Lambda@Edge function for accessing React apps in different sub-directories

index.js

exports.handler = (event, context, callback) => {
  
  // Get the request object.
  const request = event.Records[0].cf.request;
  
  // Get the host from the request and take out "www." from the host if it exists.
  let host = request.headers.host[0].value;
  host = host.replace(/^www\./, "");
  
  // Check if the host contains a subdomain.
  // Eg. support.example.com <-- true
  // Eg. example.com <-- false
  
  // If it has a subdomain, get the path to the directory where the React app's static files are stored based on the subdomain's identifier (marketing, support, portal).
  // If it doesn't have a subdomain, there will be no need for a  path because the static files will be at the root of the bucket, so return an empty string.
  const subdomainPattern = /^[a-z0-9]+\.example\.com$/;
  const dir = subdomainPattern.test(host) ? host.split(".")[0] : undefined;
  const entryPoint = dir ? `/${dir}` : "";

  // Declare the website endpoint of your Custom Origin.
  const domain = "example.com.s3-website-us-east-1.amazonaws.com";
  
  // Instruct to send the request to the S3 bucket, specifying for it to look for content within the sub-directory or at the root.
  // The key here is the 'path' property. It specifies the entry point.  It does not affect the request URI (eg. /login). 
  request.origin = {
    custom: {
      domainName: domain,
      port: 80,
      protocol: "http",
      path: entryPoint,
      sslProtocols: ["TLSv1.1", "TLSv1.2"],
      readTimeout: 5,
      keepaliveTimeout: 5,
      customHeaders: {
        // Set a referer request header to gain access to read objects in the S3 bucket.
        referer : [{ key: "referer", value: `http://${host}/` }]
      }
    }
  };

  // Change the host in the request headers to match the S3 bucket's website endpoint.
  request.headers["host"] = [{ key: "host", value: domain }];
  callback(null, request);
};