test: refactor test data generation

[SebastianElvis]

This PR refactors the tests, including

• making create_new_finality_provider generate valid finality providers
• making get_derived_btc_delegation generate valid BTC delegations
• removing the babylon-private dependency

This is the prerequisite for testing the full validation of finality providers / BTC delegations.

[maurolacy]

LGTM. Just some issues with data files name formatting, and some outdated module refs that would be nice to fix before merging.

[maurolacy]

So, if I understand correctly, the same random source needs to be used for all of the data generation?

[SebastianElvis]

Yes. This seems to be a good practice. The issue I was facing is as follows. I was using two fresh r := rand.New(rand.NewSource(time.Now().UnixNano())) to generate secret keys for covenant members and finality providers in test data. Since the code was fast, the two r end up using the same timestamp and being the same value. Then the 1st covenant secret key is same as the 1st finality provider secret key. As Babylon does not accept duplicate keys in the staking script (i.e., any duplication among FP keys, covenant keys and staker key), the tests panicked. Using the same rand source solves this issue.

[maurolacy]

Makes sense overall, but I find difficult to believe that both r have the same seed, as it's a timestamp in nanoseconds resolution. Are these routines being called in parallel or so?