# This file is the fixed prompt for `snodo models --benchmark`.
#
# The benchmark's whole point is that every run sends the exact same text, so
# a number from one run can be compared with a number from another. Changing
# the prompt below makes measurements taken before the change incomparable with
# measurements taken after it. If you change it, say why in the commit.
#
# Lines in this header are not sent to the model. Everything after the marker
# line is the prompt.
#--- benchmark prompt below this line ---
Write a Python implementation of a FIFO queue.

Requirements:
- A class named FifoQueue with enqueue(item) and dequeue() methods.
- dequeue() raises IndexError when the queue is empty.
- Include a length method and an is_empty method.
- Add a short docstring to the class.
- Return only the code, with no explanation or prose.
