The options could be extracted from the #Config which is defined in templates/config.cue. We could come up with a convention for comments and extract them to Markdown table, something like:

#Config: {
	// Runtime version info
	moduleVersion!: string
	kubeVersion!:   string

	// DOC
	// Set custom annotations and labels to all resources.
	// Note that the `app.kubernetes.io/name` and `app.kubernetes.io/version` labels
	// are automatically injected and can't be overwritten.
	metadata: timoniv1.#Metadata & {#Version: moduleVersion}

	// DOC
	// Set the number of replicas, defaults to `1`.
	replicas: *1 | int & >0
	
	// DOC
	// Set custom annotations to pods.
	podAnnotations?: {[ string]: string}
	
	// DOC
	// Set [Kubernetes pod security context](https://kubernetes.io/docs/tasks/configure-pod-container/security-context) (optional).
	podSecurityContext?: corev1.#PodSecurityContext
}

We can use the CUE AST to walk the file line by line, once we encounter a comment that starts with DOC, we would take the name and type of that field and write it to stdout as a Markdown table row.

There are many challenges with this, complex types have their own schema which can contain tens of fields, people may want to add comments to sub-fields of a complex type. It's hard to determine the defaults for complex type as those could come from the upstream schema, e.g. ClusterIP for Service.Type.

Uh oh!

Update Readme file with values available in the module #168

Description

Metadata

Metadata

Assignees

Labels

Projects

Milestone

Relationships

Development

Issue actions