This commit is supposed to add template support to post renderer args. Also, to make it possible to template arguments that are added to helm defaults, during the load, I'm removing default post renderer args from the state and putting them to each release, unless custom args are defined for the release.

Why do I need it?

I'm trying to implement some kind of audit log to the helmfile. The idea is pretty simple, I'm adding a configmap that contains an information that can be used to identify a person that has sync the release via helmfile. In the first implementation, I was using subcharts with some .Release.Set fields, added via templates, so I could just add a release like that

templates:
  apply-log:
    dependencies:
      - chart: ./charts/apply-log
        version: '0.1.0'
        alias: apply-log
    set:
      - name: apply-log.author
        value: '{{ env "USER" }}'

releases:
  - name: test
    inherit:
      - template: apply-log 

But from time to time, it's violating chart's values schemas, and also it needs to be enabled per-release, that I consider error-prone.

So I had an idea to use postRendererHooks for that, but I couldn't find a way to access a Release Name from the hook, and that's why I came up with this PR.

Example:

#!/bin/bash

WORKDIR=$(mktemp -d)
cat <&0 > "${WORKDIR}/result.yaml"

echo "---" >> "${WORKDIR}/result.yaml"

kubectl create configmap --dry-run=client "${1}-apply-log" -o yaml --from-literal author="${USER}" >> "${WORKDIR}/result.yaml"

cat "${WORKDIR}/result.yaml"

This script is supposed to take the first argument and use it for the resource name.

I have helmDefaults set like this:

helmDefaults:
  postRenderer: ./scripts/post_render_apply_log.sh
  postRendererArgs:
    - "{{ `{{ .Release.Name }}` }}"
 releases:
  - name: namespaces
    chart: ./charts/namespaces
    namespace: kube-public
    createNamespace: false
    inherit:
      - template: default-env-values

And then the result is

❯ hf diff -l name=namespaces

Comparing release=namespaces, chart=charts/namespaces, namespace=kube-public
kube-public, namespaces-apply-log, ConfigMap (v1) has been added:
-
+ apiVersion: v1
+ data:
+   author: allanger
+ kind: ConfigMap
+ metadata:
+   creationTimestamp: null
+   name: namespaces-apply-log

Why am I merging .HelmDefaults.PostRendererArgs to the .Release.PostRendererArgs. To avoid adding the args entry to each release, I need to merge all the args before templating, so I can access the release Data in the Args that are defined globally.