UNIX Command Line

You awk script does the right thing:

$ echo "ASX,23,PILE2.3,MNP,54,NEW123,LELO,67,PNP2" | awk -F, '{for(i=1;i<=NF;i++){printf("%s%s",$i,i%3?",":"\n")}}'
ASX,23,PILE2.3
MNP,54,NEW123
LELO,67,PNP2

But I don't think your sed script is quite right:

$ echo "ASX,23,PILE2.3,MNP,54,NEW123,LELO,67,PNP2" | sed 's/\([^,]*,[^,]*,[^,]*,\)/&\
/g'
ASX,23,PILE2.3,
MNP,54,NEW123,
LELO,67,PNP2

with semicolons at the end of the lines. To get around this, try using

$ echo "ASX,23,PILE2.3,MNP,54,NEW123,LELO,67,PNP2" | sed 's/\([^,]*,[^,]*,[^,]*\),/\1\
/g'
ASX,23,PILE2.3
MNP,54,NEW123
LELO,67,PNP2

hey, you are absolutely right. I missed that. Thanks for your post and your tip.

Thanks.

Hi Guys, i am very new to support analyst role. i have a big error file (75 mb) and which consists only one line!! yes only one big big line. i now have to find particular error details occured 200 times in line. since customer ID changes in every occurence only consistent line is "Payment already allocated". i now have to print 150 characters before and 100 characters after this statement. Again there is only one line in file
please help

@Ritesh,thanks for commenting.

This example might help you:

$ echo "abcdefghij0123456789" | awk -v numchr=3 'match($0,"f"){print substr($0,RSTART-numchr,numchr),substr($0,RSTART+1,numchr)}'

Output:
cde ghi

Feel free to put a comment if you require some more assistance.

sed 's/,/\n/3;P;D'