Calling C functions inside Python

In my current work, I’m developing new ways to test a Telecom framework. Firstly, I was building automation tests with cgreen to blackbox/unit test model. But the development stages were going a little more slow than I thought because write and debug C code is not so easy.

Suddenly, I though…”Hey, why do I not build these tests with python scripts ???”. And my response was: “Ctypes is your friend!”

Ctypes permits you call C functions inside python code. It’s wonderful. Let’s see how to do it in OpenSolaris environment:

First, download ctypes:

$ wget -c \
http://ufpr.dl.sourceforge.net/sourceforge/ctypes/ctypes-1.0.2.tar.gz

Install ctypes in a local path (as normal user):

$ tar -zxvf ctypes-1.0.2.tar.gz

$ cd ctypes-1.0.2

$ python setup.py install --prefix=$HOME/python

Export your PYTHONPATH environment variable:

$ export PYTHONPATH=$HOME/python/lib/python2.4/site-packages

Testing ctypes:

$ python

>>> import ctypes

OK ! Ctypes is there, but how do I use it ???

See a little example that I made to explain how to use ctypes:

• A file named test.c with a function (sum) that returns a sum of two values:

int sum (int a, int b)
{
    return (a+b);
}

• Compile it as a shared library:

$ gcc -fPIC -c test.c

$ gcc -shared -o test.o libtest.so

• Make sure that our library will be found

$ cp test.so $HOME/example

$ export LD_LIBRARY_PATH=$HOME/example

• Let’s call the function sum through a Python script

$ python

>>> from ctypes import *
>>> cdll.LoadLibrary('libtest.so')
<CDLL 'libtest.so', handle fe7806e0 at 80dfcec>
>>> libc = CDLL('libtest.so')
>>> libc.sum()
112710980
>>> libc.sum(1,2)
3

I can see a smile in your face. Yeah! Ctypes is a great trick to C programmers

Enjoy it!

Useful links:

1. http://docs.python.org/library/ctypes.html

Decompiling Python Bytecodes

Oh no ! I lost my python source code !!! And now ?

You can use an online Python Bytecode Decompiler (DePython) made by Team509, a chinese group (I guess).

Recently, I received this tip/trick site by Python-Brasil mailing list.

Online Decompiler: http://www.depython.net/

Author(s): http://www.team509.com/

Viewing your Net-Vírtua Consumption

Last month, I received an email saying that I had consumed 80% of my link quota 2 weeks earlier to end the month.

It’s boring to have to access the page at Net-Vírtua site to view your monthly consumption. To facilitate this work, I made a simple Python script:

#!/usr/bin/python

import urllib
import urllib2
import datetime
import re

class Virtua():

    __virtua_url = "http://consumo.virtua.com.br/"
    __virtua_action = "consumo.php"

    def __init__(self):
        pass

    def __prettyfy(self,html):
        separator = '\n ------------- ------ '
        preformatted  = separator + '\n'
        preformatted += '|     date    |  GB  |'
        preformatted += separator

        regex_space = re.compile('\s+')
        regex_date = re.compile('<TD CLASS="bg_azulT azul b onze" nowrap>')
        regex_cons = re.compile('<TD CLASS="bg_cinza azulE onze">')
        regex_cell = re.compile('</TD>')

        # <TD CLASS="bg_cinza azulE onze">1.04</TD>  --> GB received
        # <TD CLASS="bg_cinza azulE onze">0.06</TD>  --> GB sended
        # <TD CLASS="bg_cinza azulE onze">1.10</TD>  --> GB consumed
        # <TD CLASS="bg_cinza azulE onze">1.10</TD>  --> Total
        # when count == 3, it's signifies the daily consum
        count = 0
        seq = 0
        date = ''
        cons = ''

