create-wrangler

🌩️ Create a wrangler.toml file from a programmable template, presented as a Javascript file

Motivation

User maintains a programmable, testable file. Create, maintain and test complex configurations.

Repetitive parts can be used as variables and collections can be built with code. Some data can be excluded from the source code and made available as environment variables.

The code file is used to build a wrangler.toml file, to be used by wrangler.

Usage

npm create wrangler [--input ./wrangler.tmpl.js] [--output ./wrangler.toml]

Arguments

--input (-i): The path to the template file. Defaults to ./wrangler.tmpl.js. --output (-o): The path to the destination file. Defaults to ./wrangler.toml.

Input file

The input file exports a named function: template, which receives a single argument: the TOML object.

const subdomains = ["", "www.", "es."];
const paths = ["/", "/app*", "/admin*"];
const ZONE = "example.com";
const {
	ACCOUNT_ID = "96b57b07c3394752a79b623f49bc5163",
	KV_NAMESPAC_ID = "798d14f4b4c949d185fcf39b0f593ac7",
} = process.env;

export const template = ({ Section }) => ({
	name: "my-worker",
	main: "src/index.ts",
	compatibility_date: "2022-07-04",
	workers_dev: false,
	account_id: ACCOUNT_ID,
	env: Section({
		production: {
			name: "my-worker-production",
			main: "src/index.ts",
			routes: subdomains.flatMap((subdomain) =>
				paths.map((path) => ({
					pattern: [subdomain, ZONE, path].join(""),
					zone_name: ZONE,
				}))
			),
			vars: Section({
				LOG_LEVEL: "warn",
			}),
			kv_namespaces: [
				{
					binding: "MY_KV_NAMESPACE",
					id: KV_NAMESPAC_ID,
				},
			],
			unsafe: {
				bindings: [
					Section({
						type: "analytics_engine",
						name: "MY_WORKER_ANALYTICS",
						dataset: "MY_WORKER_ANALYTICS",
					}),
				],
			},
		},
	}),
});

Output file

The output file is a valid wrangler.toml file.

name = 'my-worker'
main = 'src/index.ts'
compatibility_date = '2022-07-04'
workers_dev = false
account_id = '96b57b07c3394752a79b623f49bc5163'

[env.production]
name = 'my-worker-production'
main = 'src/index.ts'
routes = [
	{ pattern = 'example.com/', zone_name = 'example.com' },
	{ pattern = 'example.com/app*', zone_name = 'example.com' },
	{ pattern = 'example.com/admin*', zone_name = 'example.com' },
	{ pattern = 'www.example.com/', zone_name = 'example.com' },
	{ pattern = 'www.example.com/app*', zone_name = 'example.com' },
	{ pattern = 'www.example.com/admin*', zone_name = 'example.com' },
	{ pattern = 'es.example.com/', zone_name = 'example.com' },
	{ pattern = 'es.example.com/app*', zone_name = 'example.com' },
	{ pattern = 'es.example.com/admin*', zone_name = 'example.com' },
]
kv_namespaces = [
	{ binding = 'MY_KV_NAMESPACE', id = '798d14f4b4c949d185fcf39b0f593ac7' },
]

[env.production.vars]
LOG_LEVEL = 'warn'

[[env.production.unsafe.bindings]]
type = 'analytics_engine'
name = 'MY_WORKER_ANALYTICS'
dataset = 'MY_WORKER_ANALYTICS'

package.json entry

In this example, the predeploy script will create the wrangler.toml file before the deploy script runs wrangler publish.

{
	"scripts": {
		"predeploy": "npm_config_yes=true npm create wrangler",
		"deploy": "wrangler publish"
	}
}