I can't answer for the Oracle-based III version, but in the proprietary (III Classic as I call it) you can - if you have it installed - use a telnet client to search for records, save the records to a file, and then FTP them to an external server. That's how we're getting records out of our Millennium system for ingestion into XC. In fact, I have written a perl/expect script to do that on an automated basis that i would be willing to share with anybody who's interested.
I'm currently exploring how we can use the Millennium Java client to do the same thing (if anybody knows how to do that, I would love to hear because we don't want to be depending on the telnet client since it may go away someday).
Digital Information Systems Librarian
University of North Carolina at Charlotte
9201 University City Blvd.
Charlotte, NC 28211
From: Code for Libraries on behalf of Eric Lease Morgan
Sent: Fri 2/18/2011 10:48 AM
To: [log in to unmask]
Subject: [CODE4LIB] exporting marc records from iii
How does a person go about exporting MARC records from a III system?
As you may or may not know, I spend a lot of my time developing a thing colloquially called the "Catholic Portal". It uses VUFind under the hood, and it requires me to ingest bibliographic data from a myriad of libraries.
Suppose the records I desire have the letters "crra" saved in MARC field 590$a. What is the process for connecting a III system, searching for "crra" in 590$a, and saving the result as a file of MARC records? Is there some sort of documentation I can read that will help me out in this regard?
Eric Lease Morgan