Pending block and state improvements

[mikiw]

Usage related changes

• fix of starknet_getNonce on block_id=pending

Development related changes

• after an answer from Noa is clearer that we should always have a pending block (even if it's empty) and a pending state (even if it's the same as latest state)
• new logic assumes that we always operate on a pending state and clone it to state once a new block is mined
• since pending block and state are always there block-on-demand mode is only responsible for block creation

Checklist:

• Checked out the contribution guidelines
• Applied formatting - ./scripts/format.sh
• No linter errors - ./scripts/clippy_check.sh
• No unused dependencies - ./scripts/check_unused_deps.sh
• No spelling errors - ./scripts/check_spelling.sh
• Performed code self-review
• Rebased to the latest commit of the target branch (or merged it into my branch)
• Updated the docs if needed - ./website/README.md
• Linked the issues resolvable by this PR - linking info
• Updated the tests if needed; all passing - execution info

[mikiw]

TODO: #474 is fixed by this PR buy remember to add test for that

[mikiw]

I also tried to fix hardcoded nonces in blocks_on_demand_tests but I couldn't.

I tried to use predeployed_account.lock().unwrap().set_block_id(BlockId::Tag(BlockTag::Pending)); but it led to the problem with a clone the method clone exists for struct SingleOwnerAccount<JsonRpcClient<HttpTransport>, LocalWallet>, but its trait bounds were not satisfied.

main...pending-block-improvments-test-check-nonce-bug here is a branch - Alea iacta est if anyone can fix it it would be great, maybe my rust skills are not sufficient.

I would just wait for starknet-rs update and just remove it later.

[marioiordanov]

Approving, because I am not able to understand the state logic