[Python] Which is a better way to move an iterator one step forward: `for x in iterator:...

While working on a learning task involving overlapping n-wise windows from an input iterable - similar to what itertools.pairwise does - I came across code like this:

def f(seq):
it = iter(seq)
for x in it: break
for y in it:
yield x, y
x = y


The line for x in it: break was used instead of x = next(it, None) to avoid assigning x at all if seq is empty, which would be the case with the next() call.

I'm not familiar with CPython internals or low-level details, so I can't say whether this is a justified choice. Could someone help me understand the pros and cons of both approaches for advancing an iterator by one item before a loop?

Continue reading...