Stephane Wirtel (@matrixise)

To the Python Community, Thank you for the FOSDEM !

2013-02-19T00:00:00+01:00

Python @ FOSDEM

How everything started

In September, I was at PyConFR 2012 with Ludovic Gasc and Christophe Simonis and we discussed about an hypothetical event in Belgium.

I said, and why not at FOSDEM 2013?

At beginning, Ludovic was thinking of a small joke but in fact it wasn’t.

First point

I began to make contact with some members of the FOSDEM team, the AFPy, and send some emails to the right contacts.

But this adventure was not easy to create.

Contact the FOSDEM team and ask a devroom and wait for the result
Contact a lot of potential speakers
Create the website
Create the committee
Select the talks
Publish the schedule
Wait for the FOSDEM ;-)

And I just announced this event on Hacker News This experience was really rewarding.

And now, it’s a great adventure

To the Python Community, Thank You !

I would like to tell thank you to everybody !

With your presence, this event has been a real great success.

I was very impressed and honored to meet you

Thank you so much !

Alain Meurant

You and Affinitic are a great sponsor with your involvements in the Python community. I was so happy to see your support at this event with your sponsorship. I worked with him at Aragne at the beginning of Python in Belgium (in the early 2000s).

Thank you for your help.

Alexis Métaireau

I met you at the PyConFR 2012 at Paris and you are a good guy. I discussed with you about Werkzeug, Pelican and Cornice.

For this event, you proposed a talk about the good practices for a dummy in Python. During your talk, the room was full ;-)

Thank you,

Andrea Crotti

Thank you for your talk about the “TDD from scratch”. This talk is still important and useful that apply to everyone (beginners or experts) because you show the basis of a good product / component. And sometimes, we forget those important details.

Benoit Chesneau

This is the first time wherein I can discuss with you about gunicorn and gaffer. The last time was at the pyconfr 2010, you showed the development of gunicorn and I was impressed by the project and your technical skill.

Now, we discuss via irc or twitter and I appreciate our discussion.

Thank you,

Brian Curtin

I don’t know you but thank you for your post on the PyCon blog.

Christophe Chauvet

In the first place, you wanted to be in the PostgreSQL devroom, but after the first talk you stayed in the Python devroom.

Thank you for your help in the devroom.

Hope to see you at a next OpenERP event.

Christophe Simonis

My friend for 8 years, he was with Ludovic and me at PyConFR 2012. He helped me for the “Python @ FOSDEM” website. From start to finish you were with me for this event. Thank you so much.

Eric Vernichon

Hi dude,

Thank you for your support, I really appreciate to see you at this event. Hope to see you at a next OpenERP event.

Gael Pasgrimaud

The financial officer of the “Association Francophone Python”. He managed the sponsoring from the Python Software Foundation and Affinitic.be.

Jonas Geiregat

Before this week-end, I had just discussed via email, irc or twitter but not face to face. I really appreciate his work for this edition.

Thank you.

For the next edition, would you be interested to organize with us ?

Kenneth Hoste

Your talk on EasyBuild was very interesting. We will see us at a future AFPy.ro in Belgium.

Kenneth Reitz

Thank you dude, you have received my email about this hypothetical event, and you said me “Yes, I will go to FOSDEM”.

Imagine this developer, he works for Heroku in California and he tells me, “yes I come”, it’s wonderful, awesome !

Thank you so much but I have one regret :( I didn’t tell you face to face ! Hope to see you in a future event and thank you again.

Lionel Dricot

Hi Ploum, Mister ploum, I like to read your articles on your blog and now your G+ account !

You were the first speaker to present your talk and I think that it wasn’t easy for you. I have really appreciated your presence at this Python @ FOSDEM.

Ludovic Gasc

Until the week-end of the FOSDEM, this guy does not know whether can help me for the event because his wife was pregnant.

I think you have a great gift my friend, a great success for the Python @ FOSDEM 2013 and you are daddy ;-)

Welcome in the Daddy club.

Marc-André Lemburg (PSF)

Thank you, I didn’t know you before this event, we discuss about a financial help from the Python Software Foundation. Sincerely, thank you.

Michel Cervello

Thank you for your talk on Plone.

Nelle Varoquaux

Thank you Nelle, you has been awesome, you saved my life ;-) You know why, because there was a missing speaker, and you gave an unprepared talk.

Really, Respect !

That was really nice to have a woman in the speakers’ group.

Ryan Faulkner

Thank you for your talk, I didn’t discuss with you but I appreciate your talk.

Sandro Munda

It was the first talk for this young dev, your talk was really interesting and now you have a small community to help you for the development of your talk. Great, continue !

Saul Ibarra Corretgé

This guy knows his subject, there is no secret for him about the gevent, pyuv. I didn’t discuss with you but I had a lot of questions about pyuv. Hope to see you at the next edition.

Tarek Ziadé

Tarek, who not know this python dev ? You have proposed to give a t-shirt to the speakers, this idea was original and appreciated by the speakers.

You were a member of the committee, thank you for your help and advice.

Thank you for the wine bottle, I will taste it with my family in few months.

Thierry Carrez

You have been surprised to receive a t-shirt and a package of belgian chocolate. Thank you for your talk about OpenStack and the jobs in this section.

Victor Stinner

Your talk on the AST was really awesome, You did not know if you could come because you hurt your back.

Thank you so much for your talk !

Hope to see you at the next PyCon FR ;-)

Vincent Driessen

I used his code with Python-RQ and RQ-Dashboard, I have read his blog post about the git workflow.

He was honored to be contacted by us and meet a lot of famous members of the Python Community.

Thank you and I have a dinner for you ;-) I didn’t forget.

Ralf Hildebrandt, Sjoerd Mullender

Thank your for the mailing lists (fosdem@python.org)

Juan Batiz-Benet

No news, I hope that everything goes well for him, but thank you for your talk.

