Understanding behavioural patterns in JavaScript

Abstract

JavaScript is one of the most popular programming languages. How- ever, understanding the dynamic behaviour of JavaScript apps is challenging in practice. There are many factors that hinder JavaScript comprehension, such as its dynamic, asynchronous, and event- driven nature, the dynamic interplay between JavaScript and the Document Object Model, and the asynchronous communication between client and server. In this research work, we have already proposed methods for understanding event-based and asynchronous JavaScript behaviour. To enhance the scalability of our methods, we propose a new technique that adopts bio-informatics algorithms to extract sequences of actions from execution traces that form higher-level patterns.