Massive performance upgrade 💨
romanize(): 324M ops/sec
deromanize(): 94M ops/sec
Setup, operations and all parts of the library have been performance optimized to handle conversions far faster than the original, while maintaining full coverage and maintaining the same API footprint