reynoldsnlp / pipster

pipster: The pythonic way to `pip install`.
MIT License
8 stars 2 forks source link
education install installer package package-management package-manager pip pipster python

pipster: The pythonic way to pip install

Pipster logo

| | | | --- | --- | | CI/CD | [![CI - Test and Build](https://github.com/reynoldsnlp/pipster/actions/workflows/test_build_publish.yml/badge.svg)](https://github.com/reynoldsnlp/pipster/actions/workflows/test_build_publish.yml) | | Package | [![PyPI - Version](https://img.shields.io/pypi/v/pipster.svg?logo=pypi&label=PyPI&logoColor=gold)](https://pypi.org/project/pipster/) [![PyPI - Downloads](https://img.shields.io/pypi/dm/pipster.svg?color=blue&label=Downloads&logo=pypi&logoColor=gold)](https://pypi.org/project/pipster/) [![PyPI - Python Version](https://img.shields.io/pypi/pyversions/pipster.svg?logo=python&label=Python&logoColor=gold)](https://pypi.org/project/pipster/) | | Meta | [![Hatch project](https://img.shields.io/badge/%F0%9F%A5%9A-Hatch-4051b5.svg)](https://github.com/pypa/hatch) [![linting - Ruff](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/charliermarsh/ruff/main/assets/badge/v0.json)](https://github.com/charliermarsh/ruff) [![code style - Black](https://img.shields.io/badge/code%20style-black-000000.svg)](https://github.com/psf/black) [![types - Mypy](https://img.shields.io/badge/types-Mypy-blue.svg)](https://github.com/python/mypy) [![License - MIT](https://img.shields.io/badge/license-MIT-9400d3.svg)](https://spdx.org/licenses/) [![GitHub Sponsors](https://img.shields.io/github/sponsors/reynoldsnlp?logo=GitHub%20Sponsors&style=social)](https://github.com/sponsors/reynoldsnlp) |

Intro

Install packages using pip from inside a python script or console:

>>> import pipster
>>> pipster.install("some_package")

This tool solves the age-old problem of pip installing a package only to find that you still can't import it because it was installed for the wrong interpreter. By installing from inside python, you ensure that it will be available to import from that instance of python.

We recommend using pipster.install() either in a separate python script or an interactive REPL. If using an interactive REPL, we recommend restarting python after installation.

Note: The developers of pipster hope that this tool will eventually be integrated into pip so that this functionality becomes ubiquitous. However, first pipster needs to be extensively tested and used in real life. By sharing this tool with others, and reporting bugs/issues if they arise, you are helping us achieve that goal.

Installation

To install pipster, run the following in your terminal (you may need to replace python with the name of the executable you use to run Python, such as python3, python3.11, etc.):

$ python -m pip install pipster

...and if that's not working, you can run this in a python script or console:

import subprocess
import sys
cmd = [sys.executable,  "-m", "pip", "install", "pipster"]
subprocess.check_call(cmd)

Usage

The install() function can be called in two ways.

1. Copy-paste pip install ...

If you give install() a single string that begins with pip install, then it will run that exact command as if it were given at the command line.

install("pip install some_package")`
install("pip install --user --upgrade pkg1 pkg2 pkg3")

2. Python function API

You can also pass any number of target packages along with keyword arguments corresponding to command-line options for pip install. Note that the python keyword arguments use _ instead of -.

The pipster.install() function does not validate which options are available in the command line. If you give it arguments that are not valid command-line options, then it will attempt to run pip install with those options, and pip will return an error.

The CLI options reference shows every possible command-line option and its corresponding expression as a keyword argument for install(). The keyword arguments are also shown in the docstring, which can be seen by running help(install) (type q to exit). The principles used to convert command-line options to python keyword arguments are layed out in the following sections.

Boolean options

Most boolean command-line options are set by giving <option>=True.

install("pkg", user=True)             # pip install --user pkg
install("pkg", "pkg2", upgrade=True)  # pip install --upgrade pkg pkg2

Boolean command-line options that begin with --no-, such as --no-color are set by removing the no- prefix and using <option>=False.

install("pkg", color=False)  # pip install --no-color pkg
install("pkg", deps=False)   # pip install --no-deps pkg

Key-value options

Usually, key-value options are set using a string: <option>="<value>".

install(r="requirements.txt")         # pip install -r requirements.txt
install("pkg", python_version="3.8")  # pip install --python-version 3.8 pkg

However, some key-value options can be used multiple times. In these cases, the value given should be a list of strings: <option>=["<val1>", "<val2>"].

install(r=["reqs1.txt", "reqs2.txt"])  # pip install -r reqs1.txt -r reqs2.txt

Additive options

Some command-line options are additive (notably -q and -v), and can be used up to 3 times. These can be set using either an integer or True.

install('pkg', v=True)  # pip install -v pkg
install('pkg', v=1)     # pip install -v pkg
install('pkg', v=2)     # pip install -vv pkg
install('pkg', v=3)     # pip install -vvv pkg

Omitting options

Any option that is assigned the value None is omitted from the command. This feature facilitates simpler logic in preparing keyword arguments programmatically.

user = None
if <logical test>:
  user = True
install('pkg', user=user)

Installing modules that were already imported

If you re-install, upgrade, or downgrade a package after it has already been imported, pipster will do its best to detect this and issue a warning that Python should be restarted for changes to be available.

>>> import pipster
>>> import requests
>>> requests.__version__
'2.27.0'
>>> pipster.install("requests", upgrade=True)
...
Successfully installed requests-2.28.2

WARNING: The following modules were already loaded. Restart python to see changes:
requests

Pipfiles

Pipfiles are ignored by pipster. If you use pipster from a directory that contains a Pipfile (or whose parent directories up to a depth of 3 contain a Pipfile), then pipster will issue a warning that Pipfiles are ignored.