Indexing Method for Fast Random Access to Collated Data
Original Publication Date: 1986-Sep-01
Included in the Prior Art Database: 2005-Mar-09
A mechanism for rapidly retrieving collated entries, such as phone directory listings, from a large data file is described. Only a very small portion of the data file is brought into memory, thus substantially improving performance over other file organizations. On-line access to large data files of directory-type data needs to be very fast for productive use. Since it is impractical to keep these large files in memory for immediate access, an efficient mechanism is needed to minimize the amount of data retrieved from the file with each directory look-up. The method involves pre-processing the data, collating it and developing a table of special entry points to random records in the file. Once the computation has been performed to fill the table, the table is stored as part of the file.