Major differences with "oc process" with a high impact

[ajdelaguila]

Good afternoon team,

Let me, first of all, thank you for the effort you have put into these ansible modules; they are beautiful.

I am bringing to your attention three major differences this module brings that make it more difficult when trying to replace the usage of the oc cli.

1. With the oc cli, templates can be processed locally with no connection to any OpenShift cluster which is convenient when reusing templates within a hybrid ecosystem with OpenShift and non-OpenShift clusters coexisting together.
   My proposal for this point would be to add the attribute "local" as it is in the oc cli and, when set to true, no connectivity is done to the OpenShift/Kubernetes cluster to render the objects. This would only apply when the attribute state is set to "rendered".

2. The oc cli allows multiple param files
   My proposal for this point would be to add a new attribute "parameter_files" that would allow setting up a list of parameter files.

3. The oc cli allows overwriting parameters with a precedence order when processing templates. Parameters given as literal through the command line have more priority than those loaded from param files; those loaded from param files follow the criteria "first found" so subsequent appearances of the same parameter are just skipped.
   My proposal for this point would be to load parameters using the same criteria that the oc cli does and avoid throwing an exception when duplicated parameters are provided because, sometimes, it is convenient to have default values in param files that could be overridden by a custom param file if desired.

What do you guys think? Is this something we could consider adding?