feat(codegen): refactor control stack

[g4titanx]

resolves #287

[clearloop]

lgtm! btw could you please try to resolve the issue in the transfer method here as well:

zink/examples/erc20.rs

Lines 244 to 267 in 453604b

	// 4. check transfer
	{
	// 4.1. verify balance of the caller
	let info = evm
	.calldata(&contract.encode(&[
	b"balances(address)".to_vec(),
	evm.caller.to_bytes32().to_vec(),
	])?)
	.call(address)?;
	assert_eq!(info.ret, value.to_bytes32(), "{info:?}");
	// TODO: see br_balance.rs (#287)
	// 4.2. check transfer
	/* evm = evm.commit(false);
	let info = evm
	.calldata(&contract.encode(&[
	b"transfer(address,uint256)".to_vec(),
	spender.to_bytes32().to_vec(),
	half_value.to_bytes32().to_vec(),
	])?)
	.call(address)?;
	println!("{info:?}");
	assert_eq!(info.ret, true.to_bytes32(), "{info:?}"); */
	}

if I'm not mistaken, it was caused by the same problem, if it is not caused by the same problem, please open a new issue to describe the unhandled problem in the erc20 example

[g4titanx]

if I'm not mistaken, it was caused by the same problem, if it is not caused by the same problem, please open a new issue to describe the unhandled problem in the erc20 example

they are not the same issue, i realized that after debugging this. now, the erc20 test is failing with Insufficient balance or ERC20 Insufficient allowance errors. this is a different problem related to how balances are set up and accessed in the test environment
i'd create another issue for this now, let me know what you think tho

[clearloop]

great start! getting closer to the erc20 implementation!