With numbers like that you are in a bit of a spot for trying to use them as ordered constraints. Still, though, if your flat file is in the correct order, you can simply start grabbing rows on the first match and then keep grabbing until you get a non-match.

To use a SQLish DB (even DBD::CSV) if you cannot abandon the catalog numbers and they are not unique to a row, you have a serious issue to solve in terms of assigning a unique ID number to each item at some point and translating the catalog number into the unique ID. If the catalog numbers are class-oriented, then you could assign sets of items a class ID and (with DBD::CSV) use a SQL statement like "select * from tablename where class = 'stamp_type_one' order by catalog_number;" to grab all the rows you're looking for sorted by catalog number.

The perldoc for DBD::CSV is pretty straightforward, and simple SQL is not so tough that you probably wouldn't be able to pick up most of what you need from just reading the module docs. If not, searching google or about.com for SQL should find you some decent tutorials pretty quickly.