Iterative development using the Python interpreter

Because Python is an interpreted language, a productive coding style is to write relatively self-contained functions while repeatedly testing them during development using the interactive interpreter. This is a form of unit testing, and you could almost call this Test-driven development if you copy your Python session into your docstrings using the doctest module. (But you'd be writing the tests after writing the code, which wouldn't be considered TDD by strict adherants.)

This following method might seem obvious to people with a lot of experience in Python, but I've never seen it written down anywhere. An easy way to establish this workflow is to use the reload function to reload your Python script as a module as you make changes in your text editor.

First, you start up a Python session and import your Python script as a module using the import statement. In this example, my script is called test.py.

$ python
Python 2.7.3 (default, Mar 18 2014, 05:13:23)
[GCC 4.6.3] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>>import test

Now you can make changes to test.py in your text editor and test them in your Python session.

test.py
def example(*args): return map(lambda x: x + 1, args)

>>>reload(test).example(1, 2, 3)
[2, 3, 4]

After calling your function the first time, you can re-test with the same arguments by pressing ↑ to pull the reload statement from the command history in the interpreter.

test.py
def example(*args): return map(lambda x: x + 2, args)

>>>reload(test).example(1, 2, 3)
[3, 4, 5]