Data compression and expansion services

z/OS supports compression using two different algorithms. The first algorithm exploits the occurrence of repeated characters in a data stream. Encoded data contains a combination of symbols that represent strings of repeated characters and the original characters that are not repeated. The CSRCESRV service uses this algorithm to compress data.

Data that contains many repeat characters can exploit these services most effectively. Examples include:

Data sets with fixed field lengths that might contain many blanks
80-byte source code images of assembler language programs.

Using these services with other types of data might not result in appreciable data volume reduction. In some cases, data volume might even be increased.

The second algorithm encodes data by replacing strings of characters with shorter fixed-length symbols. A key component of this technique is the symbol table, usually referred to as a dictionary. Encoded data contains symbols that correspond to entries in the dictionary. The CSRCMPSC service uses this algorithm to compress data.

This service requires more storage than the CSRCESRV services, but is a good choice if you know what your data looks like and storage is not a concern.