This page contains links to some of my experiments with ChatGPT (models: GPT-3.5 and GPT-4). Their common goal is to evaluate ChatGPT’s ability to understand various software engineering concepts and approaches, and apply them while solving coding tasks.
Links
-
Incremental coding based on test cases (and refactoring afterwards)
-
Walking Skeleton + Contract-Driven Development + Test-Driven Development
Repositories
- CSS Lexer and Parser
-
Experimental implementation of a CSS3 lexer and parser generated almost entirely by ChatGPT (under human supervision). It is written in Prolog and includes a bunch of unit and acceptance tests.