Hi Jared In cases where libraries use commercial cataloguing services, providing downloadable MARC would usually contravene the licence agreement that the library has with the source/consortium. it would add complexity to a library OPAC to distinguish between locally created MARC records which may be available for download, and 3rd party which might not. Cheers, Pru On 19/12/2012, at 7:34 AM, "X Middleton" <x@vaults.ca> wrote:
Thanks, Jared, for confirming my suspicions. It still seems weird to me how many library interfaces "display" MARC and don't provide a machine-usable format. It's better than nothing though, I suppose. At least Koha isn't one of these systems -- Koha provides so many options for downloading records it's an embarrassment of riches!
I also received a reply from someone (who doesn't seem to have CC'd the list?) who reported they copied and pasted the example link I provided into MarcEdit, and after a bit of manual re-formatting to remove "subfield markers" (I presume the | characters) and possibly adjust some spacing they were able compile the text to an importable MARC record. (They kindly attached the result for me.) To duplicate this prodigious feat, I guess I'd need to do some more research on the MARC format acceptable to MarcEdit... it would be handy to know how to do.
On Tue, Dec 18, 2012 at 12:58 PM, Jared Camins-Esakov < jcamins@cpbibliography.com> wrote:
X,
I do a fair bit of cataloging for a small library that has a lot of
old/rare books, though I am an amateur librarian. I often use worldcat to find a library that has a copy of our uncatalogued book (when I can't find any z39.50 accessible sources). These libraries often have an option to display a MARC record. Is there some way to get such a display into Koha besides manually copy/pasting bits and pieces into a new MARC record?
Not easily if they do not provide a Z39.50 server or raw MARC record download. What you are seeing when looking at the "MARC view" is a representation of the MARC format that has been designed to be easily readable for humans. MARC records are actually (usually, anyway) stored in the binary format that LC provides their records in.
For example: http://tripod.brynmawr.edu/find/Record/.b3780640/Details#tabnav
Copying and pasting the displayed record into a text file and trying to import it via "stage for import" does not work. I notice that Library of Congress has an option to download MARC records for its items, which can be 'staged for import" without a problem. Looking a the contents of the LoC downloaded MARC file, it looks almost nothing like the ones displayed by libraries which don't have a download option.
I've searched the web (i think) exhaustively for clues as to how to actually handle or convert these "display" marc records into something usable (something that looks like the LoC downloaded MARC records), but with not success. I've installed and attempted to use MarcEdit and IsisMarc hoping they might have some functionality to convert/transform this data... but no luck.
It just seems crazy that there's all this info out there, and I can't find any way to use it beyond the crudest (but better than nothing!) methods of cherry picking.
What am I missing?
Most libraries that use copy from libraries other than their national library use OCLC or Skyriver (or a similar union database) for obtaining records. It would be possible to write screen scrapers to harvest MARC data from OPACs, but since there is no standard formatting, you would need at the very least a screen scraper per ILS, if not a screen scraper per catalog to cope with minor customizations made by various libraries.
Regards, Jared
-- Jared Camins-Esakov Bibliographer, C & P Bibliography Services, LLC (phone) +1 (917) 727-3445 (e-mail) jcamins@cpbibliography.com (web) http://www.cpbibliography.com/
Koha mailing list http://koha-community.org Koha@lists.katipo.co.nz http://lists.katipo.co.nz/mailman/listinfo/koha