awk replacing text but not for the last line

Question

i tried the awk liner below(on windows command-prompt):not working properly

gawk -v var="hot" "{ if(!NR){gsub(/cool/,var,$0) ;print} else{print}}" awk_test

input file is below

 this is a cool jack
 this nota cool kack
 this obviously a cool jack

a unix solution is also feasible

The condition `(!NR)` is always false because NR (the current record number) is always a positive integer. — glenn jackman, Jul 22 '10 at 16:56

tonio · Answer 1 · 2010-07-22T06:40:14.350

0

You could try something like:

sed '$d' yourfile | sed 's/cool/hot/g' > newfile
tail -1 yourfile >> newfile

this should do the substitution on everything but the last line of the file first, and then add the last line of the original file.

edited Jul 22 '10 at 06:40

answered Jul 22 '10 at 06:33

tonio

this is ok ...but i dont have sed in my command prompt.i only have gawk.so i need an awk/gawk solution. what is actually wrong with my gawk script. – Vijay Jul 22 '10 at 06:44

score 0 · Accepted Answer · answered Jul 22 '10 at 08:47

0

awk -vvar="hot" '{gsub("cool",hot,t);print t}{t=$0}END{print}' file

answered Jul 22 '10 at 08:47

ghostdog74

score 0 · Answer 3 · answered Jul 22 '10 at 16:54

0

You could also pass the number of lines as a parameter:

gawk -v var=hot -v lines=$(wc -l < test.txt) '
    NR != lines {gsub(/cool/, var)} 
    {print}
' test.txt

answered Jul 22 '10 at 16:54

glenn jackman

3 Answers3