On Fri, 22 Jun 2001, Benedykt P. Barszcz wrote:

...

I've now got a pretty basic MARC converter built that will take all existing records in the biblio, biblioitems, items tables and convert them to MARC records. Can you tell how complete a MARC record will it be?

I'm no expert on MARC, but the plan is to make it a full MARC record. I don't have the code yet for actually writing out the data as a MARC record, but that really is trivial.

I noticed that you've added an "MARC import" link on the web page. Does it mean we would be able to import bibliographic data from an external file into koha's mysql tables? It would be wonderful.

I've been doing this for some time. My concern with this is that I've generally had to fiddle with the importer to get holdings records to work right, as everybody seems to put info in different places in the MARC record. Also, some things like Koha's "itemtype" which is pretty critical to the way Koha stores data, are hard to extract automatically out of any given MARC record. The way that I've been doing it is that the librarian's save the MARC records to their home directories on the server, and then go through and edit each imported record individually to make any necessary changes. When importing MARC records for an entire collection (when a library is switching from another program to Koha), I tweak the MARC import routines so that the data is imported correctly. This is why I'd like to see Koha adopt MARC as a base data storage format, to make importing MARC records from other sources a little less painful.

...

I stuck pretty much with the mapping I sent in my previous post, since nobody pointed out any glaring errors. I am not sure whether copyright data (f or g) fit into tag 245 according to USMARC. But I may be wrong on that.

You're right. Should have been 260 c (Date of publication, distribution, etc., may contain multiple dates (e.g., dates of publication and copyright)). Steve

On Sat, 23 Jun 2001, Benedykt P. Barszcz wrote:

software used does not make use of the indicators. Do you think that it will impede or make difficult to import data without indicators into MARC-enabled koha as you have designed it?

I can't see that being a problem. In fact, my current Koha -> MARC converter leaves all the indicators blank at the moment anyway. The only one I was thinking of automating was the 245 second indicator which tells how many non-filing characters are at the beginning of the title. I was going to search for "The", "A", or "An" to at least catch some of these. Rather english-specific solution, eh? :) Steve.