How to best match strings

I have a table where the primary key field can have values that range in length between 2 and 7 characters.A client application has an input of a string that has a length of exactly 7 characters. The application need to identify the row whose primary field contains the substring that the input string starts with, e.g, Input is 'hello', then the following rows apply, 'he', 'hello'. The approach I'm using now is to select from the table using the whole string. If no data was found, then I select using a substring of the input string without the last character, and another select without the laste two characters, and so on until I find the rows I need.
I find this approach slow and resource consuming. Can someon suggest a more efficient approach.
Thanx Received on Thu Aug 22 2002 - 18:00:01 CEST