Hi, Can someone help to write a script for formating data as given below Input: Client: shp13 Backup ID: shp13_1104610099 Policy: hp_shp13_app Creator: root Sched Label: Weekly Schedule Type: UBAK (2) Retention Level: 2 weeks (1) Backup Time: Sun Jan 02 2005 04:08:19 (1104610099) Elapsed Time: 5 second(s) Expiration Time: Sun Jan 16 2005 04:08:19 (1105819699) Kilobytes: 30620 Number of Files: 11 Keyword: Arch_appvcp Client: shp13 Backup ID: shp13_1104609818 Policy: hp_shp13_app Creator: root Sched Label: Weekly Schedule Type: UBAK (2) Retention Level: 2 weeks (1) Backup Time: Sun Jan 02 2005 04:03:38 (1104609818) Elapsed Time: 10978 second(s) Expiration Time: Sun Jan 16 2005 04:03:38 (1105819418) Kilobytes: 79988740 Number of Files: 648105 Keyword: SHP13PVCS From client to keyword the data will repeat n number of times. Out Required: Client,Backup ID,Policy,Creator,SChed Label,Schedule Type,Retention Level,Backup Time,Esalpsed Time,Expiration Time,Kilobytes,Number of Files,Keyword shp13,shp13_1104610099,hp_shp13_app,root,Weekly,UBAK (2),2 weeks (1),Sun Jan 02 2005 04:08:19 (1104610099),5 second(s),Sun Jan 16 2005 04:08:19 (1105819699),30620,11,Arch_appvcp shp13,shp13_1104609818,hp_shp13_app,root,Weekly,UBAK (2),2 weeks (1),Sun Jan 02 2005 04:03:38 (1104609818),10978 second(s),Sun Jan 16 2005 04:03:38 (1105819418),79988740,648105,SHP13PVCS Perl or Shell/awk will do Thanks

What have you got so far?

This is a non-working start. It needs the ','. sed -e 's/:/-/' infile | awk -F '-' 'BEGIN { ORS = '\012' } /Client-/,/Keyword-/ { printf %s,, $2 }'

nawk -f venu.awk infile here's venu.awk BEGIN { FS="" OFS="," RS="" } FNR==1{ for(i=1; i <= NF; i++) printf("%s%s", substr($i,1,index($i,":")-1), (i==NF) ? "\n" : OFS); } { for(i=1; i <= NF; i++) printf("%s%s", substr($i,index($i,":")+1), (i==NF) ? "\n" : OFS); }

I've provided a Korn shell solution and another poster has provided an nawk solution. If you are doing this as part of your job, I'd suggest you learn how to do this yourself by buying a book. If it is a homework assignment, I'd suggest you read the book. #!/bin/ksh OLDIFS=$IFS NEWIFS=": " exec 3<./junk.data99 while read -u3 title data1 data2 do if [[ $title = "Client" ]] then IFS=$NEWIFS titleLine="" dataLine="" titleLine="$titleLine$title," dataLine="$dataLine$data1," elif [[ $title = "" ]] then : elif [[ $title = "Keyword" ]] then IFS=$NEWIFS titleLine="$titleLine$title" dataLine="$dataLine$data1" print $titleLine print $dataLine elif [[ $data1 = "Time" ]] then IFS=$OLDIFS titleLine="$titleLine$title" dataLine="$dataLine$data1$data2" else IFS=$NEWIFS titleLine="$titleLine$title," dataLine="$dataLine$data1," fi done Jerry Lemieux

intersting... So if this's a part of your work - just "BUY a book". And if this's a homework, then "READ the book". AM I getting it right? Sorry - just kiddin'.

I meant that if it was homework, read the text book. If it was for the job, buy a book (and by extension) read it. I guess I wasn't paying much attention to what I was writing. A long day.