Inferring Pluggable Types with Machine Learning

TL;DR

This paper explores using machine learning models to automatically infer type qualifiers in pluggable type systems, aiming to ease deployment in legacy codebases by reducing manual annotation effort.

Contribution

It introduces the NaP-AST representation and evaluates multiple models, demonstrating the effectiveness of Graph Transformer Networks in inferring type qualifiers.

Findings

GTN achieves 0.89 recall and 0.6 precision.

Models perform well with around 16k classes, overfitting occurs beyond 22k classes.

Applying models reduces warnings in open-source Java projects.

Abstract

Pluggable type systems allow programmers to extend the type system of a programming language to enforce semantic properties defined by the programmer. Pluggable type systems are difficult to deploy in legacy codebases because they require programmers to write type annotations manually. This paper investigates how to use machine learning to infer type qualifiers automatically. We propose a novel representation, NaP-AST, that encodes minimal dataflow hints for the effective inference of type qualifiers. We evaluate several model architectures for inferring type qualifiers, including Graph Transformer Network, Graph Convolutional Network and Large Language Model. We further validated these models by applying them to 12 open-source programs from a prior evaluation of the NullAway pluggable typechecker, lowering warnings in all but one unannotated project. We discovered that GTN shows the best performance, with a recall of .89 and precision of 0.6. Furthermore, we conduct a study to estimate the number of Java classes needed for good performance of the trained model. For our feasibility study, performance improved around 16k classes, and deteriorated due to overfitting around 22k classes.