Defined a rule for variable and constant numeric types

Defined a rule for variable and constant numeric types
- using std::size_t for ordinary integers, those that do not impose size constraints
- using uint32_t, uint16_t or uint_8_t for integers which require specific sizes