public class TTxState implements ContractState
Constructor and Description |
---|
TTxState(CompositeKey owner,
byte[] txContent) |
Modifier and Type | Method and Description |
---|---|
Contract |
getContract()
An instance of the contract class that will verify this state.
|
CompositeKey |
getOwner() |
java.util.List<net.corda.core.crypto.CompositeKey> |
getParticipants()
A participant is any party that is able to consume this state in a valid transaction.
|
byte[] |
getTxContent() |
getContract, getParticipants
public TTxState(CompositeKey owner, byte[] txContent)
public java.util.List<net.corda.core.crypto.CompositeKey> getParticipants()
A participant is any party that is able to consume this state in a valid transaction.
The list of participants is required for certain types of transactions. For example, when changing the notary for this state (TransactionType.NotaryChange), every participant has to be involved and approve the transaction so that they receive the updated state, and don't end up in a situation where they can no longer use a state they possess, since someone consumed that state during the notary change process.
The participants list should normally be derived from the contents of the state. E.g. for Cash the participants list should just contain the owner.
public Contract getContract()
An instance of the contract class that will verify this state.
This field is not the final design, it's just a piece of temporary scaffolding. Once the contract sandbox is further along, this field will become a description of which attachments are acceptable for defining the contract.
Recall that an attachment is a zip file that can be referenced from any transaction. The contents of the attachments are merged together and cannot define any overlapping files, thus for any given transaction there is a miniature file system in which each file can be precisely mapped to the defining attachment.
Attachments may contain many things (data files, legal documents, etc) but mostly they contain JVM bytecode.
The class files inside define not only interface Contract
implementations but also the classes that define the states.
Within the rest of a transaction, user-providable components are referenced by name only.
This means that a smart contract in Corda does two things:
Define the data structures that compose the ledger (the states)
Define the rules for updating those structures
The first is merely a utility role ... in theory contract code could manually parse byte streams by hand. The second is vital to the integrity of the ledger. So this field needs to be able to express constraints like:
Only attachment 733c350f396a727655be1363c06635ba355036bd54a5ed6e594fd0b5d05f42f6 may be used with this state.
Any attachment signed by public key 2d1ce0e330c52b8055258d776c40 may be used with this state.
Attachments (1, 2, 3) may all be used with this state.
and so on. In this way it becomes possible for the business logic governing a state to be evolved, if the constraints are flexible enough.
Because contract classes often also define utilities that generate relevant transactions, and because attachments cannot know their own hashes, we will have to provide various utilities to assist with obtaining the right code constraints from within the contract code itself.
TODO: Implement the above description. See COR-226
interface Contract
public CompositeKey getOwner()
public byte[] getTxContent()