I just committed some code that Stephen Hedges and I wrote a while
back for the Nelsonville System along with a new systempref to
manage it. The code allows a library to switch 'on' the ability
to sort using the semantic information in the MARC non-filing chars.
The code is really only relevant to the 2.2 branch and it seemed a
shame to waste it. When we move to Zebra we'll use it, rather
than SQL to do the sorting.
One thing to keep in mind is that it's very proc intensive for a
large collection (over 100,000 items). So much so that NPL only
used it for a few weeks it was so slow. For the smaller systems
though it is a nice feature and won't affect search speed.
Also, it only sorts biblio.title, though it would be trivial to
add support for other fields -- drop me a note if you have a
In case you're wondering what non-filing characters are, some
versions of MARC (like MARC21) specify the number of characters
to ignore when sorting. So if your titles was 'The Jury', there
would be 4 non-filing characters; if it was 'A tall tale' there
would be 2 and so on.
Finally, I should mention that this probably only works for
MARC21, it's untested for UNIMARC but if you leave the syspref to
its default of 'off' the code won't execute so it won't cause
Let me know if anyone has questions about this ...
Joshua Ferraro VENDOR SERVICES FOR OPEN-SOURCE SOFTWARE
President, Technology migration, training, maintenance, support
LibLime Featuring Koha Open-Source ILS
[hidden email] |Full Demos at http://liblime.com/koha |1(888)KohaILS