Move variable creation out of specific resolver implementation

What does this MR do?

Refactors the secrets resolving part and moves out the variable creation out of the specific secret resolver.

Why was this MR needed?

As all secrets will be presented as variables, the resolver should return the value and the common part should use the value to prepare the variable entity. This reduce the code duplication with future improvements (like making variable type configurable and/or implementing additional secret resolvers).

What's the best way to test this MR?

What are the relevant issue numbers?

Closes #26953 (closed)