CSV null columns with gawk

pcock · 03-30-2009, 05:35 PM

Hi,

Suppose I have a file.

Code:

$ cat dat
1,newport  ,          ,5.38
2,sydnenham,     58894,2.12

How can I tell gawk to print column $1,$2 but if column $3 is blank to ignore spaces and output just ,, (with no spaces) else print values $3 $4

I tried something like this but I don't think it's right

Code:

#!/bin/bash
gawk 'BEGIN { FