(CRO-53) Balance tracking and synchronisation

[devashishdxt]

No description provided.

[devashishdxt]

This is a basic implementation of balance tracking. Chain trait is not implemented in this pull request. The implementation of Chain trait will involve querying Crypto.com Chain full nodes through ABCI.

[tomtau]

I'm fine with merging in PR, but as it mainly contains hollow abstractions that are being built up for some implementation, it'll be good to document their overall aim with some references to the intended usage -- this service abstracts over a storage layer for retrieving and permitting this information, initially backed by key-value db... this service abstracts over communication with Tendermint over JSON-RPC...

[tomtau]

@devashishdxt btw, nightly fails: https://travis-ci.org/crypto-com/chain/jobs/513949529#L818

[devashishdxt]

I have added README.md in client-core which explains the overall design of client-core and also the role and responsibilities of each trait and struct.

[devashishdxt]

You can check the documentation here: https://github.com/devashishdxt/chain/tree/transaction-storage/client-core

[tomtau]

The client-core README is pretty good!

would it be possible to make it compile with nightly: https://travis-ci.org/crypto-com/chain/jobs/513972066#L783 ?

If it's just the problem of the current nightly (and will disappear in the next release), it's ok to merge; if it's a persistent change in nightly that will stay in near future releases, it'll be good to fix it

[devashishdxt]

Nightly build failure is because of a bug introduced in latest nightly: rust-lang/rust#59553

This is a regression introduced by this pull request: rust-lang/rust#59421

[Centril]

Nightly build failure is because of a bug introduced in latest nightly: rust-lang/rust#59553

Note: The PR referenced fixed a bug introduced in some versions prior.

[estebank]

@devashishdxt the new behavior is the correct behavior and used to be the active behavior until a couple of releases ago before a parser refactor introduced the incorrect behavior. We might introduce a full release cycle of warnings in beta and stable for the new behavior depending on how extensive the fallout is. If you have reasons to believe this is an incorrect course of action, please voice them in the regression issue.

[devashishdxt]

@estebank I believe new behaviour is the correct behaviour. But, we’ll have to wait until changes are made in quote or in crates using it to fix our build.

[estebank]

Quote doesn't need to change. You need to wrap your integer loop using string interpolation with an ident so that quote doesn't interpret it as a number. That will fix your build and is backwards and forwards compatible.

[devashishdxt]

@estebank Our build fails because, one of our dependency, zeroize_derive, exposes a derive macro which internally uses quote. So, to fix this issue, one of them (quote or zeroize_derive) will have to make some changes. Please let me know if I’m missing something and if we can make any changes in our code to fix this.

[estebank]

I see. The fix will have to happen in zeroize_derive.