refactor(processors): centralize dry-run diff in AbstractProcessor - #1353

Thanks for looking into this @rickbrouwer . I think overall direction/structure is fine. Not sure if it's just me, but is it possible to make the change/logic for this a bit more easy to follow in the abstract processor?

I'm also wondering whether the abstract processor's __init__ function call the subclass function and save it as entity_in_gitlab. I think there are other areas where this can be used in future for reducing some of the boilerplates from the subclasses in future.

I think the latest update looks really good @rickbrouwer . Thanks for keep working on it. Left couple of minor comments.

Just one thing stands out I think that might need a bit more clarification. Does the raw responses from python-gitlab contain read-only API fields (id, created_at, project_id, _links, etc.). If _get_current_state returns raw API objects while _get_desired_state returns clean YAML data from gitlabform config, then _print_diff will show every extra API field as a "deletion" or "modification."

Thanks again for working on this @rickbrouwer . I feel this will make it really easy for other proecessors to consume this and show drift/diff in the dry-run.