What is it?

Pythran is a python to c++ compiler for a subset of the python language. It takes a python module annotated with a few interface description and turns it into a native python module with the same interface, but (hopefully) faster.

Installation

Pythran sources are hosted on https://github.com/serge-sans-paille/pythran.

Pythran releases are hosted on http://pypi.python.org/pypi/pythran.

1. Gather dependencies:

   • Following the easy install way:

     $> PYTHONPATH=<my_prefix>/lib/python<my_version>/site-packages \
                     easy_install --prefix=<my_prefix> ply networkx
     $> sudo apt-get install libboost-python-dev libgoogle-perftools-dev libgmp-dev
     

   • Following the debian way:

     $> sudo apt-get install libboost-python-dev libgoogle-perftools-dev libgmp-dev python-ply python-networkx python-pytest
     

2. Use the install target from setup script, in source directory:

   $> python setup.py install --prefix=<my_prefix>
   

3. Write your environment in stone:

   $> export PYTHONPATH=<my_prefix>/lib/python<my_version>/site-packages
   $> export PATH=<my_prefix>/bin:$PATH
   

or

1. The debway:

   • Add serge_sans_paille's debian server to your source.list, following the instruction given in http://serge.liyun.free.fr/serge/debian.html

   • Run the classical:

     $> sudo apt-get update
     $> sudo apt-get install pythran
     

Basic Usage

A simple pythran input could be dprod.py:

#pythran export dprod(int list, int list)
def dprod(l0,l1):
        return sum(x*y for x,y in zip(l0,l1))

To turn it into a native module, run:

$> pythran dprod.py

That will generate a native dprod.so that can be imported just like the former module.

Documentation

The user documentation is available in the MANUAL from the doc directory.

The developer documentation is available in the DEVGUIDE from the doc directory. The also is a TUTORIAL for those who don't like reading documentation.

A todo list is maintained in the eponymous TODO.

The CLI documentation is available from the pythran help command:

$> pythran --help

Some extra developer documentation is also available using pydoc. Beware, this is the computer science incarnation for the famous Where's Wally? game:

$> pydoc pythran
$> pydoc pythran.typing

Examples

See the pythran/tests/cases directory from the sources.

Contact

Praise, flame and cookies:

• pythran@freelists.org -- register at http://www.freelists.org/list/pythran first!
• #pythran on FreeNode
• serge.guelton@telecom-bretagne.eu

Authors

See AUTHORS.

License

See LICENSE file.