fixed 32-bit integer overflow in stdlib_io_npy

[degawa]

32-bit integer overflows in stdlib_io_npy_save.fypp and stdlib_io_npy_load.fypp are fixed from 2**32 to 2**24, to become the power of base-number (256) at the 4th digit a base-256 number.
To improve readability and understandability, the powers of base-number are refactored from 2**8, 2**16, 2**24 to 256**1, 256**2, 256**3, respectively.

The tasks done are summarized as follows:

• fixed and refactored stdlib_io_npy_save.fypp
• fixed and refactored stdlib_io_npy_load.fypp
• built with gfortran 10.3, ifort 2021.1, and nagfor 7.1 on Windows 10 with cmake 3.20.3
• executed cmake --build build --target test to run the test npy and confirmed it passed

closes #647

Note that config/DefaultFlags.cmake is locally changed to specify compiler flags for nagfor as follows:

elseif(CMAKE_Fortran_COMPILER_ID MATCHES "NAG")
  set(
    CMAKE_Fortran_FLAGS_INIT
    "-fpp"
    "-f2018"
    "-ieee=full"
  )
  set(
    CMAKE_Fortran_FLAGS_RELEASE_INIT
  )
  set(
    CMAKE_Fortran_FLAGS_DEBUG_INIT
    "-g"
    "-nan"
  )

[sakamoti]

I think this patch looks fine.
stdlib_io_npy_save and stdlib_io_npy_load have passed compilation phaze with nagfor7.1 (Linux version).

[awvwgk]

Thanks for sharing. Looks good to me.

[milancurcic]

Looks good, thank you!