One very helpful use it can be put to is that of working out how much of a project’s source code has been covered by unit tests and I decided to try this for the Entertainer Media Center project.

What follows is how I installed and used python-coverage to work out how much of the Entertainer Source code is actually tested when all our unittests are run. I relied upon this webpage and the command line help available from python-coverage.

First of all I installed the python module from the Ubuntu repositories:

sudo apt-get install python-coverage

This installed version 2.6

I then changed directory to where all the unittests for Entertainer are stored on my computer:

cd src/tests

After testing it in a couple of ways, I then cleared any previous history with this command:

python /usr/share/python-support/python-coverage/coverage.py -e

The actual production of the statistics is done in two stages. First you run the program which will generate the stats (In this case all the unittests):

python /usr/share/python-support/python-coverage/coverage.py -x run_tests.py

This stores a history of what lines were executed. You can then compare this history to the actual source files with:

python /usr/share/python-support/python-coverage/coverage.py -r ../*.py ../*/*.py ../*/*/*.py ../*/*/*/*.py ../*/*/*/*/*.py

This gives statistics for each individual source code file and then a set of overall statistics for the whole project. I had to use the ../*’s to make sure all the source code was used.

The overall stats that were generated for the Entertainer Project is as follows:

...
../utils/log_viewer 114 0 0%
../utils/lyrics_downloader 80 19 23%
../utils/open_feed_source_dialog 77 0 0%
../utils/preferences_dialog 200 0 0%
../utils/resolve_dialog 64 0 0%
../utils/system_tray_icon 95 0 0%
../utils/theme 88 0 0%
../utils/thumbnailer 31 27 87%
../utils/video_thumbnailer 226 126 55%
../utils/weather 106 100 94%
----------------------------------------------------------------------------------
TOTAL 13195 2206 16%

This means that out of a total of 13,195 lines of python code only 16%, 2,206 lines, is being tested by the 145 unittests that have been written so far.

Obviously increasing the coverage of the unittests will make the Entertainer much better at pinpointing bugs and fixing breaks in code as the project is developed.

At the moment I’ve add Challenge, Entry and Item classes and it can now load and save challenge data in xml format.

The next step will be added a ChallengeManager class to manager the insertion of data and automatic posting to www.blenderchallenge.com and blenderartists.org

You can comment this in Doxygen format in the following way:

class Demo:
"""
This is a description of the Demo class
@author Joshua Scotton
@version 0.1
"""
  def theDemoTest(self, testParameter):
  """
  This is the very, very, very, very, very, very, very, very, very, very long description of theDemoTest!
  @brief This is where you can put a brief description
  @see Demo
  @param testParameter String: Input Parameter is a string
  @return int Returns a 0 if successful
  """
    #do something here
    return 0


Quick Tag List
@author
@brief
@bug
@code and @endcode
@date
@file
@package
@param
@return
@see
@todo
@version
@warning

I was recently generating some documentation from the Entertainer source code and decided to share how I managed to do it.

I’m assuming python 2.4 is installed on your system.

First install doxygen and graphviz

sudo apt-get install doxygen graphviz

Download the latest version of doxypy from http://code.foosel.org/doxypy and extract to a suitable location.

I used doxypy-0.3rc2.tar.gz

Now we need to generate a configuration file, to do this run the following from your project root folder

doxygen -g config.dox

Edit the file and set the following tags:

PROJECT_NAME = #Your Project Name would go here
PROJECT_NUMBER = #Your version/revision number goes here
OUTPUT_DIRECTORY = #Where you want the documentation going
FILE_PATTERNS = *.py
RECURSIVE = YES
INPUT_FILTER = "python /path/to/doxypy.py" #set this to where you extracted doxypy
FILTER_SOURCE_FILES = YES
INPUT = #where the source files are

If you want graphs also edit the following:

HIDE_UNDOC_RELATIONS = NO
HAVE_DOT = YES
CALL_GRAPH = YES
CALLER_GRAPH = YES

Run the following:

whereis dot
and enter the output into the following config tag:

DOT_PATH = /usr/bin/dot

Next ones are optional:

MULTILINE_CPP_IS_BRIEF = YES
DETAILS_AT_TOP = YES
EXTRACT_ALL = YES
EXTRACT_STATIC = YES
SHOW_DIRECTORIES = YES
SOURCE_BROWSER = YES
ALPHABETICAL_INDEX = YES
COLS_IN_ALPHA_INDEX = 8
TOC_EXPAND = YES
DISABLE_INDEX = YES
GENERATE_TREEVIEW = YES
GENERATE_LATEX =YES

All set up now…

All we need to do now whenever you want to generate/regenerate the documentation is to run the following from the project root folder:

doxygen config.dox

Thanks to internetducttape.com as I used this blog entry as a starting point.

I’ve just joined the Entertainer Project  over on Launchpad.

From the website:

“Entertainer aims to be a simple and easy-to-use media center solution for Gnome and XFce desktop environments. Entertainer is written completely in Python using object-oriented programming paradigm. It uses GStreamer multimedia framework for multimedia playback. User Interface is implemented with Clutter UI-library, which allows sleek OpenGL animated user interfaces. Entertainer also uses other great projects like SQLite, pyIMDBb and iNotify.

The ultimate goal of the project is to create the best media center solution available for any platform. This means that Entertainer should be the best looking, most easy to use and most feature filled media center solution available. Not the easiest goal to achieve. It’s a long way, but we have a good start here!”

My first job is to add opml import capability to the rss feed reader and allow the option to sync with liferea. I’ll be checking out any bugs relating to this as well.

This can be used to import a number of reference images into a scene as an aid to modelling.

V1.0

* imports any pictures with the extension .jpg
* resizes planes to 2 blender units per 1000 pixels

V1.1

* implements changes proposed by migius on blenderartists to work on windows xp
* can choose any extension
* can choose the scale

V1.2

* cosmetic changes

Version 1.2: importrefimages12.py