Skip to main content
Back to Tutorials

Fairy, your super tester & debugger for smart contracts

Hecate2 ·  · 6 min read · Source
NEO FAIRY TESTSMART CONTRACTC#PYTHON

Introduction#

In many cases, you may enjoy testing and debugging your N3 smart contract using neo-express. However, you might encounter frustrating exceptions arising from your contracts, others' contracts, or even Neo's source code, with no clear solutions. Or, you might run out of testnet GAS. In these situations, neo-fairy-test and neo-fairy-client are available to simulate contract executions across mainnet, testnet, or your private net, making it easier to develop your next legendary contract.

1. Setup#

Getting a full Neo.CLI with Fairy plugin installed#

  • [Simpler but not recommended] Visit the latest releases of Neo to download the latest Neo-CLI executable zip files for your operating system.
  • [Recommended] Alternatively, check out how-to-debug-neo to compile and run Neo.CLI with the complete Neo source code. You will be able to debug your compiled codes.
  • Configure your Neo.CLI for mainnet, testnet or private net. Mainnet is recommended, as it requires zero configuration and provides access to the full mainnet environment.
  • [Simpler but not recommended] Install neo-fairy-test from the latest releases (version >=3.7.5.17 for this tutorial). Unzip the downloaded release and place it at Plugins/Fairy/{Fairy.dll + config.json + fairy.json + RpcServer.json}. Plugins is a directory placed beside neo-cli.exe.
  • [Recommended] Alternatively you may compile the source codes of Fairy by yourself.
  • Configure Fairy in Plugins/Fairy/RpcServer.json. Fairy is just an extended RpcServer on localhost of both IPv4 and IPv6, and on port 16868. If you are on mainnet, the initial config from my release should be good for you. For testnet T5, just change all the Network to 894710606 and all the Port to 26868.
  • Run neo-cli.exe. If everything is set up correctly, Fairy will print a blue text like ★ Fairy server running at 0.0.0.0:16868 in your Neo-CLI.

Getting a neo-fairy-client (Python required)#

Using command line, pip install --upgrade neo-fairy-client. This is a Neo RPC client that supports fairy features. Use version >=3.7.5.16 for this tutorial. Python >=3.8 required.

You can also send HTTP requests to neo-fairy-test using any tool you prefer, though neo-fairy-client is recommended for its Neo-specific functionality.

2. Test your contracts on the mainnet#

All the Fairy executions are based on the current synchronized environment of neo-cli. To access other contracts and their storage on the mainnet, ensure that Neo-CLI is fully synced to the latest block (run show state in Neo-CLI to check progress; use the offline pack at https://sync.ngd.network/ to accelerate). If you start without full block synchronization, you will test on a past mainnet state.

Deploying your contract virtually#

from neo_fairy_client import FairyClient
target_url = 'http://127.0.0.1:16868'client = FairyClient(target_url, fairy_session='Hello world! Your first contact with Fairy')scripthash = client.virutal_deploy_from_path('./bin/sc/YourContract.nef')print(scripthash)

Testing your contract#

result = client.invokefunction('main')print(result)

If your invocation writes to smart contract storage, it will be memorized in the fairy session name chosen by you. Keep using the same session (without setting another session for your FairyClient client) for continuous invocations. Also you may call other contracts on the mainnet, and they will also virtually take effect in your fairy session.

from neo_fairy_client.utils import NeoAddressclient.set_neo_balance(1_000_000_000)print(f"Your NEO balance: {client.invokefunction_of_any_contract(NeoAddress, 'balanceOf', [wallet_scripthash])}")client.invokefunction_of_any_contract(NeoAddress, 'transfer', [wallet_scripthash, Hash160Str.zero(), 1_000_000_000, None])print(f"NEO balance of zero address: {client.invokefunction_of_any_contract(NeoAddress, 'balanceOf', [Hash160Str.zero()])}")

