Neurosymbolic repair for low-code formula languages

Most users of low-code platforms, such as Excel and PowerApps, write programs in domain-specific formula languages to carry out nontrivial tasks. Often can most the program they want, but introduce small mistakes that yield broken formulas. These mistakes, which be both syntactic semantic, are hard for identify fix, even though resolved with just a few edits. We formalize problem producing edits last-mile repair problem. To address this problem, we developed LaMirage, LAst-MIle RepAir-engine GEnerator combines symbolic neural techniques perform languages. LaMirage takes grammar set constraints/rules, jointly approximate target language, uses these generate engine fix formulas language. tackle challenges localizing errors ranking candidate repairs, leverages techniques, whereas it relies on methods This combination allows find repairs satisfy provided constraints, then pick natural repair. compare state-of-the-art approaches 400 real Power Fx formulas, where outperforms all baselines. release benchmarks encourage subsequent work domains.