A Swift runtime and SDK for Vercel Serverless Functions.
Check out the intoductory blog post and YouTube tutorial for getting started:
There's two important pieces to this package that make everything work:
- An SDK that wraps the AWSLambdaRuntime and provides access to Vercel specific API like
EdgeConfig - A Swift Package Plugin which builds your code and produces a directory structure compliant with Vercel's Build Output API
import Vercel
@main
struct App: RequestHandler {
func onRequest(_ req: Request) async throws -> Response {
let greeting = EdgeConfig.default.get("greeting").string!
return .status(.ok).send("Hello, \(greeting)")
}
}import Vercel
@main
struct App: ExpressHandler {
static func configure(router: isolated Router) async throws {
router.get("/") { req, res in
res.status(.ok).send("Hello, Swift")
}
}
}import Vapor
import VercelVapor
@main
struct App: VaporHandler {
static func configure(app: Application) async throws {
app.get { _ in
"Hello, Vapor"
}
}
}And make sure to update your Package.swift with the new package:
dependencies: [
.product(name: "Vercel", package: "Vercel"),
.product(name: "VercelVapor", package: "Vercel")
]You can use any popular library to fetch data such as Alamofire or async-http-client but we also provide a convenient fetch() method directly in this package:
let obj = try await fetch("https://httpbin.org/json").json()This package provides full access to Vercel's Edge Config API. You can access the default edge config store or any additional store assigned to your project:
// Default edge config
let str = EdgeConfig.default.get("some-string-key").string
// Edge config assigned to an environment variable
let num = EdgeConfig("EDGE_CONFIG_2").get("some-int-key").intYou can add a top level public folder that will be deployed statically to Vercel's CDN.
Cron jobs are fully supported by adding a vercel.json file to the root of your project and following the Vercel documentation here: https://vercel.com/docs/cron-jobs
Running server side locally has traditionally been a huge pain, but not anymore. This package makes it trivial to run code locally:
swift package --disable-sandbox vercel devThis will build and run your Swift application and start a local server at http://localhost:7676
To deploy your project locally you need to install Docker and the Vercel CLI. Once installed you can you must link your Vercel project:
vercel linkAfter linking your project you can deploy it via Swift package manager:
swift package --disable-sandbox vercel deployswift package --disable-sandbox vercel deploy--prod- Triggers a production deploy to Vercel--product <name>- The product you want to build. Default: first target in Package.swift with the Vercel dependency--memory <number>- The amount of memory in megabytes to allocate to your function. Default 512mb--duration <number>- The maximum duration in seconds that your function will run. Default: 10s--regions <name>- Comma separated list of regions to deploy your function to. Default: iad1--port <number>- Custom port to run the local dev server on. Default: 7676
Use the following GitHub actions workflow to continuiously deploy your project to Vercel:
name: Vercel
on:
push:
branches:
- main
env:
VERCEL_ORG_ID: ${{ secrets.VERCEL_ORG_ID }}
VERCEL_PROJECT_ID: ${{ secrets.VERCEL_PROJECT_ID }}
VERCEL_TOKEN: ${{ secrets.VERCEL_TOKEN }}
jobs:
deploy:
runs-on: ubuntu-latest
container: swift:5.10-amazonlinux2
steps:
- uses: actions/checkout@v3
- uses: actions/cache@v3
with:
path: .build
key: ${{ runner.os }}-spm-${{ hashFiles('Package.resolved') }}
restore-keys: |
${{ runner.os }}-spm-
- uses: actions/setup-node@v3
with:
node-version: 16
- name: Install
run: npm install -g vercel@latest
- name: Deploy
run: swift package --disable-sandbox vercel deploy