Till von Ahnen

I was really sad for him because he had a car accident, I’m really sorry for you :(

And for the next edition ?

A larger room ;-)
A stand for the Python Language and the PSF
A lot of goodies (t-shirts, mug, flyers, stickers, cuddly toy, …)
A dinner with Python Community.

Download Ubuntu and create USB stick with your Mac

2013-02-13T00:00:00+01:00

Download Ubuntu and create USB stick with your Mac

Today I have to re-install my laptop because there is a bug in the grub boot-loader.

I tried the grub-rescue command and this one was really useful to list the file systems, etc…

But I didn’t have time to find the documentation on internet, so, first solution, re-install Ubuntu.

Here is the original link for this documentation: http://www.ubuntu.com/download/help/create-a-usb-stick-on-mac-osx

In a terminal

Download the Ubuntu ISO:

wget ftp://ftp.belnet.be/pub/ubuntu.com/releases/12.10/ubuntu-12.10-desktop-amd64.iso

Convert the iso file to img using the convert option of hdiutil:

hdiutil convert -format UDRW -o ~/path/to/target.img ~/path/to/ubuntu.iso

Run diskutil list to get the current list of devices:
```
diskutil list
```
Insert your flash media and run diskutil list agin and determine the device node assigned to your flash media (e.g. /dev/disk2).
Unmount the disk (in my case, the flash media was mapped on /dev/disk2)
```
diskutil unmountDisk /dev/disk2
```

Copy the .img file to the flash media:

sudo dd if=/path/to/target.img of=/dev/rdisk2 bs=1m

Eject the flash media:
```
diskutil eject /dev/disk2
```
Restart and boot with your USB stick.

Of course, these instructions will work fine with an other distribibution.

Test your SMTP connection within your projects

2013-01-21T00:00:00+01:00

dsmtpd - a SMTP server for debugging

Sometimes I have to use a smtp server to check my outgoing emails but I don’t want to pollute the internet with some useless emails.

So in this case, I just developped this small tools for me but you will may be interested by this project. I called dsmtpd for debugger smtp server (the last ‘d’ comes from the smtpd python lib)

Installation

This package is on pypi, so you can use pip to install it:

pip install dsmtpd

Use it

To use it, it’s very simple:

dsmtpd

2013-01-13 14:00:07,346 INFO: Starting SMTP server at 127.0.0.1:1025

By default the server will run on the port 1025 and on the loopback interface. It’s pretty easy.

Indeed, there is a small help:

dsmtpd --help

dsmtpd: A small SMTP server for the smart developer

Usage:
    dsmtpd [-i <iface>] [-p <port>] [-d <directory>]

Options:
    -p <port>      Specify the port for the SMTP server [default: 1025]
    -i <iface>     Specify the interface [default: 127.0.0.1]
    -d <directory> Specify the directory to save the incoming emails
    -h --help
    --version

Change the port

If you want to change the default port, you can do it with the ‘-p’ option:

sudo dsmtpd -p 25

2013-01-13 14:00:07,346 INFO: Starting SMTP server at 127.0.0.1:25

Change the interface

By default, dsmtpd listens to the loopback but you can change that with the ‘-i’ option:

dsmtpd -i YOUR_IP

2013-01-13 14:00:07,346 INFO: Starting SMTP server at YOUR_IP:1025

2013-01-13 14:00:19,634 INFO: 127.0.0.1:55549: stephane@wirtel.be -> stephane@wirtel.be [test Sun, 13 Jan 2013 14:00:19 +0100]

You can store the incoming emails

There is a cool feature with dsmtpd, all the emails can be caught by the mail server with the ‘-d’ parameter:

dsmtpd -d store_directory

This store directory is a maildir directory, so you will be able to use the mutt software to see the emails. It’s really useful if you want to see all the emails with the description of the headers.

mutt -f store_directory

Contributions

The project is hosted on github

I’am really sorry for the declined talks !

2013-01-16T00:00:00+01:00

Dear Samuele Reghenzi

I would like to apologize to you and to the others submitters of which the talk has been declined.

I agree with you, an email would be welcome !

I am aware than my organization for this event is not perfect because it’s my first event as main organizer. I’m really sorry for this inconvenience.

Indeed, I had written an email in the case of the talk was declined but I just forgot to send it :(

Between the development of the website, the organization of this event, my family, my job and many other things, I have just forgot to send you this email.

I am really really sorry.

Please, Could you accept my apologize ?

PS: Moreover, I would like to tell you thanks. I learned one thing today because of you, I need to improve my organization for the next edition to avoid this kind of inconvenience.

Thank you so much,

Best regards,

Stéphane Wirtel (@matrixise)

Astuce Windows Iterator

2013-01-14T00:00:00+01:00

Voici le lien original: http://sametmax.com/implementer-une-fenetre-glissante-en-python-avec-un-deque/

Window Iterator

Une astuce que j’ai pû voir sur le site de sam&max.

#!/usr/bin/env python
from collections import deque
from itertools import count
from itertools import islice
from itertools import izip

DEFAULT_SIZE=3

def window(iterable, size=DEFAULT_SIZE):
    iterable = iter(iterable)
    d = deque(islice(iterable, size), size)

    yield d
    for iterm in iterable:
        d.append(item)
        yield d

def main():
    for x in izip(count(), window(xrange(1, 13))):
        print x

if __name__ == '__main__':
    main()

New Year’s Python Meme 2012

2013-01-09T00:00:00+01:00

New Years’ Python Meme 2012

1. What’s the coolest Python application, framework or libraries you have discovered in 2012 ?

VirtualEnv (virtualenvwrapper), this tool is really awesome, you can create a new environment for your test and your development. really useful if you want to contribute to a python project. just python setup develop and let’s go.
Requests (by Kenneth Reitz) is a real advantage when you develop a http client because urllib is really bullshit.
Python-RQ (by Vincent Driessen), we can create a PubSub with Redis and it’s easy.
Gunicorn (by Benoit Chesneau), a very good WSGI server, I use it for my own project, this wsgi server can be included in your own project via the python api.

