Why virtual environments while using Python?
Python versions or packages change over time and this could break your code, ansible playbooks, molecule tests etcera. I ran into this when I began using Ansible, where the path to the old Python executable was gone due to an upgrade on my OS level. This is the global level. Using virtualenv removes this issue due to it being set to a specific location when setting up the virtualenv.
Package install
1
$ pip install virtualenv
Installs the virtualenv package.
Commands
Create a new virtual environment. Make sure you are a proper working directory, i like to use ~/code/virtual_environments/.
1
$ virtualenv [projectname]
Enter the actual project so you can start installing packages with pip. or use source instead of . (. is source)
1
$ . [projectname]/bin/activate
1
$ source [projectname]/bin/activate
Exit the virtualenv while inside one.
1
$ deactivate
While inside the virtualenv make an export of all installed packages.
1
$ pip freeze -l > requirements.txt
Why the -l? If the virtualenv has access to the global packages it is not listing the globally installed packages.
Use the requirements.txt to setup your virtualenv.
1
$ pip install -r requirements.txt
This should get you started pretty quickly with Python virtualenv.