Iterators and copy

Often it is simply not necessary to make a copy of an iterator. However there are certain cases where a copy is wanted. In these cases there three major options:

• itertools.tee()

• pickle

• copy.deepcopy()

Given these three possibilities the iterators in iteration_utilities should never be copied using copy.copy(). Generally itertools.tee() should be the best choice except when file serialization is needed. Then pickle should be used. The copy.deepcopy() function makes, like pickle a deep (recursive) copy but doesn’t support file serialization.

A simple example using itertools.tee():

>>> from iteration_utilities import roundrobin
>>> from itertools import tee
>>> # Create an iterator
>>> a = roundrobin([1, 2, 3], [4, 5], [6])
>>> # Create a new iterator
>>> # Overwrite "a" because the input for "tee" shouldn't be reused after
>>> # the call to "tee".
>>> a, b = tee(a, 2)
>>> next(a)
1
>>> next(b)
1