mjhjmtu

I have this file

AK3*BPS*2
AK4*8*0510*1
AK3*RMT*12
AK4*1*0128*7*CR

and I want this output

BPS 2 1
RMT 12 7 CR

I am using this command

awk -F* '$1=="AK3" print $2, $3 "c" ($1=="AK4" print $4, $5 ' $FileHome/badfile_$session_num.txt >> $FileHome/input_output_record.txt

but i am getting this result

BPS 2
1
RMT 12
7 CR

Any hint will be appreciated

You seem to want:

1. The last two fields of each AK3 line, followed by


2. Fields four and five from the AK4 line (field five may be missing).

$ awk -F '*' '/^AK3/ printf("%st%s", $2, $3) /^AK4/ printf("t%st%sn", $4, $5) ' data.in
BPS 2 1
RMT 12 7 CR

Another way is to directly control the output record separator(ORS)

$ awk -F'*' '$1=="AK3"ORS=OFS; print $2, $3 $1=="AK4"ORS=RS; print $4, $5' ip.txt
BPS 2 1 
RMT 12 7 CR

• by default, output field separator(OFS) is single space


• and input record separator(RS) is newline

This is easy to modify for different OFS and also shows empty field

$ awk -F'*' -v OFS=',' '$1=="AK3"ORS=OFS; print $2, $3 $1=="AK4"ORS=RS; print $4, $5' ip.txt
BPS,2,1,
RMT,12,7,CR

Plain print adds a newline (the output record separator), but printf doesn't, so we can use that on the AK3 case with an explicit separator between the fields. In the AK4 case, we want a field separator before the first field, and the simplest way to get that is to print an empty field first.

$ awk -F* '$1=="AK3" printf "%s", $2 OFS $3 $1=="AK4" print "", $4, $5 ' < input 
BPS 2 1 
RMT 12 7 CR

This is easily converted to giving tab-spaced output, just add -vOFS='t' in the beginning of the command line.

Though if you want fixed-width columns, it's probably easiest to use printf and width settings on the %s format:

$ awk -F* '$1=="AK3" printf "%-3s %-2s", $2, $3 $1=="AK4" printf " %-1s %sn", $4, $5 ' < input
BPS 2 1 
RMT 12 7 CR

Over-long fields will still extend, unless you give the maximum field width too, with e.g. %-3.3s, and this will produce broken output if the input lines are not in correct order.

How about:

awk -F '*' '
 printf "%s %s", $2, $3
 getline
 for (i=4; i<=NF; i++) printf " %s", $i
 print ""
' file | column -t

BPS 2 1
RMT 12 7 CR

With plain bash:

while IFS='*' read -r a1 b1 c1; IFS='*' read -r a2 b2 c3 rest
do
 echo "$b1 $c1 $rest//*/ "
done < file

I achieved the result by below commands

sed -n '1~2p' testfile.txt| awk -F "*" 'print $2,$3' >> op.txt
sed -n '2~2p' testfile.txt | awk -F "*" 'print $4,$5' >> op.txt

sed -n '1~2p' op.txt |sed "N;s/n/ /g" |awk 'print $0"n"'; sed -n '2~2p' op.txt| sed "N;s/n/ /g"

Output

BPS 2 1

RMT 12 7 CR