Make cfdm.read always return unicode strings from netCDF files

[davidhassell]

Currently, cfdm.read can return either unicode or byte strings, which sometimes makes testing for equality difficult, as the equals methods treat them the two data types as unequal (as does numpy).

It would be good it it was consistent, and the obvious choice is unicode.

>>> cfdm.environment(paths=False)
Platform: Linux-5.15.0-67-generic-x86_64-with-glibc2.35
HDF5 library: 1.12.1
netcdf library: 4.8.1
Python: 3.10.9
netCDF4: 1.6.0
numpy: 1.22.3
cfdm.core: 1.10.0.3
cftime: 1.6.2
netcdf_flattener: 1.2.0
cfdm: 1.10.0.3

[davidhassell]

This change constitutes a "major", rather than "minor" change (https://ncas-cms.github.io/cfdm/releases.html#versioning-strategy) because in a some cases existing code will behave differently. The one case where this has been identified is that it could cause a call to Data.equals to return the opposite result to before - due to different/same data types.

As a consequence, the major version number should be increased, and this change will go in at version 1.10.1.0