distribution.
ImageEngine uses what is called directives (as query string paramaters) to define specific output details of images you want to optimize.
To see all possible directives and their values @imageengine/imageengine-helpers
@imageengine/imageengine-helpers is a dependency on this package and you might use any of the types or helpers there included as well if needed.
There's two ways of specifying directives and both can be used together. You can define specific directives as defaults on the configuration, and/or specify them when querying through graphql.
For the configuration you can both set them at a global level for all sources and individually for each source. On your gatsby-node.js, in the @imageengine/gatsby-plugin-imageengine configuration options
options: {
sources: [
{
source: "contentful",
ie_delivery_address: "https://some-ie-url.cdn.imgeng.in/",
directives: {
format: "png",
compression: 5,
fit: "box"
}
},
{
source: "sanityio",
ie_delivery_address: "https://another-ie-url.cdn.imgeng.in/"
},
{
source: "storyblok",
ie_delivery_address: "https://another-ie-url.cdn.imgeng.in/"
},
{source: "file"}
],
ie_delivery_address: "https://yet-another-ie-url.cdn.imgeng.in/",
directives: {
format: "jpg",
sharpness: 5,
fit: "cropbox"
}
}This means that assets for Contentful would have the default directives applied to each node automatically of {format: "png", compression: 5, sharpness: 5, fit: "box"} and all remaining ones {format: "jpg", sharpness: 5, fit: "cropbox"}.
These are specified when querying for the specific ImageEngineAsset fields:
url(width: 500, height: 300, compression: 10, format: gif, fit: cropbox, sharpness: 30)The individual directives are applied on top of any global ones, and, ultimately, query level ones applied over any other. Basically the precedence is query directives -> source default directives -> global default directives.
Besides the normal directives you can "override" the distribution address set in the config by using:
url(width: 500, height: 300, compression: 10, format: gif, fit: cropbox, sharpness: 30, ie_delivery_address: "https://some-dist.com/")Notice that gatsbyImageData accepts in addition to the ImageEngine directives, the normal arguments for it, such as formats, sizes, placeholders, etc. Refer to the gatsby-plugin-image.
There's 2 ways of specifying formats. ImageEngine serves by default the best format and compression level based on the network, browser and device requesting the image can handle, unless explicitly told to use a given format. GatsbyImage on the other hand has a tighter integration with sharp in order to provide that sort of functionality natively in Gatsby by specifying multiple sources by default.
In order to provide a more natural query and resulting picture/sources for those used to ImageEngine capabilities the default formats for the gatsbyImageData resolver is [""], instead of the usual ["", "webp"].
On a gatsbyImageData query, you can specify either format or formats. format when specified overrides formats. formats when specified generates variations on the sources when included in a GatsbyImage component. Notice that both format and formats (as well as fit) are Graphqls enums and not strings.
To illustrate the differences you can refer to the following examples:
gatsbyImageData(width: 500, height: 300)
No format, nor formats specified. GatsbyImage an img tag with a srcset and sizes. The ImageEngine urls won't contain the format directive.
gatsbyImageData(width: 500, height: 300, format: jpg)
format specified. GatsbyImage will render an img tag with srcset and sizes. The ImageEngine urls will contain the format directive for jpeg.
gatsbyImageData(width: 500, height: 300, formats: [JPG]).
formats specified with a single type. GatsbyImage will render a single img tag, with srcset and sizes. The ImageEngine urls will contain the format directive for jpeg.
gatsbyImageData(width: 500, height: 300, formats: [NO_CHANGE, WEBP])
formats specified with two types, the original (NO_CHANGE) and WEBP. GatsbyImage will render a picture element, with an img tag, and one source for webp, while the fallback lets ImageEngine decide on the format. The
source urls for the NO_CHANGE format (the original) won't have a format directive, while the WEBP source will.
gatsbyImageData(width: 500, height: 300, formats: [JPG, WEBP])
formats specified with two types, the JPG and WEBP, GatsbyImage will render a picture element, with an img tag, and one source. The img tag will have the attributes of the JPG version and the source for the webp version.
All urls generated will have the format directive applied, respectively /f_jpeg and /f_webp.
The ImageEngineAsset nodes created by gatsby-plugin-imageengine expose three resolver fields that accept directives.
- url
- gatsbyImageData
- responsive_details
url returns a simple url, correctly prefixed for the configured ImageEngine engine distribution for the given source. This is useful if you want to place the url directly in an img tag src attribute, pass it as a prop to some component, make a link, or for instance as a background-image on a CSS selector.
Let's say you're integrating with Contentful and you have an ImageEngine engine configured for that Contentful source. Now you want to use the ImageEngine assets with specific quality and output settings. You can query for your Contentful node(s) in the same way as before, and include its childImageEngineAsset's url field, for instance:
query {
allContentfulAsset {
nodes {
childImageEngineAsset {
url(width: 500, height: 300, compression: 10, format: gif, fit: cropbox, sharpness: 30)
}
}
}
}Assuming you configured your Contentful source ie_delivery_address with https://some-ie.cdn.imgeng.in/ your returned Contentful assets would have a structure of:
{
data: {
allContentfulAsset: {
nodes: [
{
childImageEngineAsset: {
url: "https://some-ie.cdn.imgeng.in/skyr7ajehjkl/2BRg1oQsnOfNtA5KeDgRMh/844c4abe68cb56ec157aa906d98fe487/file-name.jpg?imgeng=/w_500/h_300/f_gif/m_cropbox/cmpr_10/s_30"
}
},
// other nodes
]
}
}
}The gatsbyImageData field is in all forms similar to the url but it also accepts the additional properties specific to the GatsbyImage component. Your query instead would look like:
query {
allContentfulAsset {
nodes {
title
childImageEngineAsset {
gatsbyImageData(width: 500, height: 300, compression: 10, format: gif, fit: cropbox, sharpness: 30)
}
}
}
}Now you can use the returned value in a GatsbyImage element, allowing you to use all the inbuilt functionality of that component, such as having the srcset, sizes and other useful attributes automatically filled, while having access to the fields in the Contentful node itself.
{data.allContentfulAsset.nodes.map(({ title, childImageEngineAsset }) => {
return <GatsbyImage key={title} image={childImageEngineAsset.gatsbyImageData} alt={title} />;
})}query {
allContentfulAsset {
nodes {
title
childImageEngineAsset {
responsive_details(width: 500, height: 300, compression: 10, format: gif, fit: cropbox, sharpness: 30)
}
}
}
}This will return an object:
{
width: 500,
height: 300,
src: "https://....some-source-with-ie.url/file.ext?imgeng=/w_500/h_300/f_gif/cmpr_10/m_cropbox/s_30",
srcSet: "urls_for_different_sizes_if_specified_in_query",
sizes: "sizes_breakdown_for_sizes_specified_in_query"
}This can be used directly in an img tag, e.g:
<img {...data.allContentfulAsset.nodes[0].childImageEngineAsset.responsive_details}>If you just want a component to help in using ImageEngine directives without graphql you might prefer @imageengine/react
If for some reason you only need to generate only urls with the right query parameters for ImageEngine you might import the helper functions in @imageengine/imageengine-helpers