making a commandline alias to a python file in a p

问题:

I've been making a python project using pipenv, and I want to be able to run it in a terminal from any location on my (linux) system. Specifically, say I have the following directory structure:

/home
  /project
    Pipfile
    main.py
  /other_dir

I would like to be able to make an alias that allows me to call main.py like so:

 /home/other_dir$ alias_to_my_proyect --some args

and run it in the virtual env, having the same behaviour as

/home/project$ pipenv run python main.py

But in another directory.

If it weren't a pipenv project, I'd just use a shebang a the start of the file and then add an alias to it in my .bashrc, but I want to use pipenv's virtual environment, but I cant find a way to do this with pipenv.

回答1:

If you want to use a specific python environment for your script you will need to point it to the interpreter of that environment. On Mac the default is that pipenv installs all virtualenvs to /Users/<user_name>/.local/share/virtualenvs/ however that can be set to different locations as described in the manual:

Pipenv automatically honors the WORKON_HOME environment variable, if you have it set — so you can tell pipenv to store your virtual environments wherever you want, e.g.:

export WORKON_HOME=~/.venvs

In addition, you can also have Pipenv stick the virtualenv in project/.venv by setting the PIPENV_VENV_IN_PROJECT environment variable.

You can find out where the exact location of the virtualenv is with pipenv --venv inside your project folder. It returns something like /Users/reedef/.local/share/virtualenvs/project-BpR9WgCa. The interpreter is in ./bin/python of that location.

If we assume that you did not set any environment variable and you are using Mac than that means that you can write a script:

#!/usr/bin/env sh
/Users/reedef/.local/share/virtualenvs/project-BpR9WgCa/bin/python /home/project/main.py

and place it somewhere in your $PATH, e.g. /usr/local/bin/my_fancy_main to let it run in that specific environment.

Note: as mentioned by @Jon in the comments, -BpR9WgCa at the end of the path is stable as it is made from the project path:

hash = hashlib.sha256(location.encode()).digest()[:6]

It should be the same as long as the project path hasn't changed.

回答2:

You should use the standard setuptools library to write a setup.py file. In particular you can write an entry_points section that names your main script:

entry_points={
  'console_scripts': [
    'alias_to_my_project = project.main.main'
  ]
}

Once you've done this, you can activate and install your package into your virtual environment

pipenv install -e .

# or without pipenv
. ~/vpy/bin/activate
pip install -e .

This will create a wrapper script in $VIRTUAL_ENV/bin/alias_to_my_project that loads the project.main Python module and calls its main function.

The wrapper script knows about the virtual environment and can be called directly without specifically activating the virtual environment. So you can do something like

ln -s $VIRTUAL_ENV/bin/alias_to_my_project $HOME/bin/alias_to_my_project
PATH=$HOME/bin:$PATH

and it will always be available.

回答3:

You can just use

#!/usr/bin/env pipenv-shebang

in your script after you install my pipenv-shebang package:

pip install pipenv-shebang