Usage

const mediaOptimiser = require('media-optimiser');

const transformedMedia = mediaOptimiser.optimiseMedia({
  fileName,
  fileBuffer
})

Prerequisites

Check prerequisites in https://sharp.pixelplumbing.com/install

Supports reading: JPEG, PNG, WebP, AVIF, TIFF, GIF and SVG images.

Default formats: JPEG, PNG, WebP, AVIF and TIFF.

On Lambda

You'll need to deploy a lambda layer with prebuilt binaries in order to run this package on Lambda.

If using Serverless Stack, then follow these steps to replace the node_modules version of Sharp with a Lambda Layer:

Build the Layer

Add a pre-built Sharp layer zip file to your project, somewhere like layers/sharp. Use the zip file from here or build your own version with any customisations you need (additional fonts for example).

Reference the layer in your Lambda

Then reference the layer in your function. Exclude Sharp by defining it in your externalModules and then add your layer. This must be an unzipped folder, so unzip the file as part of your CI flow.

const mediaOptimiser = new sst.Function(this, "media-optimiser", {
      handler: "services/core/src/media-optimiser/media-optimiser.handler",
      functionName: `${scope.stage}-${scope.name}-media-optimiser`,
      timeout: 120,
      bundle: { externalModules: [ "sharp" ] },
      layers: process.env.ENVIRONMENT !== "development" ? [ new LayerVersion(this, "sharp", { code: Code.fromAsset("layers/sharp") }) ] : undefined,
      permissions: [
        "lambda:InvokeFunction",
        "s3:*",
        "secretsmanager:GetSecretValue"
      ]
    });

To unzip the folder add the following step to your Jenkinsfile, after installing dependencies:

stage("Unzip the Sharp lambda layer") {
    steps {
      sh "unzip layers/sharp/nodejs.zip -d layers/sharp"
    }
  }

To run locally you'll still need sharp on your local machine - this is probably best added to dev dependencies:

yarn add -D sharp