        # html is a log phrase (all text in one line, it's bad)
        html_formatted = html.split('\n')

        for line in html_formatted:

            if re.findall(regex_date,line):         # get date
                date = re.sub(regex_date,'',line)
                date = date[:-5]
                date = re.sub(regex_space,'',date)  # take off blank spaces

            if re.findall(regex_cons,line):         # get daily consum
                count += 1
                cons = re.sub(regex_cons,'',line)
                cons = re.sub(regex_cell,'',cons)
                cons = re.sub(regex_space,'',cons)  # take off blank spaces

                if ( count % 4 == 0 ):              # completes a sequence (one seq == 4 TD's)
                    seq += 1

                if (count == (3 + 4 * seq)):        # daily consum
                    preformatted += '\n| %s | %s |' % (date,cons)
                    preformatted += separator

        preformatted += '\n\nTotal: %s GB' % cons
        return preformatted

    def getConsum(self,mac="cc0044bbcc99",month="",year=""):
        target = self.__virtua_url + self.__virtua_action

        month = datetime.date.today().month
        year  = datetime.date.today().year

        data =  {  
            'macadd':mac,
            'mes':month,
            'ano':year
        }

        input = urllib.urlencode(data)
        req = urllib2.Request(target,input)
        res = urllib2.urlopen(req)

        print self.__prettyfy(res.read())

if __name__ == "__main__":
    Virtua().getConsum()

Let’s turn {on|off} TV

I am not a TV addicted, but sometimes I like to watch movies on TV. As I don’t have money to spend with Pay-per-view or Cable TV, so before turn on mine TV and tune on some channel, I search at Folha Ilustrada for good films.

I think the Folha Ilustrada is better to consult, because in there all movies are classified as good, bad, not so bad, etc..

Yesterday, I was already bored to put the URL in my Firefox browser to go Folha Ilustrada and find something intersting. Then, I make a python script to bring to me the informations, look:

#!/usr/bin/python

import urllib2
import datetime
import re
from textwrap import TextWrapper
from BeautifulSoup import BeautifulSoup

class Films():

 _url          = 'http://www1.folha.uol.com.br/folha/ilustrada/filmes/'

 _today        = datetime.date.today().strftime('%A')

 _days_of_week = { 'Monday':'segunda',
 'Tuesday':'terca',
 'Wednesday':'quarta',
 'Thursday':'quinta',
 'Friday':'sexta',
 'Saturday':'sabado',
 'Sunday':'domingo'
 }

 def __init__(self):
     self.view_films()

 def view_films(self):
     regex = re.compile('localItem*')
     clean_tags = re.compile('<(/|)(div|p|h1|h3|b|i)(| class=".*")>')

     text_wrapper = TextWrapper()
     text_wrapper.width = 72

     page = self._url + self._days_of_week[self._today] + '.shtml'
     resp = urllib2.urlopen(page)
     html = resp.read()
     resp.close()

     for i in BeautifulSoup(''.join(html)).findAll('div'):
         try:
             if re.match(regex,i['class']):
                 formatted = text_wrapper.wrap(re.sub(clean_tags,'',i.__str__()))
                 for paragraph in formatted:
                     print paragraph.decode('utf8')
                 print '\n'
         except:
             pass

if __name__ == "__main__":
    Films()

Bye.

Ubuntu Contributor

As a eventual contributor to Ubuntu [1] at Launchpad [2][4], sometimes I contribute with English->Portuguese translations. Well, sometimes I use a good tool known as Open-Tran [3], written by Jacek Sliwerski, to help me with technical or language specificities on translation.

The beautiful of this tool (open-tran) is the Python xml-rpc support. Then, translation can be done in this way:

#!/usr/bin/python
# -*- coding: utf-8 -*-
import sys
from xmlrpclib import ServerProxy

lang = "pt_br"
server = ServerProxy("http://open-tran.eu/RPC2")

if __name__ == "__main__":
    if sys.argv.__len__() == 2:
        for sugg in server.suggest(sys.argv[1],lang):
            print sugg
    else:
        print "Usage: %s <words_to_translate>" % sys.argv[0]
        sys.exit(1)

    sys.exit(0)

Now, I can translate a text:

translate.py "I feel happy"

Easily, yeah ?!

If you are a Brazilian and wants to help Ubuntu and Open Source cause, you can do it through Launchpad. Visit the Brazilian community at http://launchpad.net/~ubuntu-l10n-pt-br

Meet at https://launchpad.net/~n1ghtcr4wler

Useful links:

1. http://www.ubuntu.com/
2. http://launchpad.net/
3. http://open-tran.eu/
4. http://en.wikipedia.org/wiki/Launchpad_(website)