Note: where "SPARQL store" is used here, it refers to an online (or local) RDF store with SPARQL 1.1 Update facilities. Such a store is Apache Jena Fuseki, referred to here as "Fuseki"
Schema Editor is a single-page browser-based web application designed to allow rapid creation and editing of RDF Schemas (also know as ontologies and vocabularies etc.). It is not a general-purpose RDF editor. It should work in any modern browser with Javascript enabled.
The emphasis is on creating schemas with good documentary support through the use of multi-lingual labels and comments. Standard RDFS terms are used and so subsequent inference across created schemas is possible, though this aspect isn't treated as a priority.
The editor operates against a SPARQL store, where data is persisted (subject to the configuration of the store).
Warning: if pre-existing schemas are edited, any statements involving terms outside of core RDFS may be lost.
To run Schema Editor, the following are required:
- HTTP Server
- SPARQL 1.1 Update-capable online RDF store
This repository includes an npm (node) install script that will install a minimal HTTP server and a suitable SPARQL store.
If a HTTP server and SPARQL store are already available, all that is necessary for installation is firstly to clone this repository and copy the src/public directory (with subdirectories) onto the target machine and serve as web pages.
You should then follow the instructions for Uploading Prefixes Map under Running Schema Editor below.
If facilities aren't already available, the following procedure can be followed to build a minimal HTTP server and install a SPARQL store.
These instructions should work on most Linux distros although they have only been tested on Ubuntu. There's nothing too complicated and installation should be straightforward on any OS.
- Java 8+ runtime (for Fuseki) *
- node.js (for npm-based installation and minimal HTTP server)
_ * (The version of Fuseki referred to in the install script can be built from source with Java 7, but the pre-built snapshot binaries require v. 8)_
On a bare Ubuntu machine these may be installed using:
sudo apt-get update
sudo apt-get install openjdk-8-jre
sudo apt-get install node
You should be logged into the target machine as a non-root user, and in a terminal enter:
git clone https://github.com/zazuko/schema-editor.git
cd schema-editor/
npm install
This should install Fuseki under data/store together with the depencies for the mini HTTP server.
If you have installed as above, then the command:
npm start
will set Fuseki and the mini HTTP server running.
npm stop
will stop Fuseki & the HTTP server.
Warning: npm stop is rough & ready, it's unlikely but possible that it might kill other processes
If you now point a browser at:
http://localhost:8888/
You should see the Schema Editor.
When you first see this, you will be presented with a dialog, Endpoint Settings. You should enter the appropriate values for the SPARQL store you wish to address.
If you've installed Fuseki as above, the required values will be:
http://localhost:3333/schema-edit/sparql
http://localhost:3333/schema-edit/update
On a remote server, replace localhost with the domain name or IP address of that server.
To enable the use of well-known namespace prefixes within Schema Editor (e.g. rdf: => http://www.w3.org/1999/02/22-rdf-syntax-ns#), the file data/prefixes should be uploaded to the store with the graph name http://purl.org/stuff/prefix/
This may be achieved using whatever tools are available for an existing SPARQL store, or using Schema Editor as follows:
- scroll down to Upload RDF
- enter Graph Name : http://purl.org/stuff/prefix/
- click Choose File and navigate to data/prefixes.ttl
All the parts of Schema Editor have associated help information. When the mouse cursor is over a heading it will change into a (?). Clicking will reveal the help text.
To create a new schema, click New Schema and fill in the fields.
The namespace of the schema will be used as the graph name within the SPARQL store. It must end with a "#" or "/" character.
This field will show the graph name/namespace of the schema currently being edited. The drop-down list to the right of the text field will show a list of other graphs available in the store. If you wish to edit one of these, select it.
This field will show the resource currently being edited, typically a term (class or property) in the current schema. You can select a different resource using the drop-down list or by clicking on a link to the required class/property in the left-hand column.
Clicking on Delete will remove the current resource from the schema (all triples with this resource as subject).
These are the instances of rdfs:Class and rdf:Property in the current graph. Clicking on any of these will bring them into focus in the editor.
When a class or property has been selected its details will be loaded into the fields here.
Fields with a purple border are for resources. Their values may be set in any of three ways:
- name - e.g. Cat - this will be interpreted as being in the current graph/namespace
- namespace:name (CURIE) - e.g. foaf:name - if the prefix is one of the well-know values loaded in the store (or that of the current schema), it will be interpreted as being in the corresponding well-know namespace
- full URI - e.g. http://xmlns.com/foaf/0.1/
To the right of the editable fields are two kinds of buttons:
- + - clicking on one of these will add an extra field for the same property
- lang or e.g. en - these display the language associate with the literal field (lang indicates no value has been set). Clicking on one of these will launch the Choose Language dialog
On the left you will see a list of currently available language tags. Selecting one of these and clicking Set Language will associate the tag with the literal field. If you don't see a language with which you wish to tag the literal field, you may enter its tag in the Add Language field and clicking the Add Language button will add it to the list. Select it and click Set Language to associate the tag with the literal field.
Clicking the Update button will push your changes to the SPARQL store.
Warning: if you are editing a pre-existing schema that has been uploaded to the store, any properties not seen in the editor will be wiped (Update initially wipes all properties before adding those in the fields)
To add a new class or property to the current schema, enter its name in the appropriate edit field and click Create. The term will be initialised in the store and you will be presented with the editing facilities for the term as described above.
To edit an existing schema, first enter its namespace/graph name in the corresponding field, then click Choose File and navigate to the required file. There may be a few seconds delay when uploading larger schemas (e.g. schema.org), please be patient. Currently only Turtle format files are supported.
Clicking on this button will reveal a Turtle representation of the current schema. You can save this by right-clicking and selecting Save As... (or whatever corresponds to that in your browser). The schema will remain in the store for future editing.
To address a different SPARQL store, enter its details here.
Unit tests may be run by clicking on this link.
TODO
- check rendering of this at github
- add notes about separate scripts
- write up how it works, how to extend etc.
Apache 2.0 License (c) Zazuko 2015, 2016