On Jul 28, 2004, at 9:05 AM, Eric Lease Morgan wrote:

> On the other hand, Ed Summers, who is participating in the NSF OCKHAM
> Project, is in the beginnings of writing an SRU client/server module
> in Perl. Included in this work will be some sort of CQL parser, again,
> written in Perl. Ed, will you please elaborate on the work you've done
> to date and what you will be doing in Phase II?

As you may or may not know, Ed Summers (author of Net::OAI::Harvester,
significant contributor to MARC::Record, and Perl programmer
extrordinaire) has been working with us at Notre Dame to implement the
OCKHAM alerting service where SRU plays a significant role. He has all
but completed Phase I of the task -- harvesting data from OAI providers
as well as sets of MARC records.

In short, Ed has created a set of Perl modules called Ockham::Alert.
The modules do a number of things:

   * Given the URLs of OAI repositories and dates, the modules harvest
OAI data no older than the given date and save the data to a central
store (relational database of just about any flavor).

   * Given an XML file of his design, the modules import data and save
it to the central store too. Implementors are expected to convert local
files, such a MARC records, into these XML files to support "new" items
from library catalogs or other databases. (We might move this XML
format to a more standard XML flavor such as RDF, METS, or something
else.)

   * The modules can send their content to two different applications
(plucine or swish-e) for indexing.

The next step, Phase II, will be the creation of Perl modules
implementing an SRU client/server interface to the indexes. This will
also involve the creation of a Perl module parsing CQL queries.

Phase III, will be the creation of scripts gluing the results of Phase
I and Phase II into end-user and administrative interfaces.

Ed is on time and under budget. Good work!

--
Eric Lease Morgan
University Libraries of Notre Dame