Iterator

• Aka: Cursor
• Intent: Provide a way to access the elements of an aggregate object sequentially without exposing its underlying representation.

Structure §

Image from: Gamma, Helm, Johnson, and Vissides

Applicability §

• There is a need to access an aggregate object’s contents without exposing its internal representation.
• Support multiple traversals of aggregate objects.
• Provide a uniform interface for traversing different aggregate structures

Consequences §

• It supports variations in the traversal of an aggregate.
• Iterators simplify the aggregate interface.
• More than one traversal can be pending on an aggregate

Implementation §

Iteration Controllers §

• External iterators – client controls the iterator. More flexible
• Internal iterator – the iterator controls the iteration. Easier to use.

Traversal Algorithm §

• The aggregate defines the traversal algorithm and a client invokes the next operation to change the state of the iterator
• The iterator is responsible for traversal. This is not viable if private variables need to be accessed

Robustness §

• If the list is modified while it is traversed, at the cost of memory, copy the aggregate and traverse the copy.
• Use robust iterators to ensure insertions and removals won’t interfere with traversal. One solution: register the iterator with the aggregate, on modification, the aggregate adjusts the iterator’s state or it maintains information internally.
• Make additional iterator operations according to use.
• Iterators may have privileged access. Make iterators friends of the aggregate if possible, and include protected operations for subclassing.

Integration with composites §

• Use an external iterator
• If the composite has an interface for traversal, then keeping track of the current node in the composite is better.
• Use Null Iterators to handle boundary conditions on empty objects by ensuring it is always done.