Shell Programming and Scripting

I'm doing a cleanup script for a directory using KSH. I'm keeping the file name prefixes in a text file. In a KSH, I want to read the prefix from the file, and match the pattern of the file and keep and compress(.Z) the latest 4 versions of the matched files in the directory. And I want to delete the remaining versions. Reason I'm doing like this as the files are cummulative , I want to keep only the latest files only.

Eg: My text file has prefixes like this filenames.txt abcd xyz pqr .. My directory has many files with different names My directory has files like this

Code:

abcd###.YYYYMMDDXXXXXX.txt
abcd###.YYYYMMDDXXXXXX.txt
abcd###.YYYYMMDDXXXXXX.txt  (like this with different time stamps)
xyz###.YYYYMMDDXXXXXX.dat
xyz###.YYYYMMDDXXXXXX.dat
xyz###.YYYYMMDDXXXXXX.dat

Code:

str12.YYYYMMDDXXXXXX.txt
qwefd.YYYYMMDDXXXXXX.dat

In this directory, as I 've prefixes abcd, xyz in the text file, so I want to compress the latest 4 verisons of the files with that prefix and delete the remaining versions with that prefix. I dont want to touch the files not mentioned in the text file. And I want to run the script next time I want similarly keep the latest file versions and remove the remaining versions (As next next time there are previously compressed files will be there.) At any time I need only 4 latest file verisons only. I can remove the previously compressed files.

I need to script like this KSH. Any help is appriciable

Pls. show the input file. What have you tried so far?