Skip to content

jsirois/pex

 
 

Repository files navigation

PEX

image

image

image

image

image

Contents

Overview

pex is a library for generating .pex (Python EXecutable) files which are executable Python environments in the spirit of virtualenvs. pex is an expansion upon the ideas outlined in PEP 441 and makes the deployment of Python applications as simple as cp. pex files may even include multiple platform-specific Python distributions, meaning that a single pex file can be portable across Linux and OS X.

pex files can be built using the pex tool. Build systems such as Pants, Buck, and {py}gradle also support building .pex files directly.

Still unsure about what pex does or how it works? Watch this quick lightning talk: WTF is PEX?.

pex is licensed under the Apache2 license.

Installation

To install pex, simply

$ pip install pex

You can also build pex in a git clone using tox:

$ tox -e package
$ cp dist/pex ~/bin

This builds a pex binary in dist/pex that can be copied onto your $PATH. The advantage to this approach is that it keeps your Python environment as empty as possible and is more in-line with what pex does philosophically.

Simple Examples

Launch an interpreter with requests, flask and psutil in the environment:

$ pex requests flask 'psutil>2,<3'

Save Dependencies From Pip

Or instead freeze your current virtualenv via requirements.txt and execute it anywhere:

$ pex $(pip freeze) -o my_virtualenv.pex
$ deactivate
$ ./my_virtualenv.pex

Ephemeral Environments

Run webserver.py in an environment containing flask as a quick way to experiment:

$ pex flask -- webserver.py

Launch Sphinx in an ephemeral pex environment using the Sphinx entry point sphinx:main:

$ pex sphinx -e sphinx:main -- --help

Using Entry Points

Projects specifying a console_scripts entry point in their configuration can build standalone executables for those entry points.

To build a standalone pex-tools-executable.pex binary that runs the pex-tools console script found in all pex version 2.1.35 and newer distributions:

$ pex "pex>=2.1.35" --console-script pex-tools --output-file pex-tools-executable.pex

Specifying A Specific Interpreter

You can also build pex files that use a specific interpreter type:

$ pex "pex>=2.1.35" -c pex-tools --python=pypy -o pex-tools-pypy-executable.pex

Most pex options compose well with one another, so the above commands can be mixed and matched, and equivalent short options are available.

For a full list of options, just type pex --help.

Integrating pex into your workflow

If you use tox (and you should!), a simple way to integrate pex into your workflow is to add a packaging test environment to your tox.ini:

[testenv:package]
deps = pex
commands = pex . -o dist/app.pex

Then tox -e package will produce a relocatable copy of your application that you can copy to staging or production environments.

Documentation

More documentation about Pex, building .pex files, and how .pex files work is available at https://docs.pex-tool.org.

Development

Pex uses tox for test and development automation. To run the test suite, just invoke tox:

$ tox

If you don't have tox, you can generate a pex of tox:

$ pex tox -c tox -o ~/bin/tox

Tox provides many useful commands and options, explained at https://tox.wiki/en/latest/ . Below, we provide some of the most commonly used commands used when working on Pex, but the docs are worth acquainting yourself with to better understand how Tox works and how to do more advanced commands.

To run a specific environment, identify the name of the environment you'd like to invoke by running tox --listenvs-all, then invoke like this:

$ tox -e fmt

To run MyPy:

$ tox -e check

All of our tox test environments allow passthrough arguments, which can be helpful to run specific tests:

$ tox -e py37-integration -- -k test_reproducible_build

To run Pex from source, rather than through what is on your PATH, invoke via Python:

$ python -m pex

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 99.7%
  • Other 0.3%