Make token hashable

Hash functions have been implemented for Token and Lexeme (required for token). Token hashing function relies on orth_utf8, wa and lexemes getters, which corresponds to definition of already existed == operator.

Test script has been implemented, but it is not automated (CI/CD). It was not automated because it looks like corpus2 tests are not maintained (not included in CI/CD and make test informs about failed tests) so not sure where new tests should be "attached".