Hash table implementation of an object repository

Abstract

Abstract A method uses a plurality of hash tables to provide an object repository for object oriented application development and use. The method includes storing an object identifier and a representation of the object in a first hash table and storing data about the object and the object identifier in a plurality of paired hash tables with the hash tables organized in mirrored table pairs. The data about the object include an object's class name, object methods, return types, and the data values returned by object methods. The non-inverse hash tables of the mirrored table pairs support fuzzy searches for objects while the inverse hash tables of the mirrored table pairs support searches for objects by object identifier. A system that implements the inventive method includes a first hash table for storage of data representing the object with an object identifier used as a key for storage of the representing data and a plurality of mirrored table pairs for the storage of data about objects. In the non-inverse table of the mirrored table pairs, the data about the objects are used as keys to store an object identifier. In the inverse tables of the mirrored table pairs, the object identifier is used as the key for storage of the data about the data objects. The mirrored table pairs and hash table provide an efficient and economical object repository. Such an object repository may be provided with a computer application at a nominal expense. Thus, object oriented applications may be developed without regard for whether the end user system includes an object repository.

Inventors

• 👤 Rolf G. Hunt

Bibliographic Details