As you might imagine, I didn't post anything yesterday, I was too busy celebrating. I'm now just about to import the old library database into OpenBiblio, having just done two successful test imports. Quite a lot of work went into making this happen. Yesterday I went through each field working out what had to be deleted or replaced in order to get the file conversions to work. Here's a list of what needed to be done:
Book ID number - I decided to omit this, as each book will be assigned a new number along with a barcode, once in OpenBiblio.
ISBN - I had to ask our techie in residence, Dave, to write a little script to replace all the zeros that were lost when the database was imported to IUMA. The Dog was busy working on something for Dave, so it seemed like a fair swap.
Author - I had to create a second field from this, called "statement of responsibility" - it's a librarian thing. Then, I had to delete all the multiple authors, and all the "et al" and (ed)s.
Title - All the commas had to be replaced with pipes, then the pipes were replaced again by commas in the MARC editor. In addition, all the subtitles had to be identified seperately from the main titles, another job for the MARC editor.
Statement of responsibility - remember I created this from the author field? I had to replace the commas with pipes, then replace the pipes with commas again in the MARC editor.
Edition - this just needed a bit of tidying up, replacing the numbers with the text "nth ed".
Publisher - those pesky commas needed replacing again.
Year of publication - no work needed.
Price - not a traditional field recorded in library catalogs (sorry my British chums, I've moved over to the US spelling). I had the dilemma of whether to try to include this or not, but since it would have meant more script or code writing, I decided it wasn't worth the effort. Let the accounts software sort it out.
Subject - here I had to replace the commas that seperated each subject phrase, with a right curly bracket. Once in the MARC editor, the curly brackets were replaced with $a, to indicate further subject terms.
The first test import wouldn't allow me to import all the records at once, so I had to divide the file into two, then it seemed fine. Of course, when the data is finally imported, I'll then have to get some barcode software and stick barcodes to all the books, not to mention working out how to deal with registering library users.