feat: add co-build integration in code base

[ashuralyk]

Description

co-build integration for Spore contract.

Address in Spore action definition means the intense of users if to check corresponding address, like from and to, since each Spore transaction has its from or to address, we require co-build related contracts to check their validation of addresses.

I've deleted and adjusted some duplicated and misunderstanding test cases for simplicity.

tests has been refactored.

Tasks list

• integrate co-build in Spore contract code base
• adjust test cases for adapting co-build check

[Flouse]

Is this code useless?

spore-contract/contracts/cluster/src/entry.rs

Lines 28 to 60 in 6921c15

	#[allow(unused)]
	fn process_input(
	index: usize,
	input_source: Source,
	group_cell_in_outputs: &mut Vec<usize>,
	output_source: Source,
	) -> Result<(), Error> {
	let group_id = load_cell_type(index, input_source)?
	.unwrap_or_default()
	.args();
	for i in 0..group_cell_in_outputs.len() {
	let output_index = group_cell_in_outputs.get(i).unwrap();
	let output_group_id = load_cell_type(*output_index, output_source)?
	.unwrap_or_default()
	.args();
	if group_id.as_slice()[..] == output_group_id.as_slice()[..] {
	let group_data = load_cluster_data(index, input_source)?;
	let output_group_data = load_cluster_data(i, output_source)?;
	if group_data.name().as_slice()[..] != output_group_data.name().as_slice()[..] {
	return Err(Error::ModifyClusterPermanentField);
	}
	group_cell_in_outputs.remove(i);
	return Ok(());
	}
	}
	// can not destroy a group cell now
	Err(Error::InvalidClusterOperation)
	}

Never mind, maybe it could be deleted later.

[ashuralyk]

@ShookLyngs rerun CI again once spore-sdk enabled its co-build feature

[ShookLyngs]

About the action.mol, I have found a few potential issues, please check the following:

1. The corresponding "melt" actions for ClusterProxy and ClusterAgent are not defined. ClusterProxy and ClusterAgent are both designed to be meltable, so maybe they should be added.
2. The code_hash field is not defined in both the ProxyCreate and the AgentCreate actions, but the field is defined in the Mint and the ClusterCreate actions.
3. The proxy_id field is not defined in the AgentCreate action, but when creating a ClusterAgent, a ClusterProxy must be referenced in the transaction (via payment or cell reference).

[ShookLyngs]

I have some thoughts regarding the naming of the SporeAction (just some personal suggestions):

• Inconsistent names such as Mint and ClusterCreate can be renamed to align with each other.
• Personally I feel like the naming of ClusterCreate looks off, perhaps alternatives like CreateCluster or ClusterCreation would be more appropriate.

[ashuralyk]

The code_hash field is not defined in both the ProxyCreate and the AgentCreate actions, but the field is defined in the Mint and the ClusterCreate actions

1. data_hash in proxy is not necessary, because the data itself represents cluster_id, which is included in molecule definition
2. data_hash in agent is proxy_type_hash that is used to locate proxy in celldep, I can add proxy_id to check instead

The corresponding "melt" actions for ClusterProxy and ClusterAgent are not defined. ClusterProxy and ClusterAgent are both designed to be meltable, so maybe they should be added.

in both proxy and agent contracts, there's no function to check validation of action melt, so I guess there's no need to add melt action for them

@ShookLyngs

[ShookLyngs]

1. data_hash in proxy is not necessary, because the data itself represents cluster_id, which is included in molecule definition
2. data_hash in agent is proxy_type_hash that is used to locate proxy in celldep, I can add proxy_id to check instead

Have I misunderstood why the data_hash is required in a creation action?
I thought it was for validating the code hashes. Could you please explain the design behind it? Thank you.

in both proxy and agent contracts, there's no function to check validation of action melt, so I guess there's no need to add melt action for them

Do you mean the co-build feature should be disabled when melting ClusterProxy and ClusterAgent?
I think I need more instructions about when it should be enabled/disabled.

[ashuralyk]

Have I misunderstood why the data_hash is required in a creation action?

I guess no, because current co-build design is rely on us, we have responsibility to design it to match our need, so the molecule I designed is to match my need that collected from contracts implementation

I think if you have your sense of design from sdk, feel free to post it out, and we can discuss

[ShookLyngs]

I guess no, because current co-build design is rely on us, we have responsibility to design it to match our need, so the molecule I designed is to match my need that collected from contracts implementation

From my perspective, co-build is a feature that enables typed messages in ckb wallets. To achieve this, each type of Spore-related transaction, such as creation, transfer, and melting, should have a corresponding action defined in the molecule. Otherwise, I'm not sure what would happen if a user tries to melt a ClusterProxy, would the wallet simply request the user to sign a hash instead of a typed message?

- User -> Melt Spore -> Sign for typed message: "melting spore 0x..."
- User -> Melt ClusterProxy -> ?

Need more info on this, thx.

[ashuralyk]

To achieve this, each type of Spore-related transaction, such as creation, transfer, and melting, should have a corresponding action defined in the molecule

I got your concerns, I'll add melt action for proxy and agent only with from address as parameter

[Flouse]

Cluster Proxy and Cluster Agent is planned in Stage 2.

Are you going to test them right now? or just focus on Stage 1 before 2024-01-20?

[ashuralyk]

Cluster Proxy and Cluster Agent is planned in Stage 2.

Are you going to test them right now? or just focus on Stage 1 before 2024-01-20?

test process is the same and won't take more time, and it's better to test them all, thus we don't have to pay twice for re-deploying Cluster and Spore contracts again

[ashuralyk]

@ShookLyngs proxy/agent actions are prepared, please check it out

[ShookLyngs]

Encountered some failed cases during my tests using this branch of code: sporeprotocol/spore-sdk#68.
On the other hand, Spore/Cluster related tests have passed.

Error 11:

• Transfer ClusterProxy
• Melt ClusterProxy
• Create ClusterAgent with ClsuterProxy (via cell reference)
• Melt ClusterAgent

I've only added detailed cobuild checks to the "Transfer ClusterProxy" test case, see: https://github.com/sporeprotocol/spore-sdk/pull/68/files#diff-0d580b1ad4f9eff4775817d276fc19a0c95a2e00cc6263f94bc8c394fe44b411. Maybe you can try unpack the extra witness and see what went wrong?