How to pass a bash script variable to AWK using $() and collect the ouput

Question

The problem is in following:

solaris:~/src/brk$ cat .file
one xxx 123
two yyy 123

three bbb 321

four xyz 123

Script:

solaris:~/src/brk$ cat scr.sh
... 
PATH="dist/"
LINE=$(awk '/^[ \t]*$/{next} /^[ \t]*#/{next} {printf $2" "}' .file)
echo "LINE=$LINE"
...

Output:

xxx yyy bbb xyz

I want the following output:

dist/xxx dist/yyy dist/bbb dist/xyz

How can I modify awk expression to get the desired output?

Answer 1

You can set variables in awk using the -v switch:

PATHI="dist/"
LINE=$(awk -v p=$PATHI '/^[ \t]*$/{next} /^[ \t]*#/{next} {printf p$2" "}' .file)

Answer 2

Do not use PATH as a variable name! That is already used by the shell. Always use lower-case variables in bash for this reason: the ones used by the shell are always uppercase. Than, you can pass a variable to awk :

dir="dist/"
line=$(awk -v dir=$dir '/^[ \t]*$/{next} /^[ \t]*#/{next} {printf dir$2" "}' .file)