Comment on page
Getting started
Checkout a video demo of the smart contract search engine
The following image illustrates how the smart contract search engine integrates with a) a blockchain, b) a decentralized application (DApp) and any potential c) third-party application and service.
The smart contract search engine can:
- index any Ethereum compatible blockchain
- index any number of smart contracts
- index any number of blocks
- be called via Curl
- be queried via client side Javascript
- be queried via server side NodeJS
You can quickly install and use your very own smart contract search engine via a Docker instance. Alternatively, you can install and configure the smart contract search engine for use on your own Ubuntu Server instance.
All of the configuration for the harvesting is stored in a single, config.ini, Python configuration file. The explaination for each of the configuration settings i.e. rpc endpoint, index names etc. can be found here.
The core of the smart contract search engine is a single, harvest.py, Python file. All of the harvesting functionality resides in the harvest.py file.
The harvest.py file is designed to manage all of the tasks associated with harvesting. However, due to the vast differences in blockchains (block intervals, number of blocks, number of ABIs, number of contracts etc.) the harvest.py is designed so that each of the separate tasks can be individually called, using the
-m mode flag. The documentation on how to "run once at startup" can be found here. However, before you just employ that approach, please take the time to read what each of the modes do. This Detailed explanation of harvesting modes documentation will help you taylor your search engine's operation to best suit your needs.Whilst technically not the recommended usage, the harvest.py file can also be utilized as a library in the following way. Firstly by starting a Python terminal.
# Bash syntax
cd /home/smart-contract-search-engine/python
python3.6
Then by importing the Harvest code
# Python syntax
from harvest import Harvest
Then, the following code will automatically read all of the required configuration from the config.ini file (as the Harvest() object is instantiated as the variable
harvester).harvester = Harvest()
The above code will allow any of the Harvest functions to be called individually. For example.
_addressHash = "0x123...567"
harvester.getDataUsingAddressHash(_addressHash)
The smart contract data is consumed on an ABI basis. The smart contract search engine creates a unique deterministic hash of the ABI of a given smart contract. If a DApp wants to access data which is related to an instantiation of that particular smart contract, it can query the indices using the hash as a filter. Consider the following example.
The following smart contract called
ChildContract produces an ABI.Solidity
pragma solidity >=0.4.0 <0.6.0;
contract ParentContract {
uint parentContractData = 5;
function setParentContractData(uint _parentContractData) public {
parentContractData = _parentContractData;
}
function getParentContractData() public view returns (uint){
return parentContractData;
}
}
contract ChildContract is ParentContract{
uint childContractData;
function setChildContractData(uint _childContractData) public {
childContractData = _childContractData;
}
function getChildContractData() public view returns (uint) {
return childContractData;
}
}
ChildContract ABI
[
{
"constant": false,
"inputs": [
{
"name": "_parentContractData",
"type": "uint256"
}
],
"name": "setParentContractData",
"outputs": [],
"payable": false,
"stateMutability": "nonpayable",
"type": "function"
},
{
"constant": true,
"inputs": [],
"name": "getChildContractData",
"outputs": [
{
"name": "",
"type": "uint256"
}
],
"payable": false,
"stateMutability": "view",
"type": "function"
},
{
"constant": true,
"inputs": [],
"name": "getParentContractData",
"outputs": [
{
"name": "",
"type": "uint256"
}
],
"payable": false,
"stateMutability": "view",
"type": "function"
},
{
"constant": false,
"inputs": [
{
"name": "_childContractData",
"type": "uint256"
}
],
"name": "setChildContractData",
"outputs": [],
"payable": false,
"stateMutability": "nonpayable",
"type": "function"
}
]
If we look closely at the
ChildContract we can see that it inherits from the ParentContract. If you are thinking that there should technically be two ABIs in relation to the above source code you would be right! Here is the ABI of the ParentContract.ParentContract ABI
[
{
"constant": false,
"inputs": [
{
"name": "_parentContractData",
"type": "uint256"
}
],
"name": "setParentContractData",
"outputs": [],
"payable": false,
"stateMutability": "nonpayable",
"type": "function"
},
{
"constant": true,
"inputs": [],
"name": "getParentContractData",
"outputs": [
{
"name": "",
"type": "uint256"
}
],
"payable": false,
"stateMutability": "view",
"type": "function"
}
]
Nested ABIs, like the ones shown above, are very common. In order to provide the most flexibility, the smart contract search engine stores both of the ABIs against any smart contract address which houses an instance of the above deployed
