A system for classifying and clarifying Python syntax errors for educational purposes

Abstract

Many students learning Python frequently encounter syntax errors but have difficulty understanding the error messages and correcting their code. In this thesis, we designed, implemented, and performed preliminary testing of a system for classification of syntax errors commonly made by beginning coders. Errors are classified by constructing a partial syntax tree and analyzing the node containing the error with respect to its surrounding nodes. The system aims to use the classified errors to provide more precise and instructive error messages to aid students in the debugging process.