What is SRW/U?

Executive Summary

SRW/U is an acronym for Search/Retrieve via the Web or URL and you might want to think of it as Sonne of Z39.50 sans the federated searching.

What It Is

Z39.50 is (was) a protocol -- a communication method -- for information retrieval. Originally designed to allow for the searching of remote databases, and while rather arcane, it truly was ahead of its time. (Much like the MARC record data structure, arcane but ahead if its time.)

With the advent of the Internet, the ubiquitous nature of Web browsers, and the development of XML, Z39.50 began to show more of its flaws. Enter SRW/U. Instead of opening a telnet connection to remote computers, SRW/U is usually communicated via the Web (HTTP). Instead of a dialog of commands sent back and forth between two computers resulting in a stream of relatively unstructured data, SRW/U sends a single URL to a remote computer and the remote computer sends back an XML stream. Because the XML is usually returned to a Web browser, and since Web browsers are able to "transform" XML into HTML with a built-in technology called XSLT, the search results are displayed to the user.

In many ways SRW/U is much simpler than Z39.50. It only requires a relatively modern Web browser for the user and an SRW/U server to search an index. All the other technologies come along for free.

What Can Be Done With It

SRW/U is used for searching indexes (, and to some degree databases as well). It is possible to use it to search just about any list of stuff you have. Books. Journals. Articles. Microforms. Reference sources. Names. Addresses. Telephone numbers. Internet resources. TEI, EAD, and other XML files. Etc.

SRW/U client "applications" are simply HTML forms. Characteristics of the SRW/U interactions are easy to specify in the HTML forms. Very little configuration. One of the allowable configurations is the specification of an XSLT stylesheet. This stylesheet will determine how content will be displayed in the Web browser. One stylesheet could display things in Spanish while another displays things in English. One can employ large fonts and another smaller fonts.

The SRW/U query language is extraordinarily expressive, if not too expressive. Using this query language (Common Query Language or CQL) it is possible to search for the most exact thing in any index as long as the index supports that kind of searching. ("Librarians like to search. Everybody else likes to find.")

SRW/U uses modern technology to expose indexes and databases to the Web.

Examples

Copac - the closest thing to a union catalog for the United Kingdom, includes an SRW/U interface. The following URL searches for and returns a list of three books whose metadata contains the word origami:

http://copac.ac.uk:3000/copac?query=origami&operation=searchRetrieve&version=1.1&maximumRecords=3

Ockham Alert - This NSF-sponsored digital library project regularly harvests content from other NSF projects while rotating off older materials. By encapsulating queries as RSS feeds, people can use Ockham Alert as current awareness service. Behind the scenes, Ockham Alert uses XSLT to transform search results from SRW/U XML into RSS. Try this search for origami. The result should be an RSS feed:

http://alert.ockham.org/?cmd=search&query=origami&output=RSS

University of Notre Dame institutional repository - This fledgling institutional repository of the University of Notre Dame aggregates content from three OAI data repositories, indexes the metadata, and provides an SRW/U interface to the index. The following URL points to a simple HTML form acting as an SRW/U client. View the source to see how simple it is (especially compared to Z39.50):

http://dewey.library.nd.edu/morgan/idr/sru/client.html

Who Should Be Using It

Anybody who provides access to indexes, and anybody who wants to participate in "Web 2.0" should be using SRW/U.

SRW/U is an example of Web 2.0 technology. It exploits HTTP, Web browsers, and XML as a means to disseminate and access data and information. The HTTP/Web browser combination is everywhere and owned by everyone. The use of XML effectively divorces content from display allowing the data to be used in many and unimaginable ways. ("I don't need your interface. Just give me the data.") The use of SRW/U makes it possible for you to disseminate your content (access to your indexes and databases) in a syndicated manner. It is not about people coming to your website, but rather providing ways of making your content available in the user's spaces.

Related Technologies

OpenSearch is a similar searching protocol. The query language is less expressive and the XML returned by the server is in the form of RSS by default. See: http://opensearch.a9.com/

More Information

The canonical homepage of SRW/U - http://www.loc.gov/standards/sru/.

A more thorough introduction to SRW/U - http://www.ariadne.ac.uk/issue40/morgan/.


Creator: Eric Lease Morgan <eric_morgan@infomotions.com>
Source: This essay was originally published on TechEssence at http://techessence.info/node/48.
Date created: 2006-05-24
Date updated: 2007-12-28
Subject(s): TechEssence; SRU (Search/Retrieve URL Service);
URL: http://infomotions.com/musings/what-is-srwu/