Refactor/remove mocks

[bobozaur]

There are plenty of occurrences where some occult mocking mechanism is employed by holding some global settings HashMap and setting/getting values from it.

This PR explores and aims to deal with this in multiple ways:

• remove this completely if possible
• adjust the mocks to only be employed when testing through #[cfg(test)]

This sets the ground for better decoupling the test code from production code.

Update

Unfortunately conditionally compiling mocks instead of real components based on the test profile does not work and neither does using real components in all tests. The reason is because with the runtime global flags there were tests that were using the mocks while others were not.

The problems were arising in the libvcx_core tests and the node JS wrapper tests. The tests were commented out to allow the CI to pass. In my opinion, the tests were awkward in the first place because:

• they test stuff that's already tested in aries_vcx (with real components)
• as higher level libraries, they rely on mocks, while the testing in aries_vcx does not
• there are e2e tests in the node JS wrapper which do a better job than those pseudo-integration mocked tests

[codecov-commenter]

Codecov Report

Merging #1005 (834d877) into main (ff6a060) will increase coverage by 0.11%.
The diff coverage is 80.24%.

❗ Current head 834d877 differs from pull request most recent head 1fddbde. Consider uploading reports for the commit 1fddbde to get more accurate results

@@            Coverage Diff             @@
##             main    #1005      +/-   ##
==========================================
+ Coverage   30.39%   30.50%   +0.11%     
==========================================
  Files         422      420       -2     
  Lines       26281    26005     -276     
  Branches     5100     5022      -78     
==========================================
- Hits         7987     7932      -55     
+ Misses      16070    15890     -180     
+ Partials     2224     2183      -41     

Flag	Coverage Δ
unittests-aries-vcx	30.50% <80.24%> (+0.11%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files	Coverage Δ
aries_vcx/src/common/anoncreds.rs	82.88% <ø> (ø)
aries_vcx/src/common/credentials/mod.rs	81.81% <ø> (ø)
aries_vcx/src/common/ledger/transactions.rs	49.61% <ø> (+0.38%)	⬆️
...vcx/src/common/primitives/credential_definition.rs	58.44% <ø> (+0.99%)	⬆️
...ies_vcx/src/common/primitives/credential_schema.rs	11.57% <ø> (-5.61%)	⬇️
...s_vcx/src/common/primitives/revocation_registry.rs	52.52% <ø> (+1.54%)	⬆️
...es_vcx/src/common/proofs/prover/prover_internal.rs	26.49% <ø> (ø)
...cx/src/common/proofs/verifier/verifier_internal.rs	34.64% <ø> (+0.30%)	⬆️
aries_vcx/src/core/profile/mod.rs	0.00% <ø> (ø)
aries_vcx/src/core/profile/modular_libs_profile.rs	65.38% <ø> (ø)
... and 34 more

... and 4 files with indirect coverage changes

[gmulhearn]

LGTM, just the comments about commented out CI + commented out code (nvm, i missed the description, my bad)

[gmulhearn]

Seems reasonable to me! although libvcx/node is not really my domain, so approval from Patrik or Miroslav as well would be ideal

[Patrik-Stas]

I stopped commenting eventually, but feel free to delete the commented tests

Very happy about this PR

[bobozaur]

I stopped commenting eventually, but feel free to delete the commented tests

Very happy about this PR

Alright. Will actually do that, since all the comments are kinda bloaty.