Setting witness scope#

from neo_fairy_client.utils import Signer, WitnessScopesigner = Signer(0xb1983fa2479a0c8e2beae032d2df564b5451b7a5, scopes=WitnessScope.Global)client.invokefunction_of_any_contract(..., signers=signer)  # or a list of signers

Read https://docs.neo.org/docs/n3/foundation/Transactions.html#signature-scope about what a witness scope is. Fairy also supports complex WitnessRules if you need it.

from neo_fairy_client.utils.WitnessRule import *signer = [Signer(account=wallet_scripthash, scopes=WitnessScope.WitnessRules, rules=Allow(Or(And(CalledByEntry(), ScriptHash(your_called_contract)), And(CalledByContract(your_called_contract), ScriptHash(underlying_contract)))))]# Your wallet_scripthash signature is valid if you are calling your_called_contract from entry, or your_called_contract is calling underlying_contract

Creating another fairy session from existing#

Imagine that you have deployed a new fungible token contract and minted billions of tokens for yourself. Now, you want to test if you can transfer all of them to another account and trade them on an exchange. You hope both tests start from the state where you hold billions of minted tokens. This is simple with copy_snapshot- just specify the original and new fairy session names. 

client.copy_snapshot('minted billion tokens', 'test trade')

Then, always execute your fairy transactions in the new fairy session test trade. You can either change the fairy session used in the client, by explicitly setting it:

client.fairy_session = 'test trade'client.invokefunction_of_any_contract(DEX_contract, 'trade', [amountIn, amountOut, ...])

or temporarily set the session for a single invokefunction:

client.invokefunction_of_any_contract(..., fairy_session='test trade')

Similarly, there is a contract_scripthash property in FairyClient. Set it explicitly to let your client keep calling the contract for all invokefunction.

3. Debug your contracts on the mainnet#

Compile your contract in debug mode to get .nefdbgnfo file

In neo-devpack-dotnet just add the flag --debug.

Installing DumpNef#

This is a tool that prints your source code along with its compiled Neo VM assembly. If you are using neo-devpack-dotnet in its latest source code commit, you can also invoke with flag: nccs YourContract.csproj --debug --assembly to dumpnef automatically. But it is still recommended to have a standalone dumpnef in your computer.

Setting debug information for your contract#

If there is YourContract.nefdbgnfo file placed beside YourContract.nef file, and if dumpnef is available on your computer, your FairyClient should automatically generate YourContract.nef.txt and set the debug info in the virutal_deploy_from_path called when your contract was deployed. But let's still see how to set debug info manually:

with open(r'bin/sc/YourContract.nefdbgnfo', 'rb') as f:    nefdbgnfo = f.read()with open(r'bin/sc/YourContract.nef.txt', 'r') as f:    dumpnef = f.read()client.set_debug_info(nefdbgnfo, dumpnef)

nefdbgnfo and dumpnef allows you to debug with source codes.

Setting breakpoints for your contract#

You may set source code breakpoints for any contract with debug info, or set assembly breakpoints for any contract on mainnet or in your fairy session.

# you may want to delete breakpoints from previous runclient.delete_source_code_breakpoints()print(client.set_assembly_breakpoints(0))print(client.set_assembly_breakpoints(3))print(client.set_source_code_breakpoint('YourContract.cs', 88))

Executing in debug mode#

print(rpc_breakpoint := client.debug_function_with_session('methodName', [args, ...]))print(rpc_breakpoint := client.debug_step_into())print(rpc_breakpoint := client.debug_step_out())print(rpc_breakpoint := client.debug_step_over())print(client.get_local_variables())print(client.get_arguments())print(client.get_static_fields())print(client.get_evaluation_stack())print(client.get_instruction_pointer())print(client.get_variable_value_by_name("yourVariableName"))print(client.get_variable_names_and_values())...

4. Interact with the blockchain#

Fairy functions as an RpcServer and is capable of relaying transactions. Additionally, Fairy offers enhanced features such as WebSocket support, AwaitConfirmedTransaction (available in both server and client), and replay_transaction (in the client). While there is no standalone document for Fairy, you can explore the source codes of neo-fairy-test/client for an informative experience!