Large‐scale semi‐automated migration of legacy C/C++ test code

This is an industrial experience report on a large semi-automated migration of legacy test code in C and C++. The particular was enabled by automating most the maintenance steps. Without automation this large-scale would not have been conducted, due to risks involved manual (risk introducing errors, risk unexpected rework, loss productivity). We describe evaluate method we used real-world case. benefits were that analysis, could make sure understand all relevant details for envisioned maintenance, without having manually read check our theories. Furthermore, transformations reiterate improve over complex scale source updates, until they “just right.” drawbacks that, first, had learn new metaprogramming skills. Second, scripts are readily reusable other contexts; necessarily developed ad-hoc task. Our analysis shows automated software as compared (hypothetical) alternative seems be better both terms avoiding mistakes rework because such mistakes. It necessary beneficial need avoided, if engineers create bespoke (and ad-hoc) transformation tools support it.