Browse Prior Art Database

Text Compression for Directory Files

IP.com Disclosure Number: IPCOM000062059D
Original Publication Date: 1986-Oct-01
Included in the Prior Art Database: 2005-Mar-09

Publishing Venue

IBM

Related People

Authors:
Evans, CW [+details]

Abstract

A compression technique is described which provides 50 percent reduction in file size, yet permits random access of files. Compression is optimized for use with directory-type files. This compression mechanism permits up to 18,000 entries to fit on a 360 KB diskette. The process capitalizes on the unique distribution of characters in a typical directory and on the tendency of entries to be similar to the previous entry. High performance is very important in decoding online access to directory data, so only 4-bit tokens are used to encode the directory data. The 4-bit tokens that make up a compressed file are used to look up characters from tables, or to select a new table. The state of the decoder determines which table will be used.