URI Route class for AWS Lambda@Edge to serve CloudFront

[lorezyra]

I'm hosting my blog on AWS CloudFront. I have configured a Lambda function to review the URI from CloudFront to ensure the user gets the correct folder/file from my S3 bucket. I've written a very rudimentary nodejs script, but now I need something a bit more complex as I changed my URI schema. I need to redirect users to the new URI schema. I need the router to be fast and lightweight. Express.JS is too big for my use-case.

Input is a JSON object that includes the origin URI. I need to return a JSON object with the (string) location of the file on my S3 bucket.

Is there a router class for Hexo SSG site like this already?

[yoshinorin]

@lorezyra
How about https://github.com/hexojs/hexo-generator-alias ?

[lorezyra]

Interesting!

In my case, I needed to build a new nodeJS router for my AWS CloudFront website. I had wondered if someone had already built a super-lightweight router class that I could use. Something similar to this:

const { parse } = require('url');
const Trouter = require('trouter');

const router = new Trouter();

// define the routes
router.get('/api/data/:id', (req, res) => {
  const id = req.params.id;
  // check if the ID is valid
  if (!isValidId(id)) {
    return redirect(res, '/api/data');
  }
  // prepare the URI for S3 data retrieval
  const uri = prepareUri(id);
  // set the response header with the S3 location
  const responseHeader = JSON.stringify({ 'X-My-URI': uri + '/index.html' });
  res(responseHeader);
});

router.get('/api/data', (req, res) => {
  // redirect to the new URI schema
  redirect(res, '/api/data/:id');
});

router.all('*', (req, res) => {
  // handle bad URIs
  res.status(404).send('Page not found');
});

exports.handler = async (event, context, callback) => {
  const request = event.Records[0].cf.request;
  const uri = request.uri;
  
  // parse the URI
  const { pathname } = parse(uri);
  
  // match the route
  const match = router.find(request.method, pathname);
  
  // handle the route
  if (match) {
    // create the request object
    const params = Object.assign({}, match.params);
    const query = Object.assign({}, request.querystring);
    const headers = Object.assign({}, request.headers);
    const req = { params, query, headers };
    
    // handle the route with the handler function
    const handler = match.handler;
    handler(req, (responseHeader) => {
      // create the response object
      const response = {
        status: '200',
        statusDescription: 'OK',
        headers: {
          'cache-control': [{
            key: 'Cache-Control',
            value: 'max-age=300'
          }]
        },
        body: '',
        bodyEncoding: 'text',
      };
      // set the response header
      if (responseHeader) {
        const { key, value } = JSON.parse(responseHeader);
        response.headers[key.toLowerCase()] = [{ key, value }];
      }
      // return the response
      callback(null, response);
    });
  } else {
    // handle bad URIs
    const response = {
      status: '404',
      statusDescription: 'Not Found',
      headers: {
        'cache-control': [{
          key: 'Cache-Control',
          value: 'max-age=300'
        }]
      },
      body: 'Page not found',
      bodyEncoding: 'text',
    };
    // return the response
    callback(null, response);
  }
};

// check if the ID is valid
function isValidId(id) {
  // your validation logic here
  return id && id.length > 0;
}

// prepare the URI for S3 data retrieval
function prepareUri(id) {
  // your URI preparation logic here
  return `/my-bucket/data/${id}`;
}

// redirect to a new URI
function redirect(res, uri) {
  const responseHeader = JSON.stringify({ 'Location': uri });
  res(responseHeader);
}

[lorezyra]

This is a great alternative. However, I don't want the visitor to load a page before getting redirected. I want it handled at the router on AWS. I resolved it by building my own Lambda@Edge script to inspect the HTTP traffic.

[lorezyra]

I resolved it by building my own Lambda@Edge script to inspect the HTTP traffic. Something similar to the script above.