NIP - nip isn't pip

A fast way to start and use a python project within a virtual environment.

Commands

init - Initialize a new Python project in the current folder.

install - Install all project dependencies and initialize a new venv if necessary.

add - Install a new project dependency and add it to your nipfile. (like pip install)

remove - Remove a dependency from your project. (like pip uninstall)

run - Run a script from the nipfile

Installation

Install nip from PyPi

$ pip install nip.cli

Or get the latest development version from GitHub

$ pip install git+https://github.com/RopePy/nip

Quick Start

Enter an empty directory, type nip init and answer the questions. this starts up a virtual environment for you under ./python_modules. You may also notice NIP has created these files in the cwd..

• .gitignore with the new python_modules included so you don't have to.
• nip.json file with you package details.
• setup.py file with a basic templated layout.

Add a Dependency

Run nip add <package_name> to install a new dependency. Nip creates a requirements.txt file and pins the version automatically.

If it's a development dependency the use the option -D, ex: nip add -D <package_name> which creates a dev_dependency.txt file instead

PROTIP: Any binaries included with the added package also become available for use through nip scripts .. (see below)

Scripts

You can place shortcuts to commands in the nipfile under a scripts section (see example), which you can then run with nip run <script_name> or just nip <script_name> and they will be started within the virtual environment. You can also run any binaries installed to the python_modules/bin folder.

For example, if you have tox as a dependency, but not installed globally you can run it as a script, ex:

nip.json ..

{
    ...
    "scripts": {
        "start": "python src/main.py",
        "db:migrate": "python manage.py migrate",
        ...
+       "test": "tox"
    },
    "dev_dependencies": {
+       "tox": "~=2.9.1",
        ...
    }
}

$ nip run test       // or just `nip test`

Existing Projects

Enter a directory with a nipfile.json and type nip install to install any package requirements/virtual environments necessary for the project and you're ready to go.

Example

    $ nip init
    nip - v0.2.0 - Init - Made with Love, by Duroktar © 2018

    Package Name (nip-test):
    Author: Duroktar
    Version (0.1.0):
    License (MIT):
    ⠼ Initialising.... Done.
    Finished.

nip.json

    {
        "name": "nip-test",
        "author": "Duroktar",
        "version": "0.1.0",
        "license": "MIT",
+       "scripts": {
+           "start": "python hello_world.py"
+       }
    }

    $ echo 'print("Hello world!!")' >> hello_world.py

    $ nip start
    nip - v0.2.0 - Run - Made with Love, by Duroktar © 2018

    Hello world!!

Why?

Starting a new Python project from scratch isn't hard, but you end up using similar patterns and multiple tools each time. I wanted to script away as much of that as possible without coupling everything together into something that ended up being complicated and unmanageable.

At its core, NIP is a service based framework with a middleware and hooks based strategy. It was built to take advantage of existing Python tools instead of trying to replace them. It uses pip under the hood, so if you want stop using nip just remove the nip.json file and you're left with a standard python package.

Development

I tried to make nip really easy to understand and work on, there's a useful guide here in the services section that tries to give a basic rundown of what happens when nip runs, and serves as a decent starting point for those interested in helping out.

Third Party Packages

NIP was heavily inspired by NPM/Yarn and the FeathersJS Framework.

• Click - http://click.pocoo.org/5/ - three-clause BSD License
• PySpin - https://pypi.python.org/pypi/pyspin - MIT
• Tox - https://tox.readthedocs.io/en/latest/ - MIT
• pytest - https://docs.pytest.org/en/latest/ - MIT
• flake8 - http://flake8.pycqa.org/en/latest/ - MIT
• venv_tools - ... - ...

License

MIT