Shell Programming and Scripting

HI all,

i have the source file like below,

05 BL-FEE-CYC-CDE PIC S9(03) COMP-3.
05 BL-FEE-ERR-MSG PIC X(00030).
05 BL-FEE-TYPE PIC X(00001).
418181*# 05 BL-LCYC-REL-CTD-AVG PIC S9(15) COMP-3.
418181 05 BL-LCYC-REL-CTD-AVG PIC S9(15)V9(02) COMP-3.
05 BL-LCYC-REL-CTD-AVG PIC S9(15)V9(02) COMP-3. ##define

i need to cut the previous column of "PIC" column and 2 columns after PIC.

in a row PIC may come in any column in the above 1st 3 rows its coming in 3rd column
in 4th and 5th rows it comes in 4th column

lost row it comes in 3rnd column.

i tried with AWK.

can some one help me to find the column id of PIC

Thanks
Baski

give the desired output of your example

This works, make sure you have two columns after pic to avoid bad results,

Code:

awk '{for(i=1;i<=NF;i++){if($i=="PIC"){print $(i-1) $(i+1) $(i+2)}}} ' test.txt

Code:

awk '{for(i=1;i<=NF;i++){if($i=="PIC")print $(i-1) $(i+1) $(i+2)} }' inputfile

Thanks a lot Kurmar..

I hope it will work..let me check..

Baski

Please always use code tags icon available in the menu bar while posting 'codes,sample data,sample output etc' so that it helps improving the readability

Code:

sed 's/.* \([^ ]*\) PIC \([^ ]* *[^ ]*\) *.*/\1 \2/' inputfile > outfile

Try this,

Code:

nawk '{for(i=1;i<=NF;i++) { j=i+2; k=i-2; if($i=="PIC") print $k" "$j } }' inputfile