Ease Solidity Smart Contract Compilation through Version Pragma Identification

【Author】 Tian, Zhenzhou; He, Ruikang

【Source】ADVANCES IN NATURAL COMPUTATION, FUZZY SYSTEMS AND KNOWLEDGE DISCOVERY, ICNC-FSKD 2022

【影响因子】

【Abstract】Solidity, the most popular language for developing smart contracts that run on blockchains, has experienced a fast version changing since its first release. Yet, the fact that different Solidity versions are generally incompatible, makes it torturous for developers to correctly pick the Solidity version to compile a smart contract missing version-indicative information. This work presents SolCom to ease the compilation of a Solidity smart contract. It infers from the source code of a smart contract its version pragma, with which the developers can accurately determine the right-version Solidity compiler to get the source code compiled rather than aimless attempts. SolCom processes the source code into normalized tokens, and resorts to a carefully-designed attention augmented convolutional neural network model to capture version indicative features. As the experimental evaluations show, SolCom can identify the version pragma with a rather high accuracy of 96.30%.

【Keywords】Smart contract; Version pragma; Solidity; Neural networks

【发表时间】2023

【收录时间】2023-05-07

【文献类型】理论模型

【主题类别】

区块链技术-核心技术-智能合约