nameName of your distribution.
versionVersion string of your distribution.
packagesList of Python packages (that is, directories containing modules) to include. This can be specified manually, but a call to setuptools.find_packages() is typically used instead.
py_modulesList of top-level Python modules (that is, single .py files) to include.


For further information on python packaging see:


For writing official packages there is a packaging user guide.

Purpose of

The setup script is the centre of all activity in building, distributing, and installing modules using the Distutils. It's purpose is the correct installation of the software.

If all you want to do is distribute a module called foo, contained in a file, then your setup script can be as simple as this:

from distutils.core import setup


To create a source distribution for this module, you would create a setup script,, containing the above code, and run this command from a terminal:

python sdist

sdist will create an archive file (e.g., tarball on Unix, ZIP file on Windows) containing your setup script, and your module The archive file will be named foo-1.0.tar.gz (or .zip), and will unpack into a directory foo-1.0.

If an end-user wishes to install your foo module, all she has to do is download foo-1.0.tar.gz (or .zip), unpack it, and—from the foo-1.0 directory—run

python install

Adding command line scripts to your python package

Command line scripts inside python packages are common. You can organise your package in such a way that when a user installs the package, the script will be available on their path.

If you had the greetings package which had the command line script


You could run that script by running:

python greetings/greetings/

However if you would like to run it like so:

You can achieve this by adding scripts to your setup() in like this:

from setuptools import setup

When you install the greetings package now, will be added to your path.

Another possibility would be to add an entry point:

entry_points={'console_scripts': ['greetings=greetings.hello_world:main']}

This way you just have to run it like:


Adding installation options

As seen in previous examples, basic use of this script is:

python install

But there is even more options, like installing the package and have the possibility to change the code and test it without having to re-install it. This is done using:

python develop

If you want to perform specific actions like compiling a Sphinx documentation or building fortran code, you can create your own option like this:

cmdclasses = dict()

class BuildSphinx(Command):

    """Build Sphinx documentation."""

    description = 'Build Sphinx documentation'
    user_options = []

    def initialize_options(self):

    def finalize_options(self):

    def run(self):
        import sphinx
        sphinx.build_main(['', '-b', 'html', './doc', './doc/_build/html'])
        sphinx.build_main(['', '-b', 'man', './doc', './doc/_build/man'])

cmdclasses['build_sphinx'] = BuildSphinx


initialize_options and finalize_options will be executed before and after the run function as their names suggests it.

After that, you will be able to call your option:

python build_sphinx

Using source control metadata in

setuptools_scm is an officially-blessed package that can use Git or Mercurial metadata to determine the version number of your package, and find Python packages and package data to include in it.

from setuptools import setup, find_packages


This example uses both features; to only use SCM metadata for the version, replace the call to find_packages() with your manual package list, or to only use the package finder, remove use_scm_version=True.