This project is developed in Ontotext as a part of the MOLTO Project's WP2.
It is an Eclipse plugin that is part of the Grammatical Framework plugin for Eclipse (GFEP) but can be installed and used on its own. The GF Ontology Grammar Plugin provides a GUI for calls to the MOLTO Repository Helper projects' code. It is under the form of a wizard that leads a user to create abstract, concrete and SPARQL GF grammar, based on user-friendly templates that are filled in with repository entities by the user. The wizard connects to a SPARQL endpoint and uses the ontology provided by it. Then the user has to provide a specific templates file. Next, pages for filling the templates are available. Finally, the user is prompted to export the grammars created from the templates. The exported grammars can then be edited by the GF Eclipse plugin.
The code of the plugin is publicly available at Github in two places:
- The main GFEP repository which contains code used in the released version of the plugin.
- The forked GF Ontology Grammar Plugin repository which contains the most up-to-date code.
To see how to install git, check out this page.
To "checkout" the project as read-only, you just need to perform:
git clone https://github.com/mmateva/gf-eclipse-plugin/
If you would like to contribute to the project at Github, create a Github account and follow the instructions here. A basic Git explanation is introduced on this page. Generally, all you need is on the Github help pages.
We will refer to the project's home(Git's checkout directory) as $GFEP_HOME.
Basically, what you need to do is create an Eclipse workspace(can be the default one), and in the Project Explorer perform Import → General → Existing Project into Workspace, then via the file browser select the folder
$GFEP_HOME/gf-eclipse-plugin/workspace
The result is that you get (currently) six projects that you can import in your Eclipse workspace. The one that is described in this manual is org.grammaticalframework.eclipse.ontology-grammar.
You can optionally import the rest as well, they are part of the more general GF Eclipse plugin.
After you have imported the plugin project, you can directly run it in Eclipse.
Just select the plugin.xml file in the Project Explorer, right-click it, and Run As → Eclipse Application.
Eclipse will run another Eclipse within itself, with its own classpath and Running Configuration.
The VM parameters are defined through Run As → Run Configuration, all the rest is described via the plugin's plugin.xml, build.properties, Manifest.MF; details on these.
The "child" Eclipse instance has the GF Ontology Grammar Plugin installed.
You can start the plugin from it.
See below how to Run the plugin in Eclipse.
Eclipse plugins are very well organized standalone modules that can be added to the Eclipse IDE and serve for various purposes. Generally, they are any java code that is bound to the Eclipse IDE through extension points. The two most important descriptor files are the plugin.xml and the Manifest.MF. Eclipse PDT provides a handy editor for them. More about writing Eclipse plugins can be found at Eclipse Plugins Developer's guide and on this tutorial.
Currently the plugin is not available for download from the Eclipse Plugin management pages. Hopefully, It will be soon.
Eclipse 3.7 or above is recommended
Navigate to File → New → Other... (or use Ctrl+N) from the Eclipse GUI. Write "gf" in the search box, or scroll down to the "Grammatical Framework" category. Select "GF Ontology Grammar Wizard".
Provide a SPARQL endpoint to be connected to(ontology repository).
After the username and password are populated, use the "Connect" button to verify the connection is successful.
Note: username and password can be omitted; example Repository URL = http://molto.ontotext.com/sparql
Select a templates from the file system. An example for such a file is provided in the resources folder of the plugin.
It is automatically validated, but you can still use the "Validate template" button to check if it is valid. Then you can proceed to the next page via the "Next" button.
Select a single template from the list - as on the screenshot. The binding in square brackets, e.g. [CLASS_INSTANCE], [CLASS_NAME], will be populated on the next page.
On the page for entities selection, currently one has "Classes" and "Instances". The instances on the left are the corresponding ones to the classes on the right(they are the instances of the selected class). If a class has no instances, then an [EMPTY LIST] note is shown. The template is automatically populated with a matching binding when one is selected(see the screenshot below).
If you constantly get the "LOADING" sign in the Instances box, you might need to consider a better connection speed to your SPARQL endpoint.
Use the "Add Template" button to add new templates.
Also, you can "Reset" the template, to see the binding type that is populated.
When enough templates, you can either continue with adding new template patterns("Select more templates"), or save the grammar created to a file("Export Grammar"). The first option actually brings to 4.
The grammar exportpage simply gives you the opportunity to choose where to store the ontology grammar files. After selecting the destination, you need to press the "Export Grammar" button.
This will create the following three grammar files in the destination folder:
Abstract<GrammarName>.gf- abstract ontology grammar that was createdConcrete<GrammarName>.gf- concrete ontology grammar that was created<GrammarName>SPARQL.gf- SPARQL query grammar that was created
The MOLTO Repository Helper is developed in Ontotext, also as part of the Molto project. It provides an API for creation of abstract, concrete and SPARQL GF grammars. They are based on user-friendly templates that are filled in with repository entities.