This issue proposes removing the current translation layer and instead directly interpreting the Python AST in the CSE machine.
Currently, the interpreter converts Python AST into an intermediate format, JavaScript AST, before evaluation. While functional, this adds an extra translation step that could be avoided by interpreting the AST directly.
Interpreting the AST directly can reduces complexity and make the CSE machine more efficiant. Could also more closely reflects how a real Python interpreter behaves and could improve performance and maintainability