ChildContract. The hash for the ParentContract is 0x5dc306fb7e9065cf256a57f077267b73491a0df567d2aa8c1e89250e96f87011 and the hash for the ChildContract is 0xfa13b708346165ef225d79a51acbc17c24b9a2f523b71272fc6160cd9d54ced7. We can see these hashes in the abiShaList section of the raw smart contract data shown below.{
"took": 0,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"skipped": 0,
"failed": 0
},
"hits": {
"total": 1,
"max_score": 1.6739764,
"hits": [
{
"_index": "devchaintwo",
"_type": "_doc",
"_id": "0xDd27F736AC616141b72eb67D2d79D3f6b1eD7d6f",
"_score": 1.6739764,
"_source": {
"TxHash": "0x2e8bab6c377a10747a78bad4cbcd4f56bc8789ad8e4f60848d1bcd6518cf6435",
"abiShaList": [
"0xfa13b708346165ef225d79a51acbc17c24b9a2f523b71272fc6160cd9d54ced7",
"0x5dc306fb7e9065cf256a57f077267b73491a0df567d2aa8c1e89250e96f87011"
],
"blockNumber": 2220917,
"creator": "0xb0695b88e44c27c8a203bba5aed78e2ae475cc68",
"contractAddress": "0xDd27F736AC616141b72eb67D2d79D3f6b1eD7d6f",
"functionDataList": {
"0": [
{
"functionDataId": "0x2ea80e958837c05ba351ea4d77e4247518f0ab9df296f632963417c79ceac7f4",
"functionData": {
"getChildContractData": "0",
"getParentContractData": "5"
},
"uniqueAbiAndAddressHash": "0xe666c441714e77c45920a855b1e93c9306f0c8768769627637993c9ae4d15bac"
}
]
},
"requiresUpdating": "yes",
"quality": "50",
"indexInProgress": "false"
}
}
]
}
}
If we would like to access all instances of the
ChildContract we could query the search engine using the ABI hash of the ChildContract. We can query using traditional client-side Javascript or query using server-side NodeJS Javascript. Here is an example of both.var esss = new ESSS("https://devchain-es.secondstate.io")
abi = '0xfa13b708346165ef225d79a51acbc17c24b9a2f523b71272fc6160cd9d54ced7';
esss.searchUsingAbi(abi)
.then(function(result) {
console.log(result)
})
.catch(function() {
console.log("Error");
});
let esss = require('./es-ss');
let ESSS = esss.ESSS;
let searchEngineProvider = new ESSS('https://devchain-es.secondstate.io');
abiHash = '0xfa13b708346165ef225d79a51acbc17c24b9a2f523b71272fc6160cd9d54ced7';
var abiSearch = searchEngineProvider.searchUsingAbi(abiHash);
abiSearch.then(function(result) {
console.log("Result is " + result);
})
.catch(function() {
console.log("Error");
});
The above request queries will return, deduplicated data from the smart contract, such as the following
[
{
"TxHash": "0x2e8bab6c377a10747a78bad4cbcd4f56bc8789ad8e4f60848d1bcd6518cf6435",
"abiShaList": [
"0xfa13b708346165ef225d79a51acbc17c24b9a2f523b71272fc6160cd9d54ced7",
"0x5dc306fb7e9065cf256a57f077267b73491a0df567d2aa8c1e89250e96f87011"
],
"blockNumber": 2220917,
"contractAddress": "0xDd27F736AC616141b72eb67D2d79D3f6b1eD7d6f",
"creator": "0xb0695b88e44c27c8a203bba5aed78e2ae475cc68",
"functionData": {
"getChildContractData": "0",
"getParentContractData": "5"
},
"functionDataId": "0x2ea80e958837c05ba351ea4d77e4247518f0ab9df296f632963417c79ceac7f4",
"indexInProgress": "false",
"quality": "50",
"requiresUpdating": "yes",
"uniqueAbiAndAddressHash": "0xe666c441714e77c45920a855b1e93c9306f0c8768769627637993c9ae4d15bac"
}
]
The strength of this system lies in the fact that there is no display configuration or schema required. In addition to this benefit, front-end developers are not required to learn any new skills, other than their existing HTML, JS and CSS. This system only requires that the front-end developer is aware of the smart contract which is driving the DApp which is under construction. Specifically, the frontend developer only needs to know what the ABI hash is. From there the frontend developer can go ahead and create HTML, JS and CSS to suit the requirements of the DApp. Here is an example.
The following Javascript code queries the smart contract search engine using an ABI hash. Once all instances of smart contracts which adhere to that hash are returned, the Javascript iterates through the contracts and sums the total of every contract's individual account balance.
Click here to see the front-end Javascript
function displayTotal() {
esss.shaAbi(JSON.stringify(abi)).then((shaResult) => {
var sha = JSON.parse(shaResult).abiSha3;
esss.searchUsingAbi(sha).then((searchResult) => {
var items = JSON.parse(searchResult);
var totalBodyInner = "";
var total = 0;
items.forEach(function(item) {
total = total + parseInt(item.functionData.getAccountBalance);
});
console.log(total)
totalBodyInner = totalBodyInner + "<tr id='total'><td>" + total + "</tr>";
document.querySelector("#totalBody").innerHTML = totalBodyInner;
});
}); // end of esss
}
This code (which uses the ABI hash as a filter) demonstrates that the front-end is only required to parse and iterate through the smallest amount of data. The smart contract search engine does all of the work on the server side and only passes the appropriate slices of filtered data to the DApp's JS and subsequently the HTML output.
You can use the Data Drive DApp's source code and try this demonstration inside the SecondState BUIDL tool. It literally only takes a few minutes to complete and does not require you to have any cryptocurrency or a wallet. The BUIDL tool works in any browser and even handheld devices such as iPads.