Wrapper classes StatePost, and for initial and final states sets?

[Adda0]

Question originally raised in #277 (comment).

But initial and final can still have size(), right? since they are just sets.

They will have to be unless you want to create a wrapper class SparseStateSet. It would have a member SparseSet, method num_of_states() which would only call return this->sparse_set.size(); and add the boilerplate code for all other functions unmodified, as what we intend to do for StatePost.

However, if we do this for StatePost, we could create the wrapper for sparse set, too. It is the same wrapper as StatePost is for OrdVector<SymbolPost>. Opinions?

[Adda0]

@kilohsakul @jurajsic @tfiedor @vhavlena Any opinions?

[kilohsakul]

I would probably not implement these wrappers. It is a lot of work and code and will only complicate the interface.

[jurajsic]

Would the wrapper be useful in some way? Because I don't really see it

[vhavlena]

I would keep the things as simple as possible. I see it as a complication, which is not so beneficial to me now.

[Adda0]

Unless we plan on somehow extending the wrappers in the future (with methods named specifically for states, etc.), the wrapper is useless in my opinion. The only advantage in the current stage would be getting a number of states would be consistent with Delta and Nfa.

[tfiedor]

Sadly, I don't really care about this.

[Adda0]

No one sees a reason for the wrappers. We will keep the types as they are now. I will close this issue as resolved, then.