You can look it up.
Consider a flight database in London that lists flights to a thousand U.S. cities. One alternative would be to keep a record specifying the path to each of the thousand cities. Suppose, however, that most flights to America hub though Boston, except flights to California, which hub through Los Angeles. This observation can be exploited to reduce the flight database from a thousand entries to two prefix entries (USA* --> Boston; USA.CA.* --> LA).
A problem with this reduction is that a destination city like USA.CA.Fresno will now match both the USA* and USA.CA.* prefixes; the database must return the longest match (USA.CA.*). Thus prefixes have been used to compress a large database, ...