Creating an Extendable and Developer-Friendly Framework for Visualizing Data Structures and Algorithms

Abstract

The learning of data structures and algorithms is a necessary part of many computer science programs, and visual aids are a good way of enhancing the comprehension of these abstract concepts. In this thesis, we examine how one such visual aid can be made more developer friendly, extendable, and improved upon. This was done through the refactoring, addition to, and conversion into TypeScript of the existing data structure visualization library dsvis. Several data structures and algorithms were added before and after the translation into TypeScript to be able to compare the developer experience. Refactoring and translation of the codebase provided improvements to the developer experience and readability, evaluated through both our own experience and statistical analyses of code comprehension. Finally, we conclude that while we were unable to entirely reach our ambitions (in particular concerning testing and visualization logic), we were indeed able to reduce code complexity and duplication while simultaneously introducing reusable code structures, placing the framework in a good position extendability-wise.