Question

For a project I’m working on I am processing lots of deeply nested dict-and-list data structures. Often I find myself doing a lookup that I expect to return a list with a single member. Lookups may fail outright, or simply return zero results, so I could write code like this:

try:
    value_I_need = lookup_results[0]
except IndexError:
    # handle lookup failure, keep going

But lately I’ve been wondering if it would be more future-proof to write a loop, initially assuming it only happens zero or one times:

value_I_need = None
for value_I_need in lookup_results:
    break
if value_I_need is None:
    # handle lookup failure, keep going

(I say “future-proof” because I may want to re-write or generalize this code to handle lookups with more than one result in them.) Is there anything inherently unpythonic, wrong or slower about either of these two approaches?

Answer 1

The builtin function next() does exactly that:

a = ()
b = [1, 2, 3]
print next(iter(a), "empty") # prints "empty", as a doesn't give any values.
print next(iter(b), "empty") # prints 1, the 1st value of b