[OPEN-ILS-DEV] Re: Importing Holds from Geac

[Garry Dunn]

Dan and others,

I've been assisting John/Innisfil on their evaluation of Evergreen and I 
must say congrats to the Evergreen team.  It's looking like a viable option.

My experience doesn't include PERL programming and I really don't know 
the internals of the Evergreen system that well, but I'm very well 
versed in SQL, C/PHP and the *NIX GREP utilities (awk/sed/grep/tr/...).

Using those tools, I've been toying with Innisfil's GEAC data and more 
or less successfully imported patron data (names/addresses/phone #'s/bar 
codes/etc...).  Their MARC records came in fairly easily using the 
method outlined with the Gutenberg records.

The harder part is the holdings data.  I've made some good progress, but 
there is still more to do.  We don't have a good definition of the GEAC 
data files (it's a Universe/Pick multi-value database).  There is 
basically no export utility that we know of.  I'm simply man-handling 
the raw data file and generating SQL statements which insert records 
into the appropriate tables in Evergreen.

It's coming together, but there is still more reverse-engineering to do 
with the GEAC data file.  If we're successful, I'll be sure to post our 
findings.

Relating to this, I have a question:  my scripts end up giving me 99094 
holdings records to go into the asset.call_number & asset.copy tables. 
If I import them all, they go quite happily but the OPAC web client 
never returns anything if I do a search.  If I import 20000 of the 99094 
records, searches work fine.  I'm using the Gentoo image for testing. 
Does anyone have any ideas why the larger sample set won't return search 
results?

-- 

Garry Dunn, P.Eng
www.trellisconsulting.ca