Smart Contracts

Developers have the choice of three languages for writing smart contracts:

The flagship language of Ethereum, and most popular language for smart contracts.
A Python like language for writing smart contracts.
LISP Like Language (LLL)
A low level language, Serpent provides a superset of LLL. There’s not a great deal of information for working with LLL, the following blog /var/log/syrinx and associated GitHub is a good place to start.

In order to deploy a smart contract onto the Ethereum blockchain, it must first be compiled into a bytecode format, then it can be sent as part of a transaction request as detailed in Creation of a smart contract.

Given that Solidity is the language of choice for writing smart contracts, it is the language supported by web3j, and is used for all subsequent examples.

Getting started with Solidity

An overview of Solidity is beyond the scope of these docs, however, the following resources are a good place to start:

Compiling Solidity source code

Compilation to bytecode is performed by the Solidity compiler, solc. You can install the compiler, locally following the instructions as per the project documentation.

To compile the solidity code run:

$ solc <contract>.sol --bin --abi --optimize -o <output-dir>/

Note: there are issues with installing solc via homebrew on OS X currently, please see the following thread for further information.

The –bin and –abi compiler arguments are both required to take full advantage of working with smart contracts from web3j.

Outputs a Solidity binary file containing the hex-encoded binary to provide with the transaction request (as per Creation of a smart contract).
Outputs a solidity application binary interface (ABI) file which details all of the publicly acessible contract methods and their associated parameters. These details along with the contract address are crucial for interacting with smart contracts. The ABI file is also used for the generation of Solidity smart contract wrappers.

There is also a –gas argument for providing estimates of the Gas required to create a contract and transact with its methods.

Alternatively, you can write and compile solidity code in your browser via the Browser-Solidity. Browser-Solidity is great for smaller smart contracts, but you may run into issues working with larger contracts.

You can also compile solidity code via Ethereum clients such as Geth and Parity, using the JSON-RPC method eth_compileSolidity which is also supported in web3j. However, the Solidity compiler must be installed on the client for this to work.

There are further options available, please refer to the relevant section in the Homestead documentation.

Deploying and interacting with smart contracts

If you want to avoid the underlying implementation detail for working with smart contracts, web3j provides Solidity smart contract wrappers which enable you to interact directly with all of a smart contract’s methods via a generated wrapper object.

Alternatively, if you wish to send regular transactions or have more control over your interactions with your smart contracts, please refer to the sections Creation of a smart contract, Transacting with a smart contract and Querying the state of a smart contract for details.

Smart contract examples

web3j provides a number of smart contract examples in the project directory src/test/resources/solidity

It also provides integration tests for demonstrating the deploying and working with those smart contracts in the project directory src/integration-test/java/org/web3j/protocol/scenarios.


EIP-20 Ethereum token standard smart contract

There is an active Ethereum Improvement Proposal (EIP), EIP-20 that has been created to define the standard functions that a smart contract providing tokens can implement.

The EIP-20 proposal provides function definitions, but does not provide an implementation example. However, there is an implementation provided in src/test/resources/solidity/contracts, which has been taken from Consensys’ Tokens project.

There are two integration tests that have been written to fully demonstrate the functionality of this token smart contract.

HumanStandardTokenGeneratedIT uses the generated HumanStandardTokenGenerated smart contract wrapper to demonstrate this.

Alternatively, if you do not wish to use a smart contract wrapper and would like to work directly with the JSON-RPC calls, please refer to HumanStandardTokenIT.

Solidity smart contract wrappers

web3j supports the auto-generation of smart contract function wrappers in Java from Solidity ABI files.

The web3j Command Line Tools tools ship with a command line utility for generating the smart contract function wrappers:

$ web3j solidity generate /path/to/<smart-contract>.bin /path/to/<smart-contract>.abi -o /path/to/src/main/java -p

Or by calling the Java class directly:

org.web3j.codegen.SolidityFunctionWrapperGenerator /path/to/<smart-contract>.bin /path/to/<smart-contract>.abi -o /path/to/src/main/java -p

Where the bin and abi are obtained as per Compiling Solidity source code.

