SciKits

Quick search

scikit-ci

version 0.15.0

scikit-ci enables a centralized and simpler CI configuration for Python extensions.

Download: https://pypi.python.org/pypi/scikit-ci
Homepage: https://github.com/scikit-build/scikit-ci
PyPI: http://pypi.python.org/pypi/scikit-ci
People: The scikit-build team

Description

Documentation Status https://ci.appveyor.com/api/projects/status/5to6lvgaqcrck675?svg=true https://circleci.com/gh/scikit-build/scikit-ci/tree/master.svg?style=svg https://img.shields.io/travis/scikit-build/scikit-ci.svg?maxAge=2592000 https://codecov.io/gh/scikit-build/scikit-ci/branch/master/graph/badge.svg

scikit-ci enables a centralized and simpler CI configuration for Python extensions.

By having appveyor.yml, circle.yml and .travis.yml calling the same scikit-ci command-line executable, all the CI steps for all service can be fully described in one scikit-ci.yml file.

History

scikit-ci was initially developed in May 2016 by Omar Padron to facilitate the continuous integration of the scikit-build project.

At that time, it already consisted of a driver script calling methods specific to each continuous integration service. By having each CI service calling the same driver script, there was no need to deal with implementing install/test/build steps over and over in different scripting languages (power shell, shell or windows batch). Instead all code was implemented in python code leveraging the subprocess module.

Later in early September 2016, with the desire to setup cross-platform continuous integration for other project and avoid duplication or maintenance hell, a dedicated repository was created by Jean-Christophe Fillion-Robin. By simply cloning the repository, it was possible to more easily enable CI for other projects.

While this was an improvement, all the steps were still hardcoded in the driver scripts, the project was not easily customizable. More could be done to improve the user experience.

Finally, in late September 2016, all hardcoded code was moved into standalone executable python scripts. Then, Jean-Christophe came up with the concept of scikit-ci.yml configuration file. This configuration file allows to describe the commands and environment for each step (install, test and build) specific to a project and associated continuous integration services.

Installation

PyPI

You can download the latest distribution from PyPI here: http://pypi.python.org/pypi/scikit-ci

Using pip

You can install scikit-ci for yourself from the terminal by running:

pip install --user scikit-ci

If you want to install it for all users on your machine, do:

pip install scikit-ci
On Linux, do sudo pip install scikit-ci.

If you don't yet have the pip tool, you can get it following these instructions.

This package was discovered in PyPI.