An iterable is anything for can walk. iter(x) gets an iterator from it, and next(it) returns one item at a time, raising StopIteration at the end. An iterator is used up once.

it = iter([1, 2])
next(it)  # 1
next(it)  # 2
# next(it) would raise StopIteration

A class joins in by defining __iter__. If __iter__ builds a fresh iterator each time, the object can be looped over repeatedly.

A function containing yield returns a generator, which is an iterator that pauses at each yield. Values are made on demand, so sequences can be huge or endless.

def evens():
    n = 0
    while True:
        yield n
        n += 2

def both(a, b):
    yield from a
    yield from b

yield from hands over to another iterable, which is ideal for recursive walks.

Write a class Countdown(start) whose instances can be looped over more than once, producing start down to 1. Write a generator function flatten(items) yielding every non-list value from arbitrarily nested lists.

The explanation opens here once every check passes.