If you want to monitor your processes you can use these tools.

Circus (by Tarek Ziade and the Mozilla Foundation)
Gaffer (by Benoit Chesneau)

2. What new programming technique did you learn in 2012 ?

I learned the asynchronous programming with gevent and python-rq, these libraries are really awesome.
before december 2012, I didn’t use Github excepted to fetch the sources of several projects.

3. Which open source project did you contribute to the most in 2012 ? What did you do ?

OpenERP I am a core dev for 5 years. Python via PythonFOSDEM.

I organize with some friends a python devroom at FOSDEM2013. It’s really great for the Python community, we want to improve the visibility of Python.

If you want to contribute, the repository is on GitHub

4. Which Python blog or website did you read the most in 2012 ?

The python section of Reddit and HackerNews and sometimes you can find a lot of great articles on HumanCoders

5. What are the three top things you want to learn in 2013 ?

Gateway , based on python-rq, this library is a tool to exchange the data from one source to one destination via the concept of Workers.
PrintUs , this is a prototype, but this is a printing server based on pyzmq.
LLVM I implemented a small interpreter for a minimalist lisp language and I would like to compile it with LLVM.
The foundation of the python network programming.
Erlang
Go
Scala

6. What is the top software, application or library you wish someone would write in 2013 ?

A CMS based on Python and Flask
a wiki based on git and python (flask) and with the reStructuredText support

Want to do your own list? Here’s how:

copy-paste the questions and answer to them in your blog
tweet it with the #2012pythonmeme hashtag.

Appel officiel pour les présentations Python pour le FOSDEM 2013

2012-11-20T00:00:00+01:00

Cette année, Python sera représenté par sa communauté. Si vous désirez discuter avec des utilisateurs/développeurs Python, c’est là qu’il faudra être en février !

Comme chaque année, le FOSDEM prendra place durant le premier week-end de février à Bruxelles (Belgique).

Nous disposerons d’une salle de 80 sièges dans le bâtiment K. Cette salle sera ouverte tout le dimanche 3 février.

Si vous désirez obtenir une session dans la salle Python, s’il vous plaît, allez sur le site du python-fosdem pour remplir un formulaire

L’appel est ouvert jusqu’au 21 décembre 2012.

Cette année, les propositions seront revues par un comité:

Tarek Ziade (Mozilla)
Ludovic Gasc (Eyepea)
Christophe Simonis (OpenERP)
Stephane Wirtel (OpenERP)

Merci de soumettre vos présentations et a bientot à Bruxelles pour parler Python et/ou boire une bonne bière Belge.

Bien à vous,

Stéphane Wirtel (@matrixise)

Call Of Papers FOSDEM 2013

2012-11-20T00:00:00+01:00

This year, Python will be represented by its Community. If you want to discuss with a lot of Python Users, it’s the place to be in February !

Like every year, FOSDEM will take place the first week-end of February in Brussels (Belgium).

We will have a room in the K building (80 seats). This dev room will be open all day Sunday, February 3rd.

If you want to hold a session in the Python devroom, please go to the python-fosdem web site and fill the survey.

Call for Papers is open until December 21st

This year, the submissions will be reviewed by a committee:

Tarek Ziade (Mozilla)
Ludovic Gasc (Eyepea)
Christophe Simonis (OpenERP)
Stephane Wirtel (OpenERP)

Thank you for submitting your sessions and see you soon in Brussels to talk Python and/or have some nice Belgian Beers.

Best regards,

Stephane Wirtel (@matrixise)

Python @ FOSDEM 2013

2012-10-22T00:00:00+02:00

Python @ FOSDEM 2013

Dear Python Community,

I am really glad to announce to you that we will have a python devroom at FOSDEM 2013.

FOSDEM is a free event offering open source communities a place to meet, share ideas and collaborate. It is renowned for being highly developer-oriented and brings together 5000+ geeks from all over the world.

Cheers,

Stephane

Python au FOSDEM 2013

2012-10-22T00:00:00+02:00

Python au FOSDEM 2013

Chère Communauté Python,

Je suis réellement content de vous annoncer que nous aurons une salle pour la communauté Python au FOSDEM 2013.

Le FOSDEM est un événement gratuit offrant aux communautés Open Source un lieu de rencontre, de partage d’idées et de collaboration.

Bien à vous,

Stéphane

Page de Donation de l’AFPY

2012-09-24T00:00:00+02:00

Ajouter un lien vers la page de donation de l’AFPY

Lors de la dernière Assemblée Générale de l’Association Francophone de Python (AFPY) qui s’est déroulée le 16 septembre au PyConFR je me suis demandé de quelle manière je pouvais aider cette association.

Une des réponses à ma question est la suivante, propager la page de donation.

Voilà chose dite, chose faite.

Je vous invite toutes et tous à présenter la page de donation sur votre blog.

Merci,

MailCatcher, L’outil pour le développeur

2012-02-21T00:00:00+01:00

MailCatcher

J’avais envie de vous présenter MailCatcher qui est un petit outil très utile que tout développeur doit au moins connaitre.

L’idée de MailCatcher, est de proposer un server SMTP tournant sur le port 1025 et permettant ainsi de récupérer tous les emails qui auront été émis par votre application.

Cet outil est très important si vous désirez vérifier tous les emails sortant de votre système sans devoir modifier une configuration ou commencer à bidouiller des emails.

