UNIX for Advanced & Expert Users

Select the code and click on the "#" symbol above the edit form.

The syntax looks correct. By the error message you are getting and the fact that you said this was in a script (not just on the command line) I suspect that there is a spurious, non-printable character in your script that is hosing up the scanning of the line properly and causing it to think that

Try this:
a. you can often find these non-printable chars by using "od -c filename | less"
b. re-type in the command on a new line and delete the old line altogether

Here is an example of how I reproduced your error. Look closely at the output of 'od -c' where there is actually a bad char (002 == CNTL-B) that your editor may not show you, but truly throws an error for nawk:

Code:

$ cat badnawk
#!/bin/sh
nawk -F":" -v red=2 '{print $red}' $0
$
$ sh badnawk
nawk: can't open file {print $red}
 source line number 1
$
$
$ od -c badnawk
0000000   #   !   /   b   i   n   /   s   h  \n   n   a   w   k       -
0000020   F   "   :   "       -   v 002       r   e   d   =   2       '
0000040   {   p   r   i   n   t       $   r   e   d   }   '       $   0
0000060  \n
0000061

I checked. There are no hidden characters. For double measure, I have re-typed that line of code (not copied and pasted). Still the same error.

Finally resolved.

nawk -F"," -v red="$random_variable" '{print $red}' $w_dir/$file_name > $w_dir/${column_name}

The double-quotes around the variable did it.