precision ::= unsigned_integer
This rule is called by -> approximate_num_type -> exact_num_type -> mumps_type_spec -> pl1_type_spec
-> approximate_num_type -> exact_num_type -> mumps_type_spec -> pl1_type_spec