Installation

MailCatcher est un gem Ruby, il est donc installable sur les systèmes possédant Ruby

gem install mailcatcher

Afin de l’exécuter, veuillez simplement lancer l’application via la commande suivante

mailcatcher

Attention, cette commande lancera MailCatcher en mode daemon par défaut, si vous désirez l’utiliser sans passer par un mode daemon, ajouter le paramètre -f sur la ligne de commande.

mailcatcher -f

Utilisation

J’utilise l’outil swaks que vous pouvez installer via le package manager de votre distribution. Dans mon cas, il s’agit de fedora, mais vous pouvez l’avoir dans Debian ou Ubuntu.

yum install swaks

apt-get install swaks

Une fois installé, vous pouvez utiliser l’outil comme indiqué dans l’exemple ci-dessous

swaks --from stephane@wirtel.be --to stephane@wirtel.be --server localhost --port 1025

Vérification

Comme vous pouvez le constater, mon mail est bien arrivé dans MailCatcher, et il m’est donc possible de le regarder à mon aise.

Cas d’utilisation

Utilisation avec Devise et Rails

Devise donne la possibilité d’envoyer des emails lors de l’inscription d’un utilisateur. Avec cet outil, il vous est donc possible de regarder le contenu sans devoir passer votre temps à inspecter les logs de Rails.

Références

PostgreSQL - Foreign Data Wrapper

2012-02-04T00:00:00+01:00

PostgreSQL - Foreign Data Wrapper for OpenERP

Here is a simple example, but via this connector we can connect to #OpenERP from PostgreSQL

#!/usr/bin/env python
# -*- encoding: utf-8 -*-
# copyright: 2012 - Stephane Wirtel <stw@openerp.com>
from multicorn import ForeignDataWrapper
from .utils import log_to_postgres
from itertools import cycle
from datetime import datetime

from rpc import XmlRPCConnector, Connection, Object

class OpenERPForeignDataWrapper(ForeignDataWrapper):
    def __init__(self, fdw_options, fdw_columns):
        super(OpenERPForeignDataWrapper, self).__init__(fdw_options, fdw_columns)
        self.connector = XmlRPCConnector(fdw_options['hostname'],
                                                 int(fdw_options['port']))
        self.cnx = Connection(self.connector,
                              fdw_options['database'],
                              fdw_options['username'],
                              fdw_options['password'])
        self.object_name = fdw_options['object']

    def execute(self, quals, columns):
        log_to_postgres(str(quals))
        log_to_postgres(str(columns))

        proxy = Object(self.cnx, self.object_name)

        item = {}
        for record in proxy.select([], fields=list(columns)):
            for column in columns:
                item[column] = record[column]
            yield item

multicorn

I just created a small shell script to drop and create the database, and create the server.

#!/usr/bin/env bash

PGSQL=/usr/bin/psql

DATABASE=openerp_multicorn

$PGSQL -d postgres -c "DROP DATABASE IF EXISTS ${DATABASE}"
createdb $DATABASE
$PGSQL -d $DATABASE <<-EOF
CREATE EXTENSION multicorn;

CREATE SERVER multicorn_openerp
    FOREIGN DATA WRAPPER multicorn
    OPTIONS (wrapper 'multicorn.openerpfdw.OpenERPForeignDataWrapper');

CREATE FOREIGN TABLE external_oe_users (
    login character varying,
    name character varying)
    SERVER multicorn_openerp OPTIONS (
        hostname 'localhost',
        port '8069',
        database 'openerp',
        username 'admin',
        password 'admin',
        object 'res.users'
    );

SELECT login, name FROM external_oe_users where login='admin';

EOF

stephane@atlantis:~$ ./multicorn.sh
DROP DATABASE
CREATE EXTENSION
CREATE SERVER
CREATE FOREIGN TABLE
NOTICE:  [login = admin]
NOTICE:  set(['login', 'name'])
 login | name
-------+-------
 admin | Admin
(1 row)

Installation of OpenERP 5.0 on Debian 6.0

2011-11-02T00:00:00+01:00

How to install OpenERP 5.0 on Debian 6.0

OpenERP

OpenERP is an open source comprehensive suite of business applications including Sales, CRM, Project management, Warehouse management, Manufacturing, Accounting and Human Resources. OpenERP has separate client and server components. XML-RPC interfaces are available.

In this study case, we will use the XML-RPC protocol, because it is based on the HTTP protocol containing the Host. And if you want to use this configuration for the NetRPC protocol, you will have a problem because this protocol does not contain the Host.

OpenERP

Install the dependencies

With your root access, we will install the dependencies for OpenERP v6.0.

~# useradd -m -s /bin/bash openerp
~# apt-get install postgresql
~# su postgres -c "createuser -s openerp"
~# apt-get install python-psycopg2 python-reportlab python-voject
~# apt-get install python-lxml python-pychart python-tz

Fetch and Execute OpenERP

Using the openerp user:

~$ mkdir openerp

~$ wget http://www.openerp.com/download/stable/source/openerp-server-5.0.16.tar.gz
~$ tar xfz openerp-server-5.0.16
~$ cd openerp-server-5.0.16/server/bin
~$ ./openerp-server.py --log-level=debug

Connect to the server

You can use the GTK client and you set the right information in the dialog box for the connection
You can use this python code

#!/usr/bin/env python
from xmlrpclib import ServerProxy
def get_version(hostname, port=8069):
    server = ServerProxy('http://%s:%d/xmlrpc/db' % (hostname, port,))
    return server.server_version()

def main():
    print "Version: %r" % (get_version('localhost'),)

if __name__ == '__main__':
    main()

Installation of OpenERP 6.0 on Debian 6.0

2011-11-02T00:00:00+01:00

How to install OpenERP 6.0 on Debian 6.0

OpenERP

OpenERP

Install the dependencies

With your root access, we will install the dependencies for OpenERP v6.0.

