Question : Deleting Files with Bash Script

Hello,

One of my CMS applications makes backups of my INTRANET site at 2 times, 7am and 8pm on a daily basis. As you can see from the files below is that there are web backups and intranet backups on possibly the same day, and sometimes different days in the same /usr/local/bin/cms/backup directory. What I am trying to do is delete the oldest files of both sets to leave the 2 newest. I cant write a bash script to delete anything older then a certain day, but doing this might fall into a problem area where the Web backup does not have a newer copy like the Intranet does and would delete the pervious backup. Any suggestions on some code to make into a cron job that will delete all previous files except the 2 most recent of 2 different file types? This is an issue because every Intranet backup is 5GB and every Web backup is 2GB so if i have large amount of files in there at this size my server gets quite low on space.  Any advice would be greatly appreciated.

(As seen here I want to get rid of the oldest Web backup of 2/12, and the Intranet backup of 4/12, these dates vary as I have not figured out the reason the backups are so sporadic)
20080212_*_company_usa_web_page.zip
20080304_*_company_usa_web_page.zip
20080902_*_company_usa_web_page.zip

20080412_*_company_usa_intranet.zip
20080904_*_company_usa_intranet.zip
20080916_*_company_usa_intranet.zip


Thanks in Advance.

Answer : Deleting Files with Bash Script

NOTE - This answer involves removing files.  This is dangerous.  Please test, test and test some more.  These lines work for me but you should test and check them with your technical support before using them on a real production system.  If you loose files.... its not my fault.

This is useful (but not really what you are after) :
find /usr/local/bin/cms/backup* -ctime 30 -print | xargs /bin/rm -f
This deletes files that have not been touched for more than 30 days (change the number if needed)

This is probably closer :
ls -t1 /usr/local/bin/cms/backup* | tail +5 | while read file ; do echo removing  - $file; done
This lists all the files and then tails it knocking off the first 4 (it does one less than the number specified). This is then passed into a loop which tells you each file name it will remove.

Test this and if you are happy with results then add the remove bit :
ls -t1 /usr/local/bin/cms/backup* | tail +5 | while read file ; do echo removing  - $file; rm $file; done


I hope that helps