Path: csiph.com!news.mixmin.net!weretis.net!feeder4.news.weretis.net!ecngs!testfeeder.ecngs.de!newsfeed.freenet.ag!newsfeed.kamp.net!newsfeed.kamp.net!fu-berlin.de!uni-berlin.de!not-for-mail
From: "David N. Mashburn" <david.n.mashburn@gmail.com>
Newsgroups: comp.lang.python.announce
Subject: ANN: Cpyx 0.2
Date: Thu, 10 Dec 2015 01:12:37 -0500
Lines: 72
Approved: python-announce-list@python.org
Message-ID: <mailman.108.1449738219.12405.python-announce-list@python.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.4.0
Precedence: list
Xref: csiph.com comp.lang.python.announce:1951

Announcing Cpyx version 0.2

Github: https://github.com/davidmashburn/cpyx
Bitbucket: https://bitbucket.org/davidmashburn/cpyx
PyPI: https://pypi.python.org/pypi/Cpyx

Cpyx is a utility to automatically build C and Cython code using either 
distutils or command line cython and gcc. Numpy is supported in both cases.

The two core functions are cc and cpyx which compile c sources and 
cython+c sources respecively.

* "cc" builds a shared library from all included inputs, using path 
resolution and quoting to aid in debugging issues (gcc commands should 
run verbatim from any directory).

* "cpyx" builds multiple c and cython sources into a python extension 
module.

Cpyx is quite simple (the main two functions are only ~100 lines of 
code), so it is easy to fiddle with.

Example usage:

import Cpyx
Cpyx.cpyx('some_file.pyx', ['one_file.c', 'another_file.c'])

Complete examples can be found here:
https://github.com/davidmashburn/Cpyx/tree/master/examples


NEW IN VERSION 0.2:
*All functions were significantly overhauled and function names were 
changed* (old Cdll and Cpyx/CpyxLib became "cc" and "cpyx" respectively).
Support for inlined cython code has been dropped since cython now 
supports this automatically using "cython.inline".

New helpful utilities for extracting C function information (i.e. to 
generate code using templates):
  * get_function_raw_arguments -- Get all arguments as a single string 
(everything inside the parentheses)
  * get_function_args -- Get a string for each argument
  * get_function_types_and_variables -- Get a list of [type, variable] 
for each argument
  * rebuild_function_signature -- Get a well-formatted version of the 
full function signature


Ideally, Cpyx should work anywhere gcc, cython, and python/numpy are 
installed.
The 0.1 branch was relatively well-tested and worked on Windows 
(XP/Vista/7), Linux (Ubuntu), and Mac OS X (Leopard/Snow Leopard). The 
0.2 branch has been pretty thoroughly tested on Ubuntu Linux, but let me 
know if it does not work for you on any OS.
distutils may or may not work on Windows with mingw

I know there are other tools out there (most notably pyximport and 
sage's misc/cython.py) to do this same kind of thing (and in much cooler 
ways).
But Cpyx scratches my itch, so I thought I'd put it out there. Happy 
Cythoning!


This project supports work that I did at Vanderbilt University and am doing
at Pluralsight, LLC.

Cheers!
-David Mashburn

<P><A HREF="https://github.com/davidmashburn/cpyx">Cpyx 0.2</A> - A 
gcc/cython/distutils wrapper for compiling Cython and C code directly 
from python. (10-Dec-15)