~# useradd -m -s /bin/bash openerp
~# apt-get install postgresql
~# su postgres -c "createuser -s openerp"
~# apt-get install python-psycopg2 python-reportlab python-voject
~# apt-get install python-lxml python-pychart python-tz
~# apt-get install python-yaml python-mako

Fetch and Execute OpenERP

Using the openerp user:

~$ mkdir openerp

~$ wget http://www.openerp.com/download/stable/source/openerp-server-6.0.3.tar.gz
~$ tar xfz openerp-server-6.0.3
~$ cd openerp-server-6.0.3/server/bin
~$ ./openerp-server.py --log-level=debug

Connect to the server

You can use the GTK client and you set the right information in the dialog box for the connection
You can use this python code

#!/usr/bin/env python
from xmlrpclib import ServerProxy
def get_version(hostname, port=8069):
    server = ServerProxy('http://%s:%d/xmlrpc/db' % (hostname, port,))
    return server.server_version()

def main():
    print "Version: %r" % (get_version('localhost'),)

if __name__ == '__main__':
    main()

Nginx and multiple instances of OpenERP

2011-11-02T00:00:00+01:00

Nginx and multiple instances of OpenERP

nginx

Nginx is a free, open-source, high-performance HTTP server and reverse proxy, as well as an IMAP/POP3 proxy server.

OpenERP

Install the OpenERP versions

For that, I suppose you can install OpenERP v5.0 and v6.0 but it’s not the case, you can read my tutorials for the installation of v6.0 or v5.0.

I will use the installation paths for the rest of my tutorial

Don’t forget to change the port of your instances.

For v5.0:

~$ ./openerp-server.py --log-level=debug --no-netrpc --port=9069

For v6.0:

~$ ./openerp-server.py --log-level=debug --no-netrpc --no-xmlrpcs --xmlrpc-port=10069

The configuration

In our study case, we know we use the XML-RPC protocol assigned to the port 8069.

We will use the proxy module of nginx for the solution. In my example, I will use the port 9069.

The first openerp will use the port 9069 and the second will be 10069.

For nginx, it’s very simple, just define a configuration file for the servers.

server {
    server_name openerp_v5.example.com;
    listen 8069;
    location / {
        proxy_pass http://localhost:9069;
        proxy_set_header Host $http_host;
        proxy_set_header X-Real-IP $remote_addr;
    }
}

server {
    server_name openerp_v6.example.com;
    listen 8069;

    location / {
        proxy_pass http://localhost:10069;
        proxy_set_header Host $http_host;
        proxy_set_header X-Real-IP $remote_addr;
    }
}

Don’t forget to change your DNS entries to point on the same server, in this case, localhost:

127.0.2.1   openerp_v5.example.com openerp_v6.example.com

Check the configurations

Here is a small python script to check that the proxy is well configured

#!/usr/bin/env python
from xmlrpclib import ServerProxy
def get_version(hostname, port=8069):
    server = ServerProxy('http://%s:%d/xmlrpc/db' % (hostname, port,))
    return server.server_version()

def main():
    print "Version (openerp_v5.example.com): %r" % (get_version('openerp_v5.example.com'),)
    print "Version (openerp_v6.example.com): %r" % (get_version('openerp_v6.example.com'),)

if __name__ == '__main__':
    main()

PostgreSQL: Use GNU/Parallel and pg_dump

2011-07-15T00:00:00+02:00

How to use GNU/Parallel and pg_dump ?

I work with the PostgreSQL database server and for the backup process, I use the pg_dump utility.

If you have many databases, you know there is a problem with this tool. In fact, it’s a serial process, it saves database after another and if you have a big one, you risk to wait for longtime. In my situation, this is problematic and I looked to use an other tool to parallelize the process.

But I discovered the tool GNU/Parallel.

GNU/Parallel is a shell tool for executing jobs in parallel using one or more computers.

Installation

wget http://ftp.gnu.org/gnu/parallel/parallel-20110722.tar.bz2
tar xfj parallel-20110722.tar.bz2
cd parallel-20110722
./configure --prefix=/opt/local
make
make install

The Script

#!/usr/bin/env bash
# Path to the GNU/Parallel tool
PARALLEL=/opt/local/bin/parallel

# We select all the databases we want to backup except the 'template0',
# 'template1' and 'postgres' databases
INPUTLIST=`psql -d postgres -A -t <<EOF
    SELECT datname FROM pg_database
    WHERE datname NOT IN ('template0','template1','postgres')
    ORDER BY datname ASC;
EOF`

# And we call the parallel tool with the right arguments
CMD='pg_dump --no-owner {} | gzip -9 > ${OUTPUT_DIR}/{}.dump.sql.gz'
(echo "$INPUTLIST") | $PARALLEL "$CMD"

If you have an other solution to this problem, I’m interrested

In a future post, I will show you how to use GNU/Parallel and rsnapshot to parallelize the rsync tool.

nginx, gunicorn and Flask

2011-02-24T00:00:00+01:00

How to install nginx, gunicorn and Flask

I like to work on my personal project Caviste and for the moment, I wanted to use nginx, gunicorn with my Flask application. To explain the architecture of my installation, nginx is the web server with a cache for the static content (html, css, javascript and images). the Flask application will be served by the gunicorn webserver; When there is a request, the nginx server (front-end) will forward to the gunicorn server for the dynamic content.

Flask

Here is a small project I like to follow, this one has been developped by Armin Ronacher. The documentation is certainly the best documentation for a project, pragmatic and concise. The source code is well documented and the developer is the maintainer of Werkzeug, Pygments and is in the Pocoo Team.

