Metadata-Version: 2.0
Name: dstack-factory
Version: 1.0.7
Summary: CLI that accompanies dstack-factory for building and publishing docker images for python packages.
Home-page: https://github.com/obitec/dstack-factory
Author: JR Minnaar
Author-email: jr.minnaar+dstack@gmail.com
License: MIT License
Keywords: docker python wheels images runtime automation
Platform: UNKNOWN
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: System Administrators
Classifier: Topic :: Software Development :: Build Tools
Classifier: Topic :: Software Development :: Code Generators
Classifier: Topic :: System :: Installation/Setup
Classifier: Topic :: System :: Software Distribution
Classifier: Operating System :: OS Independent
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.6
Requires-Dist: invoke
Requires-Dist: python-dotenv (>=0.5.1)
Requires-Dist: setuptools-scm
Provides-Extra: dev
Requires-Dist: Sphinx (>=1.4.1); extra == 'dev'
Requires-Dist: invoke (>=0.13.0); extra == 'dev'
Requires-Dist: python-dotenv (>=0.5.1); extra == 'dev'
Requires-Dist: setuptools-scm; extra == 'dev'
Requires-Dist: wheel (>=0.29.0); extra == 'dev'
Provides-Extra: test
Requires-Dist: coverage; extra == 'test'

dstack-factory
==============

dstack-factory consists of two components: (1) ``factory`` and (2) ``runtime``.

``factory`` is a service that builds (or downloads) python wheels from a pip requirements file or from a wheel package uploaded to the wheelhouse. If a python package is already available as wheel package, it is cached. Otherwise the source is downloaded from pip, git or any other source supported by pip and compiled/packaged as a wheel file.

Examples of python packages that currently still need to be build are:
   - psycopg2
   - lxml
   - pycrypto
   - pillow
   - weasyprint (and it's dependencies)
   - anything hosted on GitHub, e.g. forks of python packages or unreleased versions

``runtime`` is a base docker image that contains all the necessary libraries to run most popular python packages with external dependencies. For example pandas, matplotlib, weasyprint, etc. all require non-python libraries.
In addition to the base runtime docker image, ``dstack-factory`` also provides three entry point Dockerfiles to support different workflows for building and deploying docker and python based applications.


Documentation
-------------

The documentation can be found at: https://pythonhosted.org/dstack-factory/


