Fix dagvote equal depth bug , and other test infra

[kelemeno]

cast call 0xe42923350EF3a534f84bb101453D9B442d42Bf0c "readSentTreeVote(address)(address)" 0xFb60921A1Dc09bFEDa73e26CB217B0fc76c41461 0 0 0 --rpc-url "https://sepolia.era.zksync.dev" returns: 0x7921F9F70B754A014922Aca94947e23F1A424164 i.e. Zach, and Fb60.. is Marcin, but they are on the same depth.

I looked at the smart contract, I found a bug in it. I'm not sure it was the correct ones. The bug in was in findDistances here: b24246c46b540e44ac17b12a5e21fe027408dbfa . This function was not called in the last Marcin -> Zach tx, so there might be another bug. However, this bug might have been called earlier.

I also created a way to save the DAG from a chain. Next to dos: Remote debug process:

• Prepare tx replaying tool: use the dag saving tool to clone the graph + rerun previous txs (use cast to listen to previous txs), apply consistency checker
• Use same tools for previous wrong contract. Should be easy to debug. Check what broke, and that we actually caught the bug. ( copy old contract from gh ). The general process is:
  • replay, catch a bug, fix the bug, repeat until no bug is found, or until possible.

Based on debug process, knowing what wrong we can add more tests.

• write test that catches the bug in the old contract. (always write more tests).

[kelemeno]

Update:

• have save graph tool, replay tx tool. It will be useful also for other use cases.
• even the currently changed graph cannot replay all the txs, so we already caught at least some of the issues.
• there was an issue in dagconsistency checker, now even normal tests fail. (still need to be debugged). But the issue is the same, same height Dag votes, so I think it will solve the issue.