1. Problem types are composed by a collection of components (typically one or more main types and the associated extensions) each of which being a native problem type. This function returns the next component of a problem type. The type returned by the function can be decoded in the same way as for a type returned by findident. The second parameter is used to store the current location in the table of components of the type; if this parameter is NULL, the first component of the table is returned. This function returns NULL if it is called with the reference to the last component for the given problem type. Otherwise, the returned value can be used as the input parameter ref to get the following component and so on.

2. The routine will return a type 0 as the first component of problem types including an mpproblem component.

3. A problem type has at least one component: the first component of a native type is the type itself ( i.e. the parameter type receives the value of typcode).