Purge vdrtools

[bobozaur]

Attempts to remove all unnecessary code from libvdrtools, keeping only the wallet and the anoncreds data structures definitions (for the wallet_migrator crate to use).

Additionally, removed lint exemption and fixed lints in the crate while also removing the migration and vdrtools/vdrtools_anoncreds feature flags from aries_vcx and aries_vcx_core.

Lastly, dependencies have been updated and tweaked. Of particular notice are sqlx which was bumped to the latest version (0.7.1) and zeroize which was bumped to the latest version (1.6.0).

[codecov-commenter]

Codecov Report

Merging #1013 (9793048) into main (603a26c) will increase coverage by 5.46%.
The diff coverage is 34.63%.

@@            Coverage Diff             @@
##             main    #1013      +/-   ##
==========================================
+ Coverage   30.49%   35.96%   +5.46%     
==========================================
  Files         419      372      -47     
  Lines       25967    21745    -4222     
  Branches     5015     4016     -999     
==========================================
- Hits         7918     7820      -98     
+ Misses      15869    11763    -4106     
+ Partials     2180     2162      -18     

Flag	Coverage Δ
unittests-aries-vcx	35.96% <34.63%> (+5.46%)	⬆️

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/keys.rs	50.00% <ø> (ø)
aries_vcx/src/common/mod.rs	50.00% <ø> (-4.02%)	⬇️
...s_vcx/src/common/primitives/revocation_registry.rs	52.52% <ø> (ø)
aries_vcx/src/core/profile/mod.rs	0.00% <ø> (ø)
aries_vcx/src/handlers/util.rs	15.12% <ø> (-0.67%)	⬇️
aries_vcx/tests/test_credential_issuance.rs	88.95% <ø> (ø)
aries_vcx/tests/test_credential_retrieval.rs	88.38% <100.00%> (+0.80%)	⬆️
aries_vcx/tests/test_pool.rs	85.11% <ø> (ø)
aries_vcx/tests/test_proof_presentation.rs	86.36% <ø> (ø)
... and 45 more

... and 6 files with indirect coverage changes

[Patrik-Stas]

Having further look, I think there's more to be cleaned up

• I think we should be able to drop off some cargo dependencies in vdrtools directories
• there seems to be more stuff to delete in src/domain/anoncreds
• I'd suggest to delete vdrtool2credx migration
  • this would get our workspace ursa-free, which if I remember right, was major source of pain with askar addtion and sqlx update
• I believe signature_serializer.rs can be dropped and also Hash in indy-utils. Perhaps there's more stuff to delete in indy-utils. Openssl dependency can be removed.
• mac.build.sh can be deleted

[bobozaur]

* I think we should be able to drop off some cargo dependencies in vdrtools directories

Aaah right, I forgot about the dependencies! I'll work on that.

* I'd suggest to delete `vdrtool2credx` migration

@Patrik-Stas Do you mean to delete the wallet_migrator crate?

[bobozaur]

@Patrik-Stas I removed some more code and updated/tweaked dependencies.

For the other things:

• I don't think deleting the migration code is something we want. Is it? Down the line I see how we might want to incorporate the data structures from libvdrtools in there for posterity, but not really remove it.
• There's definitely code to delete from src/domain/anoncreds but it really revolves around these same data structures used by the wallet_migrator and I wouldn't really start hunting down unused methods.

[Patrik-Stas]

Do you mean to delete the wallet_migrator crate?

well, not necessarily the entire crate as it could be used for credx2anoncreds-rs or something similar, but given we have released 0.59.1 which provide migration path from vdrtools to credx, and we no longer support vdrtools, I don't see reason why to keep around the vdrtool2credx migration in particular, especially if it brings in all the ursa dependencies, that's quite heavy cost to pay.

[bobozaur]

Do you mean to delete the wallet_migrator crate?

well, not necessarily the entire crate as it could be used for credx2anoncreds-rs or something similar, but given we have released 0.59.1 which provide migration path from vdrtools to credx, and we no longer support vdrtools, I don't see reason why to keep around the vdrtool2credx migration in particular, especially if it brings in all the ursa dependencies, that's quite heavy cost to pay.

As discussed, we'll keep the data types around for a while longer just for the wallet_migrator crate. Perhaps we'll be able to move the data structures to wallet_migrator afterwards (or at least some high-level of them, without depending on ursa) and thus get rid of more things from libvdrtools.

[Patrik-Stas]

@gmulhearn if all looks good, feel free to merge

[gmulhearn-anonyome]

general question that doesn't need to be answered here: does vdr_proxy_ledger (i.e. where VdrProxySubmitter is the core component) actually need the credx feature?

You guys know this part of the code best, but it seems like the only reason vdr_proxy_ledger needs credx is because vdr_proxy_ledger is being used to hide the VdrProxyProfile implementation, which depends on IndyCredxAnonCreds (and IndySdkWallet too actually (vdrtools_wallet?)). I wonder if logically VdrProxyProfile should be hidden by credx AND vdr_proxy_ledger rather than making vdr_proxy_ledger always depend on credx

[gmulhearn-anonyome]

I guess it's a matter of perspective about whether the primary "feature" of vdr_proxy_ledger is the VdrProxySubmitter AND VdrProxyProfile, or if it's just the VdrProxySubmitter

[bobozaur]

Idk. @mirgee worked on this. Any insights? I just replaced the underlying implementation being used in the VdrProxyProfile since we're deprecating libvdrtools.

[gmulhearn-anonyome]

ye no worries, just a passing comment

[gmulhearn]

nice, good stuff!