Server-Side

This guide will walk you through setting up an AWS S3 bucket and executing a Node.js script using Puppeteer to interact with Diffusion Studio. Follow the steps below to get your environment up and running.

Prerequisites

Before proceeding, ensure you have the following:

• Node.js installed on your machine.
• An AWS Account with permissions to create and configure S3 buckets.

AWS S3 Bucket Setup

Create an S3 Bucket

1. Log in to the AWS Management Console.
2. Navigate to the S3 service and create a new bucket.
3. Note down the bucket name and region for later use.

Create an IAM User

1. In the AWS Management Console, go to the IAM service.
2. Create a new IAM user with programmatic access.
3. Attach the AmazonS3FullAccess policy to this user to grant full access to S3.
4. Save the Access Key ID and Secret Access Key securely.

Configure CORS for the S3 Bucket

1. Go to the S3 bucket you created.

2. Navigate to the Permissions tab.

3. Under the CORS configuration section, add the following rules:

   [
     {
       "AllowedHeaders": ["*"],
       "AllowedMethods": ["GET", "PUT", "HEAD"],
       "AllowedOrigins": ["*"],
       "ExposeHeaders": []
     }
   ]

   This configuration enables your application to interact with the S3 bucket without being blocked by CORS.

Project Setup

Client-Side Configuration

Create a website that assigns the core Diffusion Studio library to the global context. Below is an example of how to do this:

import * as core from '@diffusionstudio/core';
 
Object.assign(window, { core });

Ensure this website is accessible via HTTP from the server or machine where the Node.js script will run. You can achieve this by serving it with a tool like Vite.

Server-Side Configuration

The following instructions cover the Node.js setup, particularly configuring the S3 client.

Configure the S3 Client

import { S3Client } from '@aws-sdk/client-s3';
 
const s3 = new S3Client({
  credentials: {
    accessKeyId: '<S3_ACCESS_KEY>',
    secretAccessKey: '<S3_SECRET_KEY>',
  },
  region: '<S3_REGION>',
});

Replace <S3_ACCESS_KEY>, <S3_SECRET_KEY>, and <S3_REGION> with your actual AWS credentials and region.

Create a Function to Generate Presigned URLs

Presigned URLs allow temporary access to S3 objects. Here’s how to create one:

import { PutObjectCommand } from '@aws-sdk/client-s3';
import { getSignedUrl } from '@aws-sdk/s3-request-presigner';
 
async function generatePresignedUrl() {
  const id = new Date().toISOString().split('.')[0];
  const bucketName = '<S3_BUCKET>';
  const objectKey = `output/video-${id}.mp4`;
  const expiresIn = 3600; // URL expiration time in seconds (1 hour)
 
  const command = new PutObjectCommand({
    Bucket: bucketName,
    Key: objectKey,
    ContentType: 'video/mp4',
  });
 
  return await getSignedUrl(s3, command, { expiresIn });
};

Replace <S3_BUCKET> with the name of your S3 bucket.

Define the Video Processing Pipeline

This function will be executed using Puppeteer and will leverage Diffusion Studio to render a video and upload it to S3.

async function main(presignedUrl: string) {
  try {
    const composition = new core.Composition({ backend: 'webgl' });
 
    const source = await core.VideoSource.from('https://diffusion-studio-public.s3.eu-central-1.amazonaws.com/videos/big_buck_bunny_1080p_30fps.mp4');
 
    const video = new core.VideoClip(source).subclip(0, 150);
 
    const text = new core.TextClip({
      text: 'Bunny - Our Brave Hero',
      position: 'center',
      stop: 90,
      stroke: { color: '#000000' }
    });
 
    await composition.appendClip(video);
    await composition.appendClip(text);
 
    await new core.Encoder(composition).render(presignedUrl); 
 
    return 'SUCCESS';
  } catch (e) {
    return String(e);
  }
}

Execute the Script with Puppeteer

Finally, use Puppeteer to load your website, generate the presigned URL, and run the video processing function:

import puppeteer from 'puppeteer';
 
(async () => {
  const browser = await puppeteer.launch();
  const page = await browser.newPage();
 
  // Replace with the URL of your website that loads the core library
  await page.goto('http://localhost:5173');
 
  const presignedUrl = await generatePresignedUrl(); 
 
  const res = await page.evaluate(main, presignedUrl); 
 
  console.log(res);
 
  await browser.close();
})();

This script launches a headless browser, navigates to your website, and runs the video processing pipeline, uploading the output to the S3 bucket.