Thanks - Yes, probably should have been really specific
The file I'm working with will have hundreds of lines. The first few lines basically like this (incl headers)
The end of each line always has 6 commas, but before the 6 commas are the subjects each user is enrolled in. There may actually be twenty or more subjects.
It looks like the initial sed you suggested only works if there are two subjects per line. In my real file, the exact number of subjects per line is unpredictable.
User Name,Password,ID,First Name,Last Name,Subject,Other,#User Type,City,#Email,#Some Field,#Another Field
jsmi3279,xxxxxx,123456789,John,Smith,XMLT1005:MATH2987:PHYS4597:FRNC3578:ITLN590:LATN3490,,,,,,
glik2398,xxxxxx,123456789,Gill,Likely,XMLT1001:XMLT1004:MATH2987:PHYS4597,,,,,,
lbro4982,xxxxxx,123456789,Lucinda,Brown,XMLT1001:XMLT1002:ENGL3689:JAPN9257,,,,,,