Towards a unified python programming model for decentralised applications

Abstract

This project aims to contribute to the existing research in macroprogramming of smart contract systems, with a focus on developing a Python framework for decentralised application systems on the Algorand blockchain. The proposed framework utilises a Python abstract syntax tree (AST) parser to generate both on-chain and off-chain code for Algorand blockchain decentralised applications (dApp). The primary objective was to design and build a framework that provides a unified view of the entire system while significantly reducing the lines of code and minimising the time required for developing dApps. The framework was carefully designed to abstract away low-level code and provide seamless communication between on-chain and off-chain platforms. The goal was to enable developers to exploit Algorand’s smart contract functionality while minimising the time and effort required to develop dApps. The framework supports computing and storing data on both on-chain and off-chain platforms through a macroprogramming tag-based approach and Python wrapper classes. The evaluation of the project illustrated a significant decrease in required lines of code when comparing a framework-coded dApp with a hand-coded counterpart, while providing a unified view of the entire system. The contrast evaluation in gas fee between the two approaches of developing dApps proved to be insignificant and the collected data from a questionnaire of Algorand developers assisted in proposing future improvements to the implementation to further advance the framework’s performance.