KohaCon12: Koha and SRU

by Kyle Hall on Jun 7, 2012

Marcel de Rooy of Rijksmuseum presented on the subject of Koha and SRU.

SRU stands for Search/Retrieval via URL. It is a standard that allows one to search a catalog and retrieve results using standard web technologies including XML for data presentation and CQL (Contextual Query Language) for creating searches.

Koha includes an SRU almost out of the box, though some very minor is still required.

Why use SRU?

SRU is meant to replace Z39.50, which Marcel believes has not aged well. It is simpler, and uses standard web technologies ( HTTP, XML, etc. ). It is also more flexible than z39.50, and is easier to interface with and develop.

Using Zebra/Yaz

IndexData developed the Zebra indexing/retrieval engine on top of the YAZ toolkit. It actually speaks both z39.50 and SRU on the same port. Koha fetches MARC records and uses XSLT to support various output formats, including MARC XML, Dublin Core, and others.

Serving data with SRU

There are many web services that can take advantage of a Koha installation running an SRU server. These including LibraryThing ( using DublinCore ), Adlib Museum ( using a custom record schema ), and Fotografiebibliotheek, a service available in the Netherlands.

There are plans to add support for copy-cataloging via SRU, similar to Koha’s existing z39.50 support, but it has yet to be implemented.