Back in the days of VB for DOS I built a DB where I hashed the first four characters in the key to a number proportional to their alphabetic position. Ignoring spaces, numbers and punctuation, AAAA hashed to 1 and ZZZZ hashed to 32,000. The hash number represented a row in an integer array.

If the row was empty I filled it with the record number where the actual data was stored. If occupied I replaced the original value with a negative number whose absolute value pointed to a record holding the head of a linked list whose entries were the record numbers holding the data of collided keys where the linked list was maintained in alphabetical order of the keys so that a long string of collisions could be searched up and down the linked list with a binary search instead of a sequential search.

Given today’s memory sizes I now could hash a much longer key string to a much bigger number in a much bigger array, e.g. 32M, and have far fewer collisions. Still wasteful of storage, of course, but back on an old DOS 80386 machine my primitive design worked quite well.