The Iterator design pattern provides a solution:
Encapsulate the access and traversal of an aggregate in a separate Iterator object.
Clients request an Iterator object from an aggregate (createIterator())
and use it to access and traverse the aggregate.
Describing the Iterator design in more detail is the theme of the following sections.
See Applicability section for all problems Iterator can solve.

• "The key idea in this pattern is to take the responsibility for access and traversal out of the list [aggregate] object and put it into an iterator object." [GoF, p257] [BGI]
• Define separate Iterator objects:
  • Define an interface for accessing and traversing the elements of an aggregate object (Iterator | next(),hasNext()).
  • Define classes (Iterator1,…) that implement the Iterator interface.
    An iterator is usually implemented as inner class of an aggregate class.
    This enables the iterator to access the internal data structures of the aggregate (see Implementation).
  • New access and traversal operations can be added by defining new iterators.
    For example, traversing back-to-front: previous(),hasPrevious().
• An aggregate provides an interface for creating an iterator (createIterator()).
• Clients can use different Iterator objects to access and traverse an aggregate object in different ways. Multiple traversals can be in progress on the same aggregate object (simultaneous traversals).