Installation

pip install flask

Example

#!/usr/bin/env python
# -*- encoding: utf-8 -*-
"""
    deploy.py
    ~~~~~~~~~

    :author: (c) 2011 - Stephane Wirtel <stephane@wirtel.be>
    :license: BSD
"""
from flask import Flask
from flask import render_template_string

# load the middleware from werkzeug
# This middleware can be applied to add HTTP proxy support to an application
# that was not designed with HTTP proxies in mind.
# It sets `REMOTE_ADDR`, `HTTP_POST` from `X-Forwarded` headers.
from werkzeug.contrib.fixers import ProxyFix


app = Flask(__name__)
@app.route('/')
def index():
    return render_template_string('<h1>Hello World</h1>')

app.wsgi_app = ProxyFix(app.wsgi_app)

Gunicorn

I discovered gunicorn during the PyConFR 2010 at Paris, I found this project very cool, because the community is very active, and Benoit Chesneau is a good evangelist of gunicorn. He wants to create the best WSGI server, the fastest and the simplier to configure.

gunicorn, written in Python, it’s a very minimalist web server but its works is well done.

Installation

pip install gunicorn

Configuration

The configuration is easy, you just define a configuration file. Here is a small example where we specify the number of workers.

workers = 2
bind = '127.0.0.1:18000'
proc_name = 'www.example.com'
pidfile = '/tmp/www.example.com.pid'

gunicorn deploy:app

Nginx

Installation

apt-get install nginx

Configuration

The configuration of nginx (on Debian Like) is in the /etc/nginx/sites-enabled directory

upstream frontends {
    # We define the binding of the gunicorn web server
    server 127.0.0.1:18000;
}

server {
    listen 80;

    server_name localhost;

    access_log /var/log/nginx/localhost-access.log;
    error_log /var/log/nginx/localhost-error.log debug;

    location / {
        proxy_pass_header Server;
        proxy_set_header Host $http_host;
        proxy_set_header X-Forward-For $proxy_add_x_forwarded_for;
        proxy_redirect off;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header X-Scheme $scheme;
        if (-f $request_filename) {
            proxy_pass http://frontends;
            break;
        }
    }
}

Flask-OpenERP

2011-02-22T00:00:00+01:00

How to discuss with OpenERP from Flask ?

I started to work on a new extension for the Flask project. This extension will help the developer to integrate OpenERP in this web app.

GitHub: https://github.com/matrixise/flask-openerp

Enfin Papa ;-)

2011-01-21T00:00:00+01:00

Margaux 1.0

Après neuf mois de developpement, de roadmap, de milestones, nous sommes enfin parvenus à une version 1.0 de Margaux ;-)

Ma compagne et moi-même avons donc le grand plaisir de vous informer que notre superbe petite fille Margaux est née le 21 janvier 2011 à 3h26, pesant 3,270kg et mesurant 51 cm.

Regardez par vous-même

Loempia

2010-10-04T00:00:00+02:00

Loempia - Addons Registry for OpenERP

Hi all,

I want to talk about a personal project for OpenERP. This project is Loempia.

Loempia will be the official repository for the modules of OpenERP, it allows to share the modules with the community.

Why Loempia?

Currently, the community and OpenERP S.A. use LaunchPad to share the modules of OpenERP.

The problem with this way of working is that we don’t know where a module is and if this module is up to date or not. Where are the certified modules ? etc…

I think Loempia can help everybody, because it will be the main repository for the modules of OpenERP.

With Loempia, you will have the capability to share your modules and comments. You could use the tag and rating systems.

You will have the capability to make some searchs with some specific criteria.

You will be able to look for the modules for the accounting for the version 5.0 of a specific contributor.

Do you want the certified modules for OpenERP 6.0, a small search on Loempia and you will get the results.

I think it’s a very useful site and not only a marketing tool!

Why don’t you want to use PyPi?

PyPi is a very cool site if you want to share a python module, it’s very useful and it works fine. But it’s a technical site. Yes, we could use it to store the openerp’s modules and after, what? You could show the best contributors, the most downloaded modules, and promote modules with screenshots.

Some features of Loempia:

You could select a module of a given contributor.
You could select a module for a specific version of OpenERP.
How can I get the certified modules? Are these modules updated?

I hope that Loempia will help OpenERP, the Community and the developers.

What will we propose in Loempia?

Rating System
Comment System
Tag System
User / Authentication system (OpenID, you will be able to connect with your LaunchPad ID account)
Allows to register new launchpad branches or upload zip files
Generate the documentation of the modules and integrate it into Sphinx
Allows the screenshots (thanks to NaNtic)
Produce sample of PDF reports
Develop web-services for the API to register/browse addons
Allows the possibility to register a branch or a module in the buildbot of OpenERP and add the results in the module statistics
Fetch the specifications of a module in JSON, XML via an url
Download the module via an url
A python library and a command line client to handle the modules

$ loempia search <module>
$ loempia show <module+version>
$ loempia fetch <module+version>
$ loempia login <user> <version>
$ loempia push <module>
$ loempia install <module>
$ loempia update {-all | <module> | <module+version> }
$ loempia outdated

Currently, the project is hosted on http://www.loempia.org, but you can see a mock-up on http://beta.loempia.org.

Here is the mock-up of the screens with balsamic

If you have comments or suggestions, you can contact me.

2009-11-24T00:00:00+01:00

Open ERP - How to override an xml attribute in an xml view

A nice feature has been developed in the trunk version of Open ERP. It’s possible to override an xml attribute of an openerp view with the ‘attributes’ position.

For Example:

<?xml version="1.0" encoding="UTF-8" ?>
<openerp>
    <data>
        <record model="ir.ui.view" id="res_partner_form">
            <field name="name">res.partner.form</field>
            <field name="model">res.partner</field>
            <field name="type">form</field>
            <field name="inherit_id" ref="base.view_partner_form" />
            <field name="arch" type="xml">
                <xpath expr="/form/notebook/page[@string='General']/field[@name='address']/tree"
                       position="attributes">
                    <attribute name="editable">top</attribute>
                </xpath>
            </field>
        </record>
    </data>
</openerp>

Just an update, If you want to use this feature with a page, don’t forget to remove the last ‘/’ in the xpath expression.