The smart contract wrappers support all common operations for working with smart contracts:

Any method calls that requires an underlying JSON-RPC call to take place will return a Future to avoid blocking.

Construction and deployment

Construction and deployment of smart contracts happens with the deploy method:

YourSmartContract contract = YourSmartContract.deploy(
        <web3j>, <credentials>, GAS_PRICE, GAS_LIMIT,
        <param1>, ..., <paramN>);

This will create a new instance of the smart contract on the Ethereum blockchain using the supplied credentials, and constructor parameter values.

It returns a new smart contract wrapper instance which contains the underlying address of the smart contract. If you wish to construct an instance of a smart contract wrapper with an existing smart contract, simply pass in it’s address:

YourSmartContract contract = YourSmartContract.load(
        "0x<address>", web3j, credentials, GAS_PRICE, GAS_LIMIT);

Using this method, you may want to ascertain that the contract address that you have loaded is the smart contract that you expect. For this you can use the isValid smart contract method, which will only return true if the deployed bytecode at the contract address matches the bytecode in the smart contract wrapper.

contract.isValid(); // returns false if the contract bytecode does not match what’s deployed
// at the provided addres

Transaction Managers

web3j provides a TransactionManager abstraction to control the manner you connect to Ethereum clients with. The default mechanism uses web3j’s RawTransactionManager which works with Ethereum wallet files to sign transactions offline before submitting them to the network.

However, you may wish to modify the transaction manager, which you can pass to the smart contract deployment and creation methods instead of a credentials object, i.e.:

YourSmartContract contract = YourSmartContract.deploy(
        <web3j>, <transactionManager>, GAS_PRICE, GAS_LIMIT,
        <param1>, ..., <paramN>);

In addition to the RawTransactionManager, web3j provides a ClientTransactionManager which passes the responsibility of signing your transaction on to the Ethereum client you are connecting to.

Specifying the Chain Id on Transactions (EIP-155)

The RawTransactionManager takes an optional chainId parameter to specify the chain id to be used on transactions as per EIP-155. This prevents transactions from one chain being re-broadcast onto another chain, such as from Morden to Mainnet:

TransactionManager transactionManager = new RawTransactionManager(
        web3j, credentials, ChainId.MAIN_NET);

In order to avoid having to change config or code to specify which chain you are working with, web3j’s default behaviour is to not specify chain ids on transactions to simplify working with the library. However, the recommendation of the Ethereum community is to use them.

Invoking transactions and events

All transactional smart contract methods are named identically to their Solidity methods, taking the same parameter values. Transactional calls do not return any values, regardless of the return type specified on the method. Hence, for all transactional methods the Transaction Receipt associated with the transaction is returned.:

TransactionReceipt transactionReceipt = contract.someMethod(
             new Type(...),

The transaction receipt is useful for two reasons:

  1. It provides details of the mined block that the transaction resides in
  2. Solidity events that are called will be logged as part of the transaction, which can then be extracted

Any events defined within a smart contract will be represented in the smart contract wrapper with a method named process<Event Name>Event, which takes the Transaction Receipt and from this extracts the indexed and non-indexed event parameters, which are returned decoded in an instance of the EventValues object.:

EventValues eventValues = contract.processSomeEvent(transactionReceipt);

Alternatively you can use an Observable filter instead which will listen for events associated with the smart contract:

contract.someEventObservable(startBlock, endBlock).
        .subscribe(event -> ...);

For more information on working with Observable filters, refer to Filters and Events.

Remember that for any indexed array, bytes and string Solidity parameter types, a Keccak-256 hash of their values will be returned, see the documentation for further information.

Calling constant methods

Constant methods are those that read a value in a smart contract, and do not alter the state of the smart contract. These methods are available with the same method signature as the smart contract they were generated from, the only addition is that the call is wrapped in a Future.:

Type result = contract.someMethod(new Type(...), ...).get();