Related publications: [G2025]
In the article "Learning the State Machine Behind a Modal Text Editor: The (Neo)Vim Case Study" by Pierre Ganty [G2025] they used active automata-based learning to extract the state machine underlying the modal text editor Vim. They expose the various challenges to interface an active learning library with the text editor. Furthermore, they report on how they uncovered several issues and how they were dealt with by the (Neo)Vim developers. Finally, they reflect on the possible uses of automatically extracted finite-state machines beyond bug reports. The learned model you can see in Figure 1.
