The python.make file is a Makefile module for python development that use
GNU make to build a local python development environment using
virtualenv, and simplify pip package management from
outside.
Use the -f option to load and use python.make:
$ make -f python.make
Usage: make [ TARGET ... ]
These python-* make targets help you manage a python development environment
in local via virtualenv(1) and pip(1) inside it.
python-help - show this help message
python-runtime - bootstrap a virtualenv runtime environment
python-destroy - destroy the virtualenv runtime environment
python-pip - run pip(1) command inside the virtualenv runtime
python-shell - enter python shell inside the virtualenv runtime
python-exec - run shell commands inside the virtualenv runtime
python-% - install the pip module which called %
python-freeze - save installed pip modules in requirements.txt
python-module M - generate boilerplate files for new module M
The virtualenv (v1.11.4) runtime environment is located at this path:
/path/to/current/folder/runtime
Or you can write your main makefile like following to merge python.make make
targets in:
.PHONY: all
all: python-help
...
include python.make
You could even write the makefile as following to help user bootstrap build
environment if you wish not check python.make into your source repository:
.PHONY: all
all:
ifneq (,$(wildcard python.make))
include python.make
else
ifeq (,$(findstring python,$(MAKECMDGOALS)))
$(info Please run `make python` to bootstrap python.make.)
endif
.PHONY: python
python:
wget --content-disposition https://bit.ly/python-make
endif
Copy the python.make file into the top-level folder of your project.
Or download with curl directly:
curl -L -o python.make https://bit.ly/python-make
Or with wget:
wget --content-disposition https://bit.ly/python-make
Assumes the current folder is ~/work and have merged python.make make
targets in.
The make targets provided by python.make are prefixed with python- so you
can use them easily with auto-completion:
$ make python-<TAB>
python-% python-exec python-help python-pip python-shell
python-destroy python-freeze python-module python-runtime
Show help message of python.make:
$ make python-help
...
Bootstrap your development environment in local:
$ make python-runtime
mkdir -p ~/work/runtime;
python ~/work/.cache/virtualenv/virtualenv.py ~/work/runtime;
New python executable in ~/work/runtime/bin/python
Installing setuptools, pip...done.
If there is Python 3 in your base environment and the python executable is
python3, you may build the runtime with it:
$ make VIRTUALENV_OPTS="--python=python3" python-runtime
mkdir -p /vagrant/python-make/runtime;
python /vagrant/python-make/.cache/virtualenv/virtualenv.py --python=python3 /vagrant/python-make/runtime;
Running virtualenv with interpreter /usr/bin/python3
Using base prefix '/usr'
New python executable in ~/work/runtime/bin/python3
Also creating executable in ~/work/runtime/bin/python
Installing setuptools, pip...done.
By default the base python environment in the created runtime will be the interpreter that virtualenv was invoked with. So you can do this instead:
$ make VIRTUALENV_RUN_BY=python3 python-runtime
mkdir -p /vagrant/python-make/runtime;
python3 /vagrant/python-make/.cache/virtualenv/virtualenv.py /vagrant/python-make/runtime;
Using base prefix '/usr'
New python executable in ~/work/runtime/bin/python3
Also creating executable in ~/work/runtime/bin/python
Installing setuptools, pip...done.
With this tweak, you can build the runtime with customized python version other than the system one:
$ python --version
Python 2.7.6
$ python3 --version
Python 3.4.0
$ source runtime/bin/activate
(runtime)$ python --version
Python 3.4.0
However, you will have to maintain the custom python base runtime by your self.
Search pip modules in virtualenv:
$ make python-pip search ralc
. ~/work/runtime/bin/activate; pip search ralc;
ColanderAlchemy - Autogenerate Colander schemas based on SQLAlchemy models.
python-mistralclient - Mistral Client Library
ralc - Rate Calculator.
referral_candy - ReferralCandy Python API Client
List installed pip modules in virtualenv:
$ make python-pip list
. ~/work/runtime/bin/activate; pip list;
pip (1.5.4)
setuptools (2.2)
wsgiref (0.1.2)
Install any pip module you want in virtualenv, with prefix python- added in front of package name:
$ make python-ralc
. ~/work/runtime/bin/activate; \
pip install --download-cache ~/work/.cache/pip ralc;
Downloading/unpacking ralc
Using download cache from ~/work/.cache/pip/https%3A%2F%2Fpypi.python.org%2Fpackages%2Fsource%2Fr%2Fralc%2Fralc-0.1.tar.gz
Running setup.py (path:~/work/runtime/build/ralc/setup.py) egg_info for package ralc
Installing collected packages: ralc
Running setup.py install for ralc
Installing ralc script to ~/work/runtime/bin
Successfully installed ralc
Cleaning up...
Run any command in virtualenv while using -- to avoid options captured by make:
$ make python-exec -- ralc 30 20
. ~/work/runtime/bin/activate; ralc 30 20;
> 30.00 x 20.00 = 600.00
Run python interactively with all modules you installed in local:
$ make python-shell
. ~/work/runtime/bin/activate; python;
Python 2.7.5 (default, Mar 9 2014, 22:15:05)
[GCC 4.2.1 Compatible Apple LLVM 5.0 (clang-500.0.68)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>>
Take a snapshot of current dependencies in requirements.txt:
$ make python-freeze
. ~/work/runtime/bin/activate; pip freeze > requirements.txt;
Create new module foo in current folder with module skeleton generated:
$ make python-module foo \
version=1.2.3 \
author='Jeff Hung' email=name@example.com \
url=https://github.com/jeffhung/python.make \
summary='one line summary to describe this new module' \
description='multiline description of this new module' \
license=mit copyright='blah~'
. ~/work/runtime/bin/activate; pip freeze > requirements.txt;
Making new python module 'foo'...
mkdir -p foo;
Generate foo/__init__.py
Generate setup.py
Generate MANIFEST.in
Generate LICENSE
The version, author, email, url, summary, description, and
license parameters are optional. Please specify them to override the default
values.
By default the author, email, and url parameters will be automatically
discovered from your git and github configurations. The author will be the
value of your user.name git configuration. The email will be the value of
your user.email git configuration. And url will be a github repository url
composed by your github username and name of this new python module.
By default license is empty. If license is empty, the new python module is
considered a private work that all rights are reserved to the author (auto
discovered or specified by the author parameter). Otherwise, a LICENSE file
will be generated using the lice tool. That means the valid value of
license parameter could be any of the followings: agpl3, apache, bsd2,
bsd3, cddl, cc0, epl, gpl2, gpl3, lgpl, mit, or mpl.
Destory and wipe out the local virtualenv runtime environment:
make python-destroy
rm -rf ~/work/runtime;
See also: