[Feature] Support historical logging for delegated, bonded, and unbonding` mappings.

[d0cd]

This PR supports historical logging for the delegated, bonded, and unbonding mappings in credits.aleo.

If the history feature is enabled, the VM will store the above mappings as JSON to a set local directory (the same directory as the .ledger).

This data is stored and updated as follows:

• A top-level .history* directory is created at the storage path.
• When a block i is added to the ledger, a subdirectory block-<i> is created containing: block-<i>-delegated.json, block-<i>-bonded.json, and block-<i>-unbonding.json.
• If the history directory is deleted, it is created once again.
• Data is stored as a JSON string for easier consumption. We considered storing it at bytes, but opted for usability over compression.

Pruning and management of the .history directory is left to the node runner.

Note. The entirety of this PR is gated under the history feature.

CI passed mostly here. The failing tests are succeeding locally and unrelated to this PR.

[vicsn]

LGTM!

As for CI (thx for pushing this already!):
Can you also push a clippy fix? I assume maybe a new Rust version came out.
Can you also run the synthesizer tests locally to double check they pass?

[Meshiest]

LGTM. We use a tool that uses a similar mechanism that captures the raw binary values of all mappings and they are quite compact (under 4-16KB per block)

[d0cd]

LGTM. We use a tool that uses a similar mechanism that captures the raw binary values of all mappings and they are quite compact (under 4-16KB per block)

Do you have opinions on whether the data should be stored as raw JSON strings or as bytes?

[apruden2008]

We ran our own internal CI against this and is passing. Given the fact this is gated behind a CLI flag, I think this is a safe feature to add in here at this point in the development timeline. Thank you @Meshiest and @damons for your review

[apruden2008]

See comment above. LGTM based on internal CI passing and prior reviews both inside and outside of Provable

[HarukaMa]

I just saw this and I'd highly recommend anyone who use this to store the jsons in a fs that supports compression, since with a busy network those data could get huge really fast. Had to manually implement a similar function to debug my explorer before. Also one of the reasons I don't keep a whole history of bonded and deletaged mappings - they are really huge on a live network.