This script builds a new marc_word

This script builds a new marc_word
table with a reduced number of tags (only those
tags that should be searched) allowing for
faster and more accurate searching when used
with the SearchMarc routines.  Make sure that
the MARCaddword routine in Biblio.pm will index
characters >= 1 char; otherwise, searches like
"O'brian, Patrick" will fail as the search
routines will seperate that query into "o",
"brian", and "patrick".  (If "o" is not in the
database the search will fail)