Now, all data types use the same template and now specialized code for sorting. This is possible by extended the template to a more generic framework. This shift to a more generic solution enables us to implement generic composed data types like tuples and arrays as well as the design of possible future algorithms and data structures like radix trees.
Keep in mind that the current design is limited to fixed-size data types and especially disallows slices, vectors and strings. There are ideas to extend the framework to enable these use cases in future versions.
Apart from the new framework, there are some other new features and fixes included in this release:
- more supported types: char, isize, usize, arrays, tuples
- extended benchmarks
- simple macro to create alias implementations
- bug fixes for sort strategy
- extended docs