README.md

launch-gen

Generate code from launch YML

Running

Build it

make build

Run it

./bin/launch-gen <path-to-launch-yml>

Migrating to use in a Golang repo

This assumes you have a go mod repo.

1. Grab the latest version of launch-gen.

go get github.com/Clever/launch-gen

2. Within your repo's Makefile, declare launch-gen to be built on every install_deps invocation. Example:

install_deps:
    go mod vendor
    go build -o bin/launch-gen -mod=vendor ./vendor/github.com/Clever/launch-gen

3. Include as a blank import, to make sure launch-gen is declared as a dependency in go.mod. See Clever/template-wag for an example.

4. On the entry point of your application, include a stanza to generate the config. This should match the launch YAML in launch/.

// generate launch config
//go:generate sh -c "$PWD/bin/launch-gen -o launch.go -p main $PWD/launch/<your-application>.yml"

If you scope the current project's /bin/ to the Makefile's path, then you can simplify the stanza:

// generate launch config
//go:generate launch-gen -o launch.go -p main $PWD/launch/<your-application>.yml

5. Ensure you call go generate ./... in the Makefile for paths that would be relevant for building or running the application. This allows the //go:generate stanza to run, which helps with building.

For example:

PKGS := $(shell go list ./... | grep -v /vendor | grep -v tools)
$(PKGS): generate golang-test-all-strict-deps
	$(call golang-test-all-strict,$@)

# Before the `test` recipe is called, `make generate` is called as a prerequisite.
test: $(PKGS)

# Before the `build` recipe is called, `make generate` is called as a prerequisite.
build: generate
	$(call golang-build,$(PKG),$(EXECUTABLE))

# Create a target that will run `go generate` when `make generate` is called.
generate:
	go generate ./...

6. Run make generate. launch.go should be within the specified directory.

7. Call InitLaunchConfig() during startup of your program, and use it when needed.