recur.py

# given  a list like this:
the_list = [1,2,3,4,5]

# We could call do_something() for each item in the list like this:

for li in the_list:
    do_something(li)


# This is a recursive self-calling function to do the same
def joes_loop(the_list):
    if the_list: # anything left?
       do_something(the_list[0]) # do somethnig on the first item of the list
       joes_loop(the_list[1:])   # call ourself with the rest of the list


# One will work in map-reduce, spark, parallel, the other won't....
# this is because the iterator (li) is a side-effect, a "temp variable" that is
# not thread-safe (simply put)