<record id="base.view_crm_partner_info_History" model="ir.ui.view">
    <field name="name">res.partner.crm.history.inherit1</field>
    <field name="model">res.partner</field>
    <field name="type">form</field>
    <field name="inherit_id" ref="base.view_partner_form"/>
    <field name="arch" type="xml">
        <xpath expr="/form/notebook/page[@string='History']" position="attributes">
            <attribute name="invisible">False</attribute>
        </xpath>
    </field>
</record>

RubyFlow - a new RSS flow for Ruby and Rails

2008-04-23T00:00:00+02:00

RubyFlow - un nouveau flux pour Ruby et Rails

Voici quelques semaines, Peter Cooper a lancé un nouveau projet nommé RubyFlow.

Il est vrai que la majorité des flux RSS que j’utilisais se tarissaient, et cela m’ennuyait. Ne plus avoir de news concernant Ruby et Rails, arrrgghhh !!!, c’était insoutenable.

Maintenant, des centaines de news viennent à moi, oh que je suis content lol.

Voici la version française de ce site RubyFlow FR

Code Cooking

2008-03-23T00:00:00+01:00

Du code, c’est de la cuisine, il faut avoir la technique et les ingrédients.

Le développement de logiciel, je suis désolé pour certains, mais pour moi, il s’agit d’un art. Je suis écrivain en logiciel, et j’essaie d’écrire le meilleur code qu’il soit. Je ne m’amuse pas à trouver des phrases bien tournées pour les mettres ensemble. J’essaie de comprendre et de créer quelque chose qui soit facilement lisible, compréhensible et maintenable très facilement.

Je discute avec un ami que je n’ai plus vu depuis des années, et cela me chagrine de l’entendre me dire tout le contraire de ce que j’aime. Le goût du bon code. Jean Pierre Coffe dirait “Le goût de la bonne bouffe”, et bin moi, c’est la même chose mais pour du code. Je n’aime pas voir un code insipide, je veux du code, qui me parle et où la lecture me permettre une compréhension très rapide.

Il est clair que l’on doit vivre de son travail, moi je vis de mon code, c’est bête, mais on pourrait me comparer à un chef coq, et j’apprend chaque jour pour continuer à mitonner des bons petits plats :d

Une personne qui se respecte, écrira du bon code, du moins, c’est ce que je pense. Je ne peux m’imaginer la chose autrement.

Nouveau Travail chez OpenERP (anciennement TinyERP)

2008-03-19T00:00:00+01:00

Nouveau Travail chez OpenERP (anciennement TinyERP)

A partir du 14 mai 2008, nous travaillerons chez Tiny.be qui est la société éditrice de TinyERP.

TinyERP est un logiciel de gestion d’entreprise, dont la rivalité avec SAP et d’autres projets n’est plus à faire.

J’espère que l’équipe sera contente de nous recevoir, et que nous même, aurons un bel avenir.

2008-02-02T00:00:00+01:00

New Release of the “Programming Ruby”

Cool, a new release of the last beta version of “Programming Ruby”. I believe that Andy and Dave will prepare a very good book. I look forward to having a paper version.

Thanks again for your work.

Buy this book, it covers the last version of Ruby 1.9

The Pragmatic Programmers

Après la SOAP, le dessert

2007-12-05T00:00:00+01:00

Après la SOAP, le dessert !

Vous souvenez-vous, dans mon précédent billet, j’indiquais que j’allais devoir utiliser mouseHole afin de vérifier le bon fonctionnement d’un connecteur SOAP.

J’étais prêt à mettre la main à la pate, quand tout à coup, j’ai trouvé soapUI.

Il s’agit d’un outil de test pour WebServices.

A première vue, il avait l’air simple, mais j’ai pû constater qu’il était génial. Importer mon WSDL, générer un mock webservice et tester mon connecteur en quelques minutes, sans écrire une seule ligne.

Je ne vous dis pas la joie que j’ai pû ressentir en me disant que j’avais gagné quelques jours de développement à cause d’un manque d’outil pour debugger mes messages SOAP.

La chose vraiment intéressante, il est possible de scripter cet outil via Groovy afin de modifier les responses à la volée. Sincèrement, j’avais la bouche ouverte :d

SOAP + Delphi + C++

2007-12-04T00:00:00+01:00

A la SOAP !!

Voilà, pour le boulot, je vais devoir créer un connecteur SOAP. La partie la plus marrante, c’est que je ne dispose pas d’outils pour vérifier mes messages. Donc, je me suis entrain de penser à écrire un proxy http écrit en ruby, et qui fera intermédiaire entre moi et le webservice que je dois attaquer.

Je vais essayer tout à l’heure cet outil : mouseHole

J’espère que son utilisation m’aidera, car de cette manière, il m’aidera à valider mon code plus rapidement.

Pour ne pas oublier, je développe principalement en C++ Builder 6 et Delphi 7.

Je n’ai pas le choix, c’est les choix technologiques de mon employeur.

2007-12-03T00:00:00+01:00

Win32OLE, le retour

Suite à mon précédent article sur Win32OLE

Je viens de découvrir une méthode de classe très intéressante, surtout qu’elle appartient à la classe WIN32OLE. Il s’agit de la méthode

WIN32OLE::const_load( ole_object, module_or_class )

Vous souvenez-vous du code de mon précédent article ? J’avais défini un ensemble de constantes hiérarchisées à l’aide de modules.

module Word
    module PageSetup
        module Orientation
            Portrait = 0
            Landscape = 1
        end
    end
    # .....
end

En découvrant cette méthode, j’ai pû refactoriser mon code, afin de le rendre plus lisible et d’utiliser directement les constantes définies par la couche OLE. Voici donc, un peu de refactorisation de mon précédent code.

