The documentation for the csv.writer.writerows method says that it expects "a list of row objects", when it really will accept any iterable that yields rows (such as a generator). While it's often nice for code to be more accepting than the documented requirements, I think the docs in this case should state that writerows() expects an iterable, rather than misinforming users that a list is required.

This documentation issue was brought up in a Stack Overflow question here: http://stackoverflow.com/questions/28636848/csv-writer-writerows-takes-iterator

I expect the necessary documentation patch will be pretty trivial, and if nobody else gets to it first, I will try to provide one when I have enough time to update my cpython checkout (not soon, alas).