AWS Lambda (API Gateway)
Drop-in recipe for using Hunch from an AWS Lambda, as a handler for an AWS API Gateway.
Prerequisite
Start with the tutorial on indexing an 11ty site.
Folder Structure
Instead of sending the Hunch output to _site/hunch.json
we'll create a new folder called _lambda
that we can zip up as a Lambda image.
Note: if you're not familiar with this approach, the AWS docs (here and here) will tell you more. The summary of it is that the
.zip
file needs to hold everything accessible to the Lambda runtime.
/_lambda
/hunch.json
/index.js
/dist
/hunch.js
/normalize.js
The files are as follows:
hunch.json
- This is the Hunch output, aka the compiled index file.index.js
- The code that the Lambda will execute.dist/
- These are files bundled with Hunch. For this demo we'll manually copy+paste the files.dist/hunch.js
- The core search functionality.dist/normalize.js
- Normalize HTTP query parameters to a Hunch query object.
File Contents
For this demo, look in the ./node_modules/hunch/dist
folder and manually copy those files over to _lambda/dist
.
Then, in the index.js
file we have this simple code:
// Load the compiled index:
import { readFile } from 'node:fs/promises'
const index = JSON.parse(await readFile('./hunch.json'))
// Create an instance of Hunch using that data:
import { hunch } from './dist/hunch.js'
const search = hunch({ index })
// This will normalize the API Gateway query parameters.
import { normalize } from './dist/hunch.js'
// This is the exported Lambda event handler function:
export const handler = async event => {
// A sample API Gateway request would have something like:
// event.queryStringParameters = {
// q: 'fancy words',
// 'facets[tags]': 'cats,-rabbits',
// }
// So we need to normalize it to a HunchQueryObject:
const query = normalize(event.queryStringParameters)
const results = search(query)
return {
statusCode: 200,
body: JSON.stringify(results),
headers: { 'content-type': 'application/json' },
}
}
If you don't need authentication or anything else, that's really all there is to it.
Deploy
If you want to manually deploy the _lambda
folder, the AWS docs will walk you through that, but using a deploy tool like serverless
(see the recipe here) instead.