# Définition d'un module qui va servir à agréger les différentes constantes
# qui sont définie dans le composant OLE de Word.
module Word
end

begin
    word = WIN32OLE.new( "Word.Application" )
    # La méthode WIN32OLE::const_load charge les différentes constantes
    # depuis l'instance Word.
    # De cette manière, il n'est pas nécessaire de créer une correspondance
    # entre constante et valeur numérique
    WIN32OLE.const_load( word, Word )

    document = word.Documents.Add()
    document.PageSetup.PaperSize = Word::WdPaperA4
    document.PageSetup.Orientation = Word::WdOrientLandscape
ensure
    word.Quit()
end

Je vous donne un avis, je n’ai pas pû complètement la documentation lors de l’article précédent, ce qui m’a fait perdre du temps. Un conseil, lisez les exemples, la documentation et tout ce qui peut aider. Références

WIN32OLE::const_load

2007-11-25T00:00:00+01:00

Ruby, OLE et Word

Comment générer des documents Words avec OLE et Ruby ?

Je vais être honnête avec vous, je ne suis pas pour l’utilisation de Word, mais les personnes travaillant avec moi, utilisent majoritairement cet outil et ne jurent que par celui-ci. Afin d’utiliser correctement OLE avec Ruby, il faut avoir à sa disposition plusieurs outils, :-)

win32ole Elle fait partie de la librairie standard de Ruby et d’une très grande utilité ;-) sans ça, il est impossible de pouvoir utiliser Word via OLE.
ole_browser.rb qui est un script permettant d’introspecter les objets disponibles. Cet outil a été écrit par Dave Burt et Masaki Suketa.

#!/usr/bin/env ruby

require "win32ole"

Définition d’une nouvelle méthode basée sur l’opérateur ‘/’ afin de sucharger la classe String.

class String
    def / (other)
        File.join(self, other)
    end
end

Etant donné que je n’avais pas envie d’écrire tout le temps du code inutilement, et que j’adore refactoriser, j’ai ajouté une petite méthode au module Kernel. Celle me permet de gagner un peu de lisibilité, mais pas grand chose, je la met dans la zone des Helpers.o

module Kernel
    def with(obj)
        yield obj
    end
end

Bon, commençons par le plus intéressant, je pense que le plus intéressant est tout de même de charger Word :D donc, voici comment charger cet outil.

word = WIN32OLE.new( "Word.Application" )

Dans mon exemple, je ne veux pas voir l’interface de Word, de ce fait, je n’active pas sa visibilité.

word.Visible = false

Etant donné que cet article parle de la création de document, il ne s’agit pas d’ouvrir un document Word et de lui rajouter des infos, on part de zéro.

document = word.Documents.Add()

Afin que vous ne soyez pas perdu dans des nombres magiques, j’ai ajouté une petite liste de constantes, permettant de cette manière une utilisation plus aisée des composants.

module Word
    module PageSetup
        module Orientation
            Portrait = 0
            Landscape = 1
        end
        module Size
            PaperA4 = 7
        end
    end

    module Font
        module ColorIndex
            Black = 1
            Blue = 2
            Turquoise = 3
            BrightGreen = 4
            Pink = 5
            Red = 6
            Yellow = 7
            White = 8
            DarkBlue = 9
            Teal = 10
            Green = 11
            Violet = 12
            DarkRed = 13
            DarkYellow = 14
        end
    end

    module ParagraphAlignment
        Left = 0
        Center = 1
        Right = 2
        Justify = 3
        Distribute = 4
        JustifyMed = 5
        JustifyHi = 7
        JustifyLow = 8
        ThaiJustify = 9
    end
end

Cette partie est bête, mais spécifie le format de la page, de cette manière j’indique le format A4 en Paysage ( LandScape )

document.PageSetup.PaperSize = Word::PageSetup::Size::PaperA4
document.PageSetup.Orientation = Word::PageSetup::Orientation::Landscape

Pour toutes les sections se trouvant dans le document, nous allons ajoutés un en-tête et un pied de page. Création de l’entête avec une police de caractères ayant la couleur Violet, et dont le contenu sera “En-tête d’exemple”

for section in document.Sections
    with section.Headers( 1 ).Range do |range|
        range.Font.ColorIndex = Word::Font::ColorIndex::Violet
        range.Text = "En-tête d'exemple"
    end
    with section.Footers( 1 ).Range do |range|
        range.ParagraphFormat.Alignment = Word::ParagraphAligment::Right
        range.Font.ColorIndex = Word::Font::ColorIndex::DarkRed
        range.Text = "Stephane Wirtel < stephane DOT wirtel AT gmail DOT com>"
    end
end
with word.Selection do |sel|
    # Ajout d'une image,
    picture = sel.InlineShapes.AddPicture( Dir.pwd() / 'change_directory.png', nil, true )

    # Bon, on insère un break, c'est à dire une nouvelle page
    sel.InsertBreak()

    [ "Je ne suis pas Superman, mais j'y arrive :D",
    "Une option en plus que les autres :D" ].each do |line|
        sel.TypeText( "#{line}\n" )
    end
    sel.InsertBreak()

    # Spécification de la taille de la police à utiliser
    with sel.Font do |font|
        font.Name = "Bitstream Vera Sans Mono"
        font.Size = 10
        font.Bold = false
    end

    # On ajoute l'heure avec la date de cette manière, on peut déjà savoir la date de création du document
    sel.TypeText( Time.now.to_s )
    sel.InsertBreak()
end

Et si on enregistrait le document, ça serait cool, vous ne pensez pas ? Tout ce travail pour ne pas le sauver, ça serait débile :D Donc, tadaaaaam

document.SaveAs( Dir.pwd() / 'test.doc' )
document.close()

Maintenant il ne reste plus qu’à fermer l’application Word, si nous ne le faisons pas, il restera une instance Word ouverte, ce qui peut provoquer des problèmes.

word.Quit()

Imaginez maintenant la création d’un DSL pour simplifier tout ce code, ou aussi l’utilisation de Textile