This is the repository for the ARDC Vocabulary Registry. See also the repositories for the corresponding Portal, Linked Data API, and the documentation repository.
The Registry implements a vocabulary metadata catalogue and repository, and provides an API for a vocabulary portal and for machine-to-machine clients.
Please refer to INSTALL.md for detailed installation instructions.
The Registry should be deployed in Tomcat. (Deployment to other containers may be possible; it has not been tested.)
The Registry combines the functions of a metadata catalogue and a repository. The repository functions were previously undertaken by a project known as the Vocabulary Toolkit; those functions are now fully integrated into this Registry.
This is an overview of the repository functions:
- Get project metadata from PoolParty
- Harvest
- PoolParty
- Sesame
- SPARQL endpoint
- File
- Transform
- Extract a list of SKOS concepts in JSON format
- Create a tree of SKOS concepts in JSON format to support a browse visualisation
- Import
- Upload RDF data into a Sesame repository
- Publish
- Create a configuration file for use with the Elda library
- Unpublish
- Unimport
- Unharvest
- Backup
- Download project exports from PoolParty
The API is accessed via HTTP(S).
This section provides some background information on the technology used to implement the Registry.
The Registry has been developed using Java 8, and a Java 8 Runtime Environment is required. Specifically, the code uses lambda expressions.
The Registry uses JPA for database access. Hibernate is used as the implementation provider for JPA. The JPA layer provides database independence. Nevertheless, because ARDC software has for a long time used MySQL, the MySQL JDBC driver is included, and the software has only been tested using MySQL.
The HTTP-based API is implemented as a set of restlets using JAX-RS. Strictly speaking, the interface is not RESTful. Jersey is used as the implementation provider for JAX-RS.
The OpenRDF Sesame libraries are used extensively to implement the Registry features.
The Registry uses Apache Commons components. Currently, Configuration, IO, and Lang (both versions 2 and 3) are used. (Commons IO is required by Sesame, but it is also invoked directly from the Registry code.)
The Registry exposes its API methods through Swagger (also known as
Open API version 2). The default web.xml exposes the Swagger
description of the API at the URL /swagger.json.
The Vocabulary Registry is licensed under the Apache License,
version 2.0. See LICENSE for details.