…n of radix sort

Add common typedef for sort_contig_fn_ptr_t from merge_sort header
and radix_sort header files into a new file.
Used it in cpp files.

Supported values for kind are "radixsort", "mergesort", "stable".
The default is None (same as "stable").

For stable, radix sort is used for bool, (u)int8, (u)int16.

Radix sort is supported for integral, boolean and real floating point
types.

…eter

The intent is to reduce the build time, build memory footprint, and
binary size of the sorting_impl module.

With this change it stands at 46MB, before it was 72MB.

…x_impl

With this change, _tensor_sorting_impl goes back to 17MB, and
_tensor_sorting_radix_impl is 30MB. The memory footprint of linking
should be greatly reduced, speeding up the building process,
reducing the required memory footprint, and providing better
parallelisation opportunities for the build job.

The build time on my Core i7 reduced from 45 minutes to 33 minutes.

…late-parameter-to-an-argument

Change descending from template parameter to an argument

sort.cpp -> merge_sort.cpp, argsort.cpp -> merge_argsort.cpp

Refined exception texts thrown when implementation function
pointer is found missing.