Netlify has recently reached 1 MILLION DEVELOPERS. That’s a lot of happy consumers. To celebrate that, here‘s a step-by-step guide to deploy your very own Netlify Function with TypeScript.
📹 If you prefer, you can have this post as a screencast video.
Please don’t forget to 👍 and 🔔 on your way out if it helped you. It would help me a ton!!! 💕
Also, check the whole source code on github/monster-as-a-service
Configuring netlify.toml
Our first needed step is to configure our Netlify deploy. If you prefer, you can set this up on the dashboard as well, but I find the configuring file more straightforward. To each their own, though.
[build]
command = "yarn build"
functions = "lambda"
# this is actually a Rewrite
# totally optional, but makes up for
# a much better UX
[[redirects]]
from = "/"
to = "/.netlify/functions/index"
status = 200
Teaching TypeScript to Babel
Now we need to setup our .babelrc
so Babel can use the TypeScript compiler (tsc
)
{
"presets": [
"@babel/preset-typescript",
[
"@babel/preset-env",
{
"targets": {
"node": true
}
}
]
]
}
Create your script tasks
Almost done, run over to your package.json
and add the two tasks you’re going to need:
"build": "netlify-lambda build src",
"ts-check": "tsc --noEmit --lib ES2015 ./src/*.ts"
The only mandatory one is build
, but I imagine you will want to be able to check all your types at some point. For any production level TypeScript code I’d consider a good practice to check your types on CI pipeline, at the very least.
🚨 That's because Babel will not check your types, simply run the compiler.
Bare bones function with TypeScript
The only thing you need to have a working Netlify Function is to export a handler
method from your .js
or .ts
file. Netlify Functions run AWS-Lambdas under the hood. So, to have your handler
typed, you can simply add types/aws-lambda
and you’re good to go.
Here’s a useful example:
import { APIGatewayEvent, Context } from 'aws-lambda'
export async function handler (
event: APIGatewayEvent,
context: Context
) {
const { msg } = event.queryStringParameters
return {
statusCode: 200,
headers: {
'Content-Type': 'application/json'
},
body: JSON.stringify({ msg })
}
}
In that case, if you hit https://your-lambda.netlify.app?msg=BlackLivesMatter
if will return the following json
:
{
"msg": "BlackLivesMatter"
}
I recommend you to install netlify-dev
globally so you can simulate production environment in your machine. If you have the package, you can run netlify dev
and it will boot up a server for you and expose your lambda on https://localhost:8888
by default.
If linked to your Netlify Account, the CLI will also allow you to deploy your project without pushing it to anywhere.
Deploy
Netlify offers you a few ways of deploying your code. The most common is by setting Continuous Deployment by integrating it with your Github, Gitlab, or Bitbucket account.
Another way to do it though, is using Netlify-Dev
CLI and hitting netlify deploy
.
💌
In case you have found this post useful, please consider sharing it with your network, that would help me a lot to continue sharing posts like this and others. 😁
Top comments (9)
Thanks, nice tutorial. That .babelrc with
targets node:true
made it work (the example given in netlify-lambda README did not work for me without adding that).Also, the redirect trick was nice, haven't used Netlify redirects before. But I changed it a little bit since I have a project with multiple functions:
the sample snippet doesn't build for me:
fixed it as follows:
(note, this won't actually enforce runtime type checks, but it will make it compile)
Interesting... it works well for me, code runs as expected and
ts-checks
return no warnings or errorsPlus
{ msg }
definitely exists in type[name: string] : string
😜ok, I see the issue, the tsconfig I was inheriting from (node12) sets strict: true, which does not pass in this case
specifically, it seems that this fails because strict implies strictNullChecks, and given that this type is nullable it won't pass (if I cast it as
{ [name: string] : string }
it's ok)Great article, thanks 🤟
Thanks a lot, Gregori!
I'm very glad you liked it! 😁
Nice tutorial!
Thanks a lot, Alexandru!
I'm glad it was useful! 💪