python-env-utils

.. image:: https://travis-ci.org/yunojuno/python-env-utils.svg?branch=master :target: https://travis-ci.org/yunojuno/python-env-utils

.. image:: https://badge.fury.io/py/python-env-utils.svg :target: https://badge.fury.io/py/python-env-utils

.. image:: https://codecov.io/gh/yunojuno/python-env-utils/branch/master/graph/badge.svg :target: https://codecov.io/gh/yunojuno/python-env-utils

env utils

This library extends the standard library's getenv function, allowing you to coerce the return value into a type.

And that's it.

It's been released as a library because every project we have includes the same requirements - read in environment variables, coerce them into the correct type.

The problem is that environment variables are always stored as strings, but Python will evaluate any string (even "") as True if cast to a boolean. This is almost never the desired behaviour. If you set an environment variable to "", "0" or "False", you want it to be False.

.. code:: python

>>> os.environ['foo'] = "0"
>>> val = os.getenv('foo')
>>> val
"0"
>>> bool(val)
True

env_utils.get_env will coerce the value into the type you require. The package contains basic helper functions that coerce booleans, integers, decimals, floats, dates, lists and dictionaries.

.. code:: python

# FOO=0
>>> env_utils.get_env('FOO')
"0"
>>> env_utils.get_bool('FOO')
False
>>> env_utils.get_int('FOO')
0

# FOO=foo,bar
>>> env_utils.get_list('FOO', separator=',')
['foo', 'bar']

# FOO='{"foo": true}'
>>> env_utils.get_dict('FOO')
{'foo': True}

# FOO=2016-11-23
>>> env_utils.get_date('FOO')
datetime.date(2016, 11, 23)

You can supply any function you like to coerce the value, e.g.:

.. code:: python

>>> import os
>>> os.getenv('FOO_NAME')
'bob'
>>> class Foo(object):
...     def __init__(self, name):
...         self.name = name
>>> coerce = lambda x: Foo(x)
>>> import env_utils
>>> foo = env_utils.get_env('FOO_NAME', coerce=coerce)
>>> foo.name
>>> 'bob'

Installation

The library is available at pypi as 'python-env-utils', and can installed using pip::

$ pip install python-env-utils

Tests

The tests can be run using tox:

.. code:: shell

$ tox