Accurate description of program inputs remains a critical challenge in the field of programming languages. Active learning, as a well-established field, achieves exact learning for regular languages. We offer an innovative grammar inference tool, V-Star, based on the active learning of visibly pushdown automata. V-Star deduces nesting structures of program input languages from sample inputs, employing a novel inference mechanism based on nested patterns. This mechanism identifies token boundaries and converts languages such as XML documents into VPLs. We then adapted Angluin’s L-Star, an exact learning algorithm, for VPA learning, which improves the precision of our tool. Our evaluation demonstrates that V-Star effectively and efficiently learns a variety of practical grammars, including S-Expressions, JSON, and XML, and outperforms other state-of-the-art tools.

@article{JiaT24,
  title = {V-Star: Learning Visibly Pushdown Grammars from Program Inputs},
  author = {Xiaodong Jia and Gang Tan},
  journal = {Proceedings of the ACM on Programming Languages},
  volume = {8},
  number = {PLDI},
  year = {2024},
  doi = {10.1145/3656458}
}