Add Makefile target and the code to create new API

Signed-off-by: manuelbuil <[email protected]>

Author: manuelbuil

Dockerfile

@@ -5,6 +5,11 @@ RUN apk add --no-cache bash git gcc musl-dev
 WORKDIR /src
 COPY . .
 
+RUN GOPROXY=direct go install sigs.k8s.io/controller-tools/cmd/[email protected]
+RUN GOPROXY=direct go install github.com/elastic/[email protected]
+
+RUN go generate ./...
+
 RUN --mount=type=cache,id=gomod,target=/go/pkg/mod \
     --mount=type=cache,id=gobuild,target=/root/.cache/go-build \
     ./scripts/build
@@ -16,9 +21,7 @@ COPY --from=builder /src/bin/helm-controller /bin/
 FROM golang:1.24-alpine3.22 AS dev
 ARG ARCH
 ENV ARCH=$ARCH
-RUN apk add --no-cache bash git gcc musl-dev curl
-RUN GOPROXY=direct go install golang.org/x/tools/cmd/goimports@gopls/v0.18.1
-RUN GOPROXY=direct go install sigs.k8s.io/controller-tools/cmd/[email protected]
+RUN apk add --no-cache bash git curl
 RUN if [ "${ARCH}" != "arm" ]; then \
     curl -sSfL https://raw.githubusercontent.com/golangci/golangci-lint/master/install.sh | sh -s v1.64.7; \
     fi
@@ -37,6 +40,9 @@ RUN ./scripts/package
 FROM scratch AS artifacts
 COPY --from=package /src/dist/artifacts /dist/artifacts
 
+FROM scratch AS crds
+COPY --from=builder /src/pkg/crds/yaml/generated/ /
+
 FROM alpine:3.22 AS production
 COPY bin/helm-controller /usr/bin/
 CMD ["helm-controller"]

Makefile

@@ -2,13 +2,18 @@ IMAGE_NAME ?= helm-controller
 ARCH ?= amd64
 
 .DEFAULT_GOAL := ci
-.PHONY: build test validate package clean
+.PHONY: build test validate package clean generate-crd
 
-build:
+build:  generate-crd
 	DOCKER_BUILDKIT=1 docker build \
 		--target binary \
 		--output type=local,dest=. .
 
+generate-crd:
+	DOCKER_BUILDKIT=1 docker build \
+		--target crds \
+		--output type=local,dest=./pkg/crds/yaml/generated .
+
 validate:
 	docker build --target dev --build-arg ARCH=$(ARCH) -t $(IMAGE_NAME)-dev .
 	docker run --rm $(IMAGE_NAME)-dev ./scripts/validate

README.md

@@ -4,7 +4,7 @@ helm-controller
 A simple way to manage helm charts with Custom Resource Definitions in k8s.
 
 ## Manifests and Deploying
-The `./manifests` folder contains useful YAML manifests to use for deploying and developing the Helm Controller. This simple YAML deployment creates a HelmChart CRD + a Deployment using the `rancher/helm-controller` container. The YAML might need some modifications for your environment so read below for Namespaced vs Cluster deployments and how to use them properly.
+The `./manifests` folder contains useful YAML manifests to use for deploying and developing the Helm Controller. However, we recommend using the `deploy-cluster-scoped.yaml` or `deploy-namespaced.yaml` manifests included in the releases. The YAML might need some modifications for your environment so read below for Namespaced vs Cluster deployments and how to use them properly.
 
 #### Namespaced Deploys
 Use the `deploy-namespaced.yaml` to create a namespace and add the Helm Controller and CRD to that namespace locking down the Helm Controller to only see changes to CRDs within that namespace. This is defaulted to `helm-controller` so update the YAML to your needs before running `kubectl create`
@@ -22,22 +22,26 @@ To remove the Helm Controller run `kubectl delete` and pass the deployment YAML
 ## Developing and Building
 The Helm Controller is easy to get running locally, follow the instructions for your needs and requires a running k8s server + CRDs etc. When you have a working k8s cluster, you can use `./manifests/crd.yaml` to create the CRD and `./manifests/example-helmchart.yaml` which runs the `stable/traefik` helm chart.
 
-#### Locally
+### Build
+Run `make build` to build the binary and, opitonally, generate new CRDs if the API changed. We recommend execute `make` to additionally run validation and testing.
+
+#### Running it locally
 Building and running natively will start a daemon which will watch a local k8s API. See Manifests section above about how to create the CRD and Objects using the provided manifests.
 
 ```
 go build -o ./bin/helm-controller
 ./bin/helm-controller --kubeconfig $HOME/.kube/config
 ```
 
-#### docker/k8s
-An easy way to get started with docker/k8s is to install docker for windows/mac and use the included k8s cluster. Once functioning you can easily build locally and get a docker container to pull the Helm Controller container and run it in k8s. Use `make` to launch a Linux container and build to create a container. Use the `./manifests/deploy-*.yaml` definitions to get it into your cluster and update  `containers.image` to point to your locally image e.g. `image: rancher/helm-controller:dev`
+#### Running it on k8s
+Use the `deploy-cluster-scoped.yaml` or `deploy-namespaced.yaml` manifests in the assets of each release to run the helm-controller in Kubernetes.
 
 #### Options and Usage
 Use `./bin/helm-controller help` to get full usage details. The outside of a k8s Pod the most important options are `--kubeconfig` or `--masterurl` or it will not run. All options have corresponding ENV variables you could use.
 
-## Testing
-`go test ./...`
+## Testing/Validating
+`make test`
+`make validate`
 
 ## License
 Copyright (c) 2019 [Rancher Labs, Inc.](http://rancher.com)