Hannibal
Hannibal is a PHP7 adaptive engine for 3D Web digital heritage artefacts in Web-Based Virtual Museums (WBVMs). It uses also Python to execute a lot of needed processes such as for decimating 3D models, and for communicating with a backend (Omeka DAMS). It uses a forward chaining expert system hooked to a MySQL Database Management System (DBMS). Hannibal is a middleware that works with Omeka Digital Asset Management System (DAMS) as a backend and with a leaflet JS map-based Web-Based Virtual Museum on the front end.
Hannibal is presented in repository in a way that allows you to use it in your virtual museum and with your backend. You can also alter it to be used in any adaptive application that requires doing an action based on a set of decisions. Hannibal has a component that decimates an uploaded 3D model into lower fixed resolutions and creates a 360 image sprite of the model. But this component is shared in the form of seperate repositories in BlenderPythonDecimator and 3DMeshBulkSimplification , TransformMeshToGIFSprite.
The adaptive engine is named after Hannibal Barca, the hero of the Phoenician Carthaginians who was himself adaptive and cunning in his war tactics against the Romans during the Punic wars.
Hannibal was integrated and tested in a small subset clone of a WBVM (for the Shetland musuem (Scotland)) built for the EULAC Web-Based Virtual Museum initiative, a complete web-based virtual museum infrastructure developed by researchers in the Open Virtual Worlds research group at the University of St Andrews, that aims to replicate all the functions of traditional museums on the web providing curators a management interface where they can upload digitised 3D models and other media and their metadata, in addition to providing documentation in the form of wiki articles.
Notez Bien
Due to copyrighted material the Web-Based Virtual Museum clone subset pertaining to the EULAC will not be uploaded. I do not have any permission of showing digitised material pertaining to the musuem. In addition the aim of this repository is to show the expert system and the adaptive engine (precisely the detection component of network conditions and the WebGL Benchmark component).
Requirements
You need a Linux OS (probably you can make things work in other OSs)
The following are configurations done on a Ubuntu Server 16.04.X I would assume you have Apache Web server. If not install the following packages:
sudo apt-get install apache2 apache2-doc apache2-utils
#Install php7 (PS: it will install php7 now not php5 by default)
sudo apt-get install php
You should have MySQL on your server if not, install the following:
#For MySQL server
sudo apt-get install php7.0-mysql #for php7
sudo apt-get install mysql-server mysql-client
#Other Necessary Packages
sudo apt-get install build-essential
sudo apt-get install libncurses5-dev libncursesw5-dev libreadline6-dev
sudo apt-get install libbz2-dev libexpat1-dev liblzma-dev zlib1g-dev
Install the Python virtual environment. It is usually a better practice to install Python packages in a Python virtual environment that in the whole server system using pip of the system.
sudo apt-get install python-virtualenv
Since our WBVM is talking to Omeka so we are using the Omeka Client Python: omeka-client-py by Jim Safley. The needed Python modules to install are markdown, httplib2, pyyaml, and urllib3.
So you have to make sure Apache can have access to all Python modules required in this library. As an easy and headache-free technique is to install the Python modules as I have mention in a Python Virtual Enviornment inside your Apache /var/www/html/SOMETHING. So navigate to such directory.
virtualenv env
#This will install a small python environment inside the folder with pip and other essential modules.
#To check the pip version of the virtual environment
env/bin/pip --version
#To check the Python version of the virtual environment
env/bin/python --version
#Now activate the python virtual environment
source env/bin/activate
#Now install all the needed python modules inside the python virtual environment
env/bin/pip install markdown
env/bin/pip install httplib2
env/bin/pip install pyyaml
env/bin/pip install urllib3
I had some problems installing the EULAC WBVM containing Hannibal on many Linux servers and this was due to problems in permissions so make sure you have these set correctly.
Since both Hannibal and the WBVM talks to MySQL DBMS so make you sure you configure these correctly.
#Setting up MySQL users and root user
sudo mysql -uroot –p
#This should ask you for the password you entered before during the installation, enter it. Write this down (important).
#THEN (IMPORTANT) install phpmyadmin
sudo apt-get install phpmyadmin
This will start installing the packages. You will be asked which Web Server is to be used. Choose apache2 and follow along.
Now for the Nginx server and the folks who want to use it, Hannibal itself without its backend (backend of decimated models) works gracefully with this web server but not Omeka (the backend of Hannibal and WBVM). Omeka needs the Apache Rewrite rules to work and to access the .htaccess files. I tried to create a solution for nginx following material online but to no avail. So the only way to make the system work with with Apache web server.
For Omeka to work gracefully with Apache Web Server you need to enable mod_rewrite in Apache.
sudo a2enmod rewrite
#Restart apache2 after that for things to take effect
sudo systemctl restart apache2
You need also to change /etc/apache2/apache2.conf for effective rewrite. So have something similar to the following:
<Directory /var/www/>
Options Indexes FollowSymLinks
AllowOverride None
Require all granted
</Directory>
and change it to
<Directory /var/www/>
Options Indexes FollowSymLinks
AllowOverride All
Require all granted
</Directory>
then restart Apache
sudo service apache2 restart
Omeka Important things
Inside php.ini in apache2 the default upload size is 2MB. Obviously this does not help especially when uploading 3D models so we need to put it as high as possible especially 3D models have large sizes on disk. Assuming we are using php7
sudo nano /etc/php/7.0/apache2/php.ini
The default lines that control the file size upload are normally 2M, Change these default values to your desired maximum file upload size. It is good to put a high number such as 1024MB.
#php.ini Options (setting used by Omeka EULAC)
post_max_size = 1024M
upload_max_filesize = 1024M
Other common resource settings include the amount of memory PHP can use in a script and this is set by memory_limit:
memory_limit = 128M
or max_execution_time, which defines how many seconds a PHP process can run for:
max_execution_time = 30
When you have the php.ini file configured for your needs, save the changes, and exit the text editor. Restart the web server to enable the changes. For Apache web server on Ubuntu server 16.04, this command will restart the web server:
sudo service apache2 restart
When POSTing files to omeka (meaning using REST API via Python), you need also to install ImageMagick tool. Omeka uses ImageMagick.
To check if ImageMagick is present, check that the following gives an output.
convert –version
If not you need to install additional packages before it. Check that the gcc package is installed, then install ImageMagick itself:
sudo apt-get install imagemagick
Then install all php extensions of ImageMagick:
sudo apt-get install php-imagick #This will install extensions for PHP7 by default
Restart Apache and Verify
sudo service apache2 restart
Now verify by:
php -m | grep imagick
You should get the word ‘imagick’.
License
This program is licensed under MIT License - you are free to distribute, change, enhance and include any of the code of this application in your tools. I only expect adequate attribution and citation of this work. The attribution should include the title of the program, the author (me) and the site or the document where the program is taken from.