JISC IE Metadata Schema Registry

Introduction

The JISC IE Metadata Schema Registry (IEMSR) website software is a web-based user interface to the data stored in the IEMSR data server. It provides a view over the metadata information stored in the data server, such as created by the desktop schema creation tool.

Download Website Server software

The SourceForge schemas project hosts the CVS repository and provides live browsing of the CVS for the source code and checking out of IEMSR sources from CVS.

Releases can be found the files area

Running the Website Server

The WebSite Server software is a java web application, downloadable as a .war file so that it may be deployed directly into the Apache Tomcat servlet container as a web application.

There is a demo of the release dated 2005-06-30 running at the IEMSR website.

This software has been developed and deployed using these technologies:

• Java - tested with Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2_06-b03)
• Apache Tomcat - a servlet container. For more information see http://jakarta.apache.org/tomcat/. This application has been tested with Tomcat versions 4.1.31 and 5.0.x.
• Struts - http://struts.apache.org/userGuide/. The required libraries are included in the downloadable .war file.
• Struts tiles - http://struts.apache.org/userGuide/dev_tiles.html. The required libraries are included in the downloadable .war file.
• Java Server Pages (JSP)
• JavaBeans

It also uses these standards:

• XML
• SPARQL (the query language for RDF)
• Model-View-Controller (MVC) Architecture (as implemented via Struts)

The application source code may be found in the webapp's /WEB-INF/source directory of the software release. Struts libraries and other .jar files required are included in the /WEB-INF/lib directory. The Apache Software License applies.

Further technical information may be found in the README text file included with the software.

Todos

Development work

• Finish what is currently partial support for free text searching - currently there is a mixture of supported and unsupported searches which all combine a text string with the type of object being searched for. For further developer's information see the README that comes with the latest version of the website application software.
  See 3.14 HOWTO: Alter the SPARQL queries used to get data from the IEMSR Registry server (or configure Free Text queries
• Add support for new SPARQL results syntax. The current software version supports first version only.
• Complete the support for XML binding schemas when fuller data is in the Data Server.
• Complete the (almost finished) query support for RDF Data Sources that refer to IEMSR 'things' other than just DCAPs (and configure view layer appropriately).
• VIEW LAYER - refering to sections in IEMSR website view mockups by Pete Johnston ( PowerPoint or PDF)
  Includes:
  • Add SORT. e.g. for DCAPs by name, title etc. and sorting for other views required too.
  • Display of e.g. LOMAPs/DCAPs - - complete XML Binding schema (expressed by) and match the view of the LOMAP to the website spec.
  • Alter the drop down list options for searching on a per-tab basis. For further developer's information see section [3] 10. "HOWTO: DROP DOWN LIST: change the options that appear in the Search drop-down list in any one tab view" in the README that comes with the latest version of the website application software.
  • On the DC tab add:
    
    • Browse Properties for resources of type rdf:Property
    • Browse Encoding Schemes for those classes that are the objects of iesmsr:encodingScheme
  • Add data to the XML Schema Bindings Tab.
  • Show Term Status for the browse view of DC properties (as per currently shown for DC classes) in the view of a DC Metadata Vocabulary. Also show status of DC Metadata Vocabularies
• Develop a more sophisticated free text query capability on the Data Server and adapt the website client to this.
• DC Vocabulary - increase the number of typed 'things' that a DC vocab can contain additional to properties and classes, i.e. Vocabularies - DC/non-LOM such as show in this example query using old website software.
• Make the view of a LOM Data Element also show links to its child elements as well as parent element.
• Implement load up of data on startup in the servlet container (Tomcat).
• LOMAP: to assist end-user orientation/navigation, implement storing which LOMAP is being currently browsed when browsing through lom data elements and provide some sort of visual image to indicate where in the tree hierarchy of LOM Data Elements the end-user view is conceptually positioned
• Harmonise terms with IEMSR desktop tool.
• Make the refresh button refresh whichever tab page user currently in rather than always revert to Agency tab.

Known minor bugs

• (application) Uri for JORUM DC Application Profile - leads to error in website, check why.
• (data) Agencies that publish non-LOM vocabularies are not found in the data in the Data Server.
• (application) In general check all SPARQL queries to the Data Server retrieve unique results rather than containing too many 'OPTIONAL' clauses (for further developer's information see section 3.14."HOWTO: Alter the SPARQL queries used to get data from the IEMSR Registry server (or configure Free Text queries)")
• Parse the users search string for speech marks and other similar characters.
• (DatabaseService.java) getallRDFClasses for a DC Metadata Vocabulary - implement this using the concept URI. It currently searches by the name of a DC Metadata Vocabulary which is slow.