Automatically Fixing Breaking Changes of Data Science Libraries

Abstract

Data science libraries are updated frequently, and new version releases commonly include breaking changes. These are updates that cause existing code to not compile or run. Developers often use older versions of libraries because it is challenging to update the source code of large projects. We propose CombyInferPy, a new tool to automatically analyze and fix breaking changes in library APIs. CombyInferPy infers rules from the history of library source code in the form of Comby templates, a structural code search and replace tool that can automatically transform code. Preliminary results indicate CombyInferPy can update the pandas library Python code. Using the Comby rules inferred by CombyInferPy, we can automatically fix several failing tests and warnings. This shows this approach is promising to help developers update libraries.