I wrote the code for Fork/Join example in java to update the benchmarks from Parallel stream vs sequential stream vs for-each loop processing in java 8. I wanted to see how my fork/join converter algorithm competes against parallel streams.

The first change was to split original KCalConverter into two classes:
* KCalConverter – it uses ConverterKt::convert to convert Abbrev into AbbrevKcal
* KCalComplexConverter – it uses ConverterKt::complexConvert to convert Abbrev into AbbrevKcal

The second change was to add the code that benchmarks fork/join converter:

The results of running this code are close to the results of running parallel stream converters

and reinforces the conclusion from Fork/Join example in java to choose the streams just because it is simpler to use.