Skip to content

Latest commit

 

History

History
141 lines (107 loc) · 4.71 KB

README.md

File metadata and controls

141 lines (107 loc) · 4.71 KB
Raw

OWL/Turtle to HTTP

It's really what the name says: you can make an HTTP interface for the URIs of an ontology.

There are two main scripts in this repo:

  • server: a Node.js server that dynamically content-negotiates your turtle file in application/json, text/turtle, and text/html
  • serve: a generation script that uses your turtle file to generate the files in application/json, text/turtle, application/rdf+xml and text/html, and accompanying .htaccess to support content negotiation.

The latter needs an additional Apache web server to actually serve your files.

installing

Be sure to have nodejs installed. If you install nodejs from package manager consider https://github.com/joyent/node/wiki/Installing-Node.js-via-package-manager

cd owlturtle2http/
npm install

executing

server: Running a node.js server

./server

After this, the server will be running on port 8080. You can change this in the file server

serve: Generating all files

./serve -h
usage: serve [-h] [-v] [-o ONTOLOGY] [-a] [-i]

Serving your turtle ontologies

Optional arguments:
  -h, --help            Show this help message and exit.
  -v, --version         Show program's version number and exit.
  -o ONTOLOGY, --ontology ONTOLOGY
                        label of the turtle file you want to serve, as
                        specified in the config file. If provided, the
                        version can also be added (e.g., "apps4X:0.4", if so,
                        the turtle file will also be published under apps4X/0.
                        4/. The name of the turtle CANNOT be `assets` (as all
                        site assets reside in the `assets`-folder)
  -a, --all             (re-)serve all turtle files in the config file
  -i, --index           (re-)render the index.html file. Don't do this if you
                        have your own index.html file! PS: this actually
                        tries to be smart, and links to all subfolders that
                        have a .htaccess file in them, so is not necessarily
                        depending on config.files

Configuration

config.example.json shows an exemplary configuration. The actual config file should be config.json. It has three parts:

  • outpath: the file path (relative or absolute) to where to generate you static files to. This could be, e.g., the root of your web server, or a subfolder thereof (also virtual hosts are possible, see below). For basic (UNIX) configurations, this is usually /var/www/html.
  • namespacebase: the url path (relative) to where your namespaces will reside for serve (for server, this is by default /ns, but can be changed in the file). E.g., if you publish all your ontologies on http://example.com/ns, namespacebase should be /ns.
  • files: a dictionary of your ontologies: their labels and the (relative or absolute) path to their turtle file. These labels need to be used for the -o input parameter for ./serve. Both serve and server will use these labels as the shortname of the ontologies you want to publish, e.g., the ontology with shortname apps4X will be generated under {outpath}/{namespacebase}/apps4X in serve, or served under /ns/apps4X in ./server.

Demo

see for example: http://semweb.mmlab.be/ns/apps4X

Or ./server on your own install: http://localhost:8080/ns/apps4X/

Configuring apache

server

a2enmod proxy_http
a2enmod proxy
# Insert These 2 lines in your sites_available/000-default 
#> ProxyRequests On
#> ProxyPassReverse /      http://localhost:8080/
# Restart your apache
sudo service apache restart

Add this .htaccess in the right directory:

 Options +FollowSymLinks -Indexes -MultiViews
    
 <IfModule mod_rewrite.c>
        
 	RewriteEngine on
        
        # Simple URL redirect:
        RewriteCond %{REQUEST_FILENAME} !-d
	RewriteCond %{REQUEST_FILENAME} !-f
        RewriteRule ^(.*) http://localhost:8080/ns/$1 [P]

</IfModule>

server

You can set config.outpath to (a subfolder of) the root path of your web server and be done with it. But, e.g., if you want your namespace to reside on a subdomain (e.g., http://ns.example.com), you can use following VirtualHost Configuration:

<VirtualHost *:80>
    DocumentRoot "[absolute outpath]"
    ServerName ns.localhost
    ErrorLog "logs/webgrind-error.log"
    CustomLog "logs/webgrind-access.log" common
    <Directory "[absolute outpath]">
        Options Indexes FollowSymLinks Includes ExecCGI
        AllowOverride All
        Require all granted
    </Directory>
    LogLevel warn
</VirtualHost>

License

MIT