Re: [CODE4LIB] marc2oai
On May 29, 2007, at 8:23 PM, Conal Tuohy wrote: I don't have an answer to your question as such, but ... if you check the OAI Registry at UIUC, you can see there are hundreds of repositories providing a few different flavours of MARC: http://gita.grainger.uiuc.edu/registry/ListSchemas.asp This is a great pointer. I provides a lot of food for thought, and there is no reason to re-invent the wheel. Thank you. code4lib++ -- Eric Morgan
[CODE4LIB] marc2oai
Does anybody here know of a MARC2OAI program? A few days ago on the NGC4Lib mailing list there was some discussion on how authority (as well as bibliographic) records could be shared among libraries. One answer, proposed by Stephens Owen, was OAI. After thinking about it a bit more, I think this is entirely feasible. For example, such a system could work like this: 1. Dump MARC records to a file system. 2. Ingest records into a simple database with fields such as local key, foreign key, date stamp, deleted record flag, and record in communications format. To implement OAI sets, the database may be normalized into a small number of tables for things like author names, series, formats, and main subject headings. 3. Implement an OAI interface against the database complete with Dublin Core, MARCXML, MODS metadata schemas. 4. Provide a means to mark database records as deleted as well as to add additional records. 5. Go to Step #1. Given the existence of N number of such implementations, libraries could then do a number of things such as: * Harvest their own content, index it, and provide rudimentary access to the catalog. * Harvest other library's content, index it, and create a union catalog. * Harvest selected content from many repositories, index it, and create a topic-specific catalog. * Harvest content, combine it with other OAI-accessible content, and create a more than just books catalog. * Harvest content, index it, search it, and download a MARC record in communications format facilitating a sort of copy cataloging process. So, do you know of any software in existence that makes it relatively trivial to make a set of MARC records in communications format accessible as an OAI data provider? -- Eric Lease Morgan University Libraries of Notre Dame (574) 631-8604
Re: [CODE4LIB] marc2oai
-Original Message- From: Code for Libraries [mailto:[EMAIL PROTECTED] On Behalf Of Eric Lease Morgan Sent: Tuesday, May 29, 2007 1:53 PM To: CODE4LIB@listserv.nd.edu Subject: [CODE4LIB] marc2oai Does anybody here know of a MARC2OAI program? Eric, I have a small script that does this, it is fairly quite simple. Probably about 100 lines of code or so. I have a nightly cron script that gets any new/modified marc records from the past 24 hours out of the catalog and then runs marc2xml on the dump file. Then I have a small script that breaks up the large marcxml files into individual xml files and imports them into SOLR! I then can use an XSL stylesheet such as the LOC's marc2oai to produce an OAI document or the marc2rdf, etc on the full marcxml files (since solr doesn't have the original record). I have yet to incorporate my OAI server code into this, but since it is already written, it would be a fairly easy merge. This is all built into my NextGen OPAC that I am working on and hope to open-source sometime this summer. So sorry, im not allowed to hand out the code just yet :( Thanks Andrew
Re: [CODE4LIB] marc2oai
Well, that's an impressive teaser, anyway, Andrew Looking forward to your release! On 5/29/07, Andrew Nagy [EMAIL PROTECTED] wrote: -Original Message- From: Code for Libraries [mailto:[EMAIL PROTECTED] On Behalf Of Eric Lease Morgan Sent: Tuesday, May 29, 2007 1:53 PM To: CODE4LIB@listserv.nd.edu Subject: [CODE4LIB] marc2oai Does anybody here know of a MARC2OAI program? Eric, I have a small script that does this, it is fairly quite simple. Probably about 100 lines of code or so. I have a nightly cron script that gets any new/modified marc records from the past 24 hours out of the catalog and then runs marc2xml on the dump file. Then I have a small script that breaks up the large marcxml files into individual xml files and imports them into SOLR! I then can use an XSL stylesheet such as the LOC's marc2oai to produce an OAI document or the marc2rdf, etc on the full marcxml files (since solr doesn't have the original record). I have yet to incorporate my OAI server code into this, but since it is already written, it would be a fairly easy merge. This is all built into my NextGen OPAC that I am working on and hope to open-source sometime this summer. So sorry, im not allowed to hand out the code just yet :( Thanks Andrew
Re: [CODE4LIB] marc2oai
On May 29, 2007, at 4:09 PM, Andrew Nagy wrote: Does anybody here know of a MARC2OAI program? I have a nightly cron script that gets any new/modified marc records from the past 24 hours out of the catalog and then runs marc2xml on the dump file. Then I have a small script that breaks up the large marcxml files into individual xml files and imports them into SOLR! I then can use an XSL stylesheet such as the LOC's marc2oai to produce an OAI document or the marc2rdf, etc on the full marcxml files (since solr doesn't have the original record). I have yet to incorporate my OAI server code into this, but since it is already written, it would be a fairly easy merge. Fun. I was thinking of doing something in Perl but no never mind. Any system that would the the following would work for me: 1. Export (to a file system) MARC data in communications format 2. Do magic. 3. Make MARC records accessible via OAI. Such an implementation could be operating system independent, standards compliant, and vendor-neutral. -- Eric Morgan