Iterators

An iterator is an object that lets you loop through an iterable, usually by invoking, implicitly or explicitly, a next method.

• In Python, iterators are objects that expose the __next__ dunder method which can be invoked to retrieve the next value in the iterable.

An iterable is an object containing a collection of items that you can get an iterator from, usually, via a method. Iterables are stateless and have no concept of what element is the ‘current’ element in a traversal – that is what the iterator handles.

• In Python, that method is __iter__.
• In C++, that method is usually begin.

Many programming languages give you a for-loop variant that basically serves as syntactic sugar in using looping through the items in an iterable.

• In Python, when you say for item in iterable, what happens behind the scenes is the iterable’s __iter__ method gets called to get an iterator, then item is assigned to whatever the iterator’s __next__ method returns. The for loop stops when __next__ raises a StopIteration exception.
• In C++, they’re called range-based for loops with the syntax for (T item : iterable) statements.
• In C#, they’re called foreach loops with the syntax foreach (T item in iterable) statements.
• In JavaScript, they’re called for of loops with the syntax for (let item of iterable) statements
• … and so on.

Iterator Design Pattern

The purpose of iterators is to let the user access the elements of a data structure through a consistent interface, regardless of whether they’re iterating through items in a vector, a binary search tree, a graph, a hash map, etc. All these data structures will provide a way for the user to get an iterator from them that can be used in a for loop, for example.