Improving allocation cost for Validation

[elichai]

Hi :) It would be great if we could create a validation without allocating (and also paying for a HashSet when you only have 1/2 entries), as I have a bunch of issuers that I need to differentiate but I don't want to store a bunch of Validation objects and also some of their fields outisde (e.g. duplication across tid/issuer etc in azure) It would be great if we can enable reference based validation, this can be done via:

1. A trait Validation (e.g with fn contains_issuer(issuer: &str) -> bool etc. with struct Validation implementing it, and downstream users may implement them however they want.
2. Use some sort of enums inside Validation or provide a ValidationRef type

If this is something that is desired I can try to propose and implement the desired outcome

[Keats]

The idea with Validation was really that you keep them around so it was not designed with that kind of optimization in mind. The trait approach would be better but I don't know if enough people want that to be worth doing