Hi there. I'm looking to do a basic shell script whereby there are four fields in a record: NAME POSTALCODE PAID OWED There are about 15 records in a file that I'm reading in. I've been cutting them as follows: #!/bin/sh #Rent FILE=`cat <rent_payments` while [ $EXITFLAG -eq 0 ] do LINE=`echo $FILE | cut -f1 -d" "` #This is supposed to stop the field at the end of each line. There's a fatal flaw in this whereby some people's names have spaces in. if [ -z $LINE ] then EXITFLAG=1 else NAME=`echo $LINE | cut -d -f1` POSTCODE=`echo $LINE | cut -d -f2` PAID==`echo $LINE | cut -d -f3` OWED==`echo $LINE | cut -d -f4` echo -n "Postcode: $POSTCODE Name: $NAME PAID: $PAID" >>paid echo "" >>paid echo -n "Postcode: $POSTCODE Name: $NAME Owed: $OWED" >>owed echo "" >>owed fi done The problem is this... when the program runs, it loops forever. It does not stop. And when I ctrl-c it, the 'paid' file contains only the details from the first record and it doesn't go onto the next line. Is there any way I can do anything about this? I've looked all over and I can't find anything! Any help would be appreciated. Thanks.

Perhaps I should have pointed out that the records look like this: Davids,RE454JH,120,0 This is name,postcode,paid,owed. With ',' being the field delimiter.

Rather than try and fix a very broken script ;-), it would be better to use what is the right tool for this job in the first place: awk. awk -F',' {print "Postcode: " $2 " Name: " $1 " PAID: " $3 > "paid"} {print ... > "owed"}' rent_payments Where "..." means fill it in yourself. There is no need to write scripts for extremely simple things like this that can be done with a single command. Having said that, if the object of the exercise is to learn bash rather than solve the problem, please say so.

s/-F',' {/-F',' '{/

If I had a very large file to parse I would probably use perl: #!/usr/bin/perl open (FILE, "rent_payments")||die "Can't find file!"; @records=<FILE>; foreach $line (@records){ @rent=(split(',', $line)); print "Postcode:$rent[1]...Name:$rent[0]...Paid:$rent[2]\nPostcode:$rent[1]...Name:$rent[0]...Owed:$rent[3]\n"; }

Thankyou. Perl is not an option though. The awk command is something I had considered using but I was informed by a 'knowledgable' friend that it would be messy to do as I wanted to read records in (about 20) and then split them up into owed and paid and for the thing to tell me in descending order who owes me what. That's easy but actually cutting it is proving hard to do. When the script runs, it always returns the first record from the file and doesn't go past the first line. I'd rather it be a shell script and not something run from the command line as basically, all I want to do is this: In the middle of the month, I was to see what was paid by tenants by running ./rent. I want the script to delete any previous instances of the owed and paid files and add something to the top of each of them (echo "Paid" and echo "Owed" will suffice). Then using the method shown in the first post, produce an ordered (in order of most owed to least owed in owed file) list of payments in two seperate files. I've done something like this before when I used to go to college and I know it could have been done in cut but my college days are behind me now. I've looked at my older books and they're no help. There is no object of excerise but more of proof of concept on my part.

So put the awk command in the script! It is hardly messy and you will not get your script to work at all the way it was because it is trying to process the whole file as a single variable, losing the line structure. You could do it with cut but it is not the right way to do it - if you use script constructs properly for reading the file like read and IFS redefinition then cut is again redundant.