jayvdb/pundler

Python bundler-alike alternative to virtualenv

Pundle

Changelog

• PUNDLEENV can include comma separated envs
• Pipfile initial support. Only strings as versions now. Do not calculates hashes and do not use it yet.
• New setup.py support with mocking of setuptools.setup
• Added python shell try package feature. To use it use pundle.use("package_name==0.1"). Version is optional.
• Added environments support. To use it just make files like requirement.txt, requirements_dev.txt, requirements_test.txt. To activate env use like PUNDLEENV=dev pundle ...
• Added VCS support for urls like git+https://github.com/mitsuhiko/jinja2.git@85820fceb83569df62fa5e6b9b0f2f76b7c6a3cf#egg=jinja2-2.8.0. Push exactly like this formatted str to requirements.txt
• Added initial support for setup.py requirements. Helpful for package development.

What is it all about?

Pundle get rid of virtualenv, because I think that virtualenv is pile of garbage and we must get rid of it.

• Pundle installs all packages and their versions to special folder. And mount pinned, frozen versions on activate step.
• After that you program will use exactly this versions that were pinned in frozen.txt.
• If you change branch or edit requirements.txt or frozen.txt, pundle will note you about you need make install new packages or freeze newly added packages. It will not let you use packages that have not bin pinned. You will never fall in situation where you test old version of package.

Why not Pipenv, I heard it is "for humans"?

I don't think that anything that is based on virtualenv can be "for humans".

How to

Install:

> pip install pundle

or just place pundle.py where python can find it

Create requirements.txt or setup.py (we will support .toml a bit later). You can pin versions if you need it, or just place package name. Pundle will pin it anyway as well as all of it dependencies.

Reveal all dependencies, pin versions, download and install everything:

> python -m pundle

Where actually it will install? Pundle use special folder .pundledir/python-version/package-name-version for every seperate package and version.

To make it short create alias:

alias pundle='/usr/bin/env python -m pundle'
pundle install

After packages install, frozen/pinned, we want to use them, you know, import, right?

import pundle; pundle.activate()

Or we can try to use pundle features:

# to execute entry point
pundle exec some_package_entry_point
# to run python script
pundle run my_script.py
# run module like python -m
pundle module some.my.module

To add VCS to requirements.txt use git+url#egg=my_package-0.1.11 form.

Pundle console

To start console with Pundle activated use

> pundle console [ipython|ptpython|bpython]

You will have pundle_suite object inserted to environment. You can use it to call pundle_suite.use("trafaret_schema") for example.

Python shell usage

You can use pundle to expirement in python shell:

>>> import pundle
>>> pundle.use('django==1.11.1')  # will download and install django
>>> import django

Or you can use it in script:

>>> import pundle
>>> pundle.use('django')
>>> pundle.use('arrow')
>>> pundle.use('trafaret')
>>>
>>> import django
>>> import arrow
>>> import trafaret

Environments

Pundle support environments. You can create seperate requirements file with suffix like requirements_dev.txt. Pundle will create frozen_dev.txt that will track common requirements + dev requirements.

To use dev environment use PUNDLEENV=dev environment variable:

bash> PUNDLEENV=dev pundle run myscript.py

or common usage:

bash> PUNDLEENV=test pundle exec pytest

For setup.py file pundle uses extras_require as environments. For example if you have extras_require = {'test': ['pylint', 'pyflakes']} then you can use pylint with PUNDLEENV=test pundle exec pylint.

More usage info

Upgrade package:

pundle upgrade django

Upgrade all packages:

pundle upgrade

List of all entry points:

pundle entry_points

Do not hesitate to pundle help ;)

Howto

Q: How to use custom index url or extra index?

A: use PIP_EXTRA_INDEX_URL or any other pip environment variables.