Database storage and retrieval method using a declining stage size and repetitive searches

摘要

A hashing database system that efficiently handles collisions comprises a list of data blocks divided into a plurality of stages, each of which is geometrically smaller than the preceding stage. The key of the data to be stored or retrieved (the "desired key") is hashed and, along with the stage number, used as an input into a staging algorithm to derive an offset into the list. One or more data blocks in the list surrounding the hashed-to data block are checked to determine whether the key field in the data block matches the desired key. If the keys match, then the desired data block is found. If the desired key is not found, then the hashed key is again used as an input into a staging algorithm along with the stage, and a new offset is derived and used as an index into the next stage of the index list. These steps are repeated until the key is found or the last stage is reached. The last stage is an overflow where the data is guaranteed to be found or space is guaranteed to be found for insertion of new data if not found in one of the stages.