Model-driven approach to smart contract development with automatic code generation

Abstract

Existing smart contract development languages are unique to their related block chain platforms. Thus, multiple versions of the same smart contract must be developed for different block chains. This thesis presents a model-driven approach to smart contract development with automatic code generation. This approach encompasses a new XML-based programming language to develop smart contracts and a code generator that transforms those contracts to a platform-specific programming language. The XML-based language was discussed, and all supported constructs were illustrated. The code generation tool currently supports Ethereum and Neo smart contracts that are written in Solidity and C# respectively. The language and its supporting code generation tool were tested with two real-world smart contract examples. The properties and capabilities of the language and tool were demonstrated by generating Solidity and C# code correctly and efficiently. In comparison to Solidity and C#, the new language required fewer lines of code to represent the same smart contracts. Another advantage of this language is that developers can learn one language instead of several which reduces the learning curve and development time.