Virtual environments with virtualenvwrapper
virtualenvwrapper is a set of shell functions built on top of the
virtualenv python module, and make it even easier to create and manage python virtual environments.
The documentation for this project is quite good, but here I wanted to write a reminder (mostly for me) about the configuration of this tool on a Linux server.
You can install
pip install virtualenvwrapper
Then you will need to create a (hidden) folder in your
home directory and call it
Lastly, you will need to configure your terminal to execute
virtualenvwrapper commands. This is done by adding 2 lines of code to a bash configuration file.
Here’s the **catch:
If you are working on your computer you have to add these 2 lines to your
# ~/.bashrc (your local machine) export WORKON_HOME=$HOME/.virtualenvs source /usr/local/bin/virtualenvwrapper.sh
If you are working on a remote server via SSH you have to add the very same 2 lines, but in a different file, the
# ~/.bash_profile (your linux server) export WORKON_HOME=$HOME/.virtualenvs source /usr/local/bin/virtualenvwrapper.sh
Thanks to this answer on askubuntu I found out that this difference is due to the different access modality:
- on your local machine you are accessing the console in interactive, non-login mode, and the
~/.bashrcfile will be sourced;
- on a remote server (e.g. a DigitalOcean droplet) via SSH you are accessing the console in interactive, login mode, and the
~/.bash_profilefile will be sourced.
If you want to know more about these bash files, see here.
If you are already using
virtualenv you will probably know the
source bin/activate command. This is a hook that sets an environment variable called
VIRTUAL_ENV, another one called
PYTHON_HOME and a few others. The command
deactivate is another hook that unset the same environment variables previously set.
virtualenvwrapper defines some additional hooks, like
predeactivate. You can use these hooks to set additional environment variables, set aliases, ect. See here for some examples that you might find useful.
Here are the most common
mkvirtualenv YOUR_VIRTUALENV # create virtual environment (and activate it) mkvirtualenv YOUR_VIRTUALENV --python=python3.5 # create virtual enviroment and specifiy the python version workon YOUR_VIRTUALENV # activate virtual environment rmvirtualenv YOUR_VIRTUALENV # remove virtual environment deactivate # deactivate current active virtual environment