Firefox didn't choose the C++ regex implementation because rewriting pieces of code with well defined interfaces is hard. It was for a few reasons: 1) at the time they made the decision, there was no existing rust regex library that matched the way that JavaScript worked, 2) they didn't want to write their own/rewrite existing code because regex libraries are complicated beasts, 3) the C++ code is mature and well tested, so the primary benefit of a rust implementation would be more speed (because it's easier to safetly reason about shared data) they didn't think that optimizing the regex engine was a priority.