Set specific environment variables activating a Python virtual environment

When using Python virtual environments it can be useful to have certain environment variables automatically set when we activate it. For example we can have a projects that needs to connect to a database or we may need specific settings to run a service. Automating this process makes sure we don't forget about setting everything we need.

Storing the configuration

I strongly suggest to store all the configuration in the root of your project in a file named .env which you need to take care of adding to your .gitignore.

An example of .env file:

MY_USERNAME=user1
MY_PASSWORD=secret
MY_SERVICE_URL=http://localhost:5000

How to activate the configuration

Python virtual environments have a few scripts that are beign executed in certain situations. In particular there is one named postactivate that is being run after you activate trhe virtual environment.

You need to first find the folder where it's located and to do this, you need to first activate your virtual environment. Now you need to run cdvirtualenv in your terminal and your working directory will be set to where the virtual environment is located.

At this point edit postactivate file with (you can replace nano with your preferred editor) nano bin/postactivate and add the following script:

#!/bin/zsh
# This hook is sourced after this virtualenv is activated.
set -a
source .env
set +a

Note: I'm using zsh here but you can replace it with bash (the syntax should be the same).

Test the configuration

To make the configuration effective you need to deactivate your virtual environment and activate it again. The postactivate script will run automatically and you can now check that your environment variables have been set correctly (the ... have been placed to replace other existing values):

(test-env) ➜ env
...
MY_USERNAME=user1
MY_PASSWORD=secret
MY_SERVICE_URL=http://localhost:5000

As you can see the values configured in the .env file have been correctly set.