Check (store, macro) tests using the Lurk Alpha Evaluator have an equivalent

[huitseeker]

The following tests are using the alpha evaluator, and should have a "moral" LEM equivalent (or a good reason not to) before destruction:

Store tests

https://github.com/lurk-lab/lurk-rs/blob/dae0d6c478a83c80e3b3753c29f89e4ecd2a6313/src/store.rs#L1938-L1996
https://github.com/lurk-lab/lurk-rs/blob/dae0d6c478a83c80e3b3753c29f89e4ecd2a6313/src/store.rs#L1998-L2097
https://github.com/lurk-lab/lurk-rs/blob/dae0d6c478a83c80e3b3753c29f89e4ecd2a6313/src/store.rs#L2099-L2184

Macro tests

https://github.com/lurk-lab/lurk-rs/blob/dae0d6c478a83c80e3b3753c29f89e4ecd2a6313/lurk-macros/tests/lurk_macro_test.rs#L54-L86
https://github.com/lurk-lab/lurk-rs/blob/dae0d6c478a83c80e3b3753c29f89e4ecd2a6313/lurk-macros/tests/lurk_macro_test.rs#L88-L112

[arthurpaulino]

So, the concept of opaque pointers didn't materialize explicitly in LEM.

An opaque pointer can be expressed by a Ptr::Atom whose hash (hopefully) corresponds to the hash of some other pointer (potentially with children). At any point, if we need an API to create opaque pointers, we implement it by calling Store::hash_ptr to get the corresponding ZPtr and then translate it back to a Ptr::Atom with the same tag and value.

About those macros, I feel like we won't need them anymore, given our current testing suite, which follows the same standard without the need for those macros.

[porcuquine]

I may not follow completely, but its important for the Lurk-data manipulation API to make it easy to create opaque pointers. Are you suggesting this will be lost?

This isn't just about testing. Even though we created formal commitments to provide a 'better way', it was useful and correct that we could roll our own poor-man's hiding commitments using Cons values. This (and it's just an example) required the verifier to be able to construct claims involving opaque Cons pointers.

[arthurpaulino]

Not lost. What I meant is that a function to create opaque pointers could be added on demand (and it's almost an one-liner). But I added it to #879 regardless.

[huitseeker]

AFAICT, #879 fixes the equivalence for the first three, and the last two are covered by evaluation tests.

[huitseeker]

Closed in #879