salutations, we really wanted to stick with the BerkeleyDB solution, so, we found a solution to the problem: we generate the dict file in TXT format, and, then, we convert it to Berkeley DB by using the command "db-load" at command-line. for this, besides the BerkeleyDB module of Perl, we also installed the Windows installer of BerkeleyDB at http://www.oracle.com/technology/software/products/berkeley-db/index.html. then, we discovered that the command (in the command-line)

db_load -c duplicates=1 -T -t hash -f dict.txt dict.db
[download]

converts "dict.txt" (with keys and values separated by a newline character, and each pair of lines being a record) to the BerkeleyDB database "dict.db", allowing duplicate keys. this solution turned out to work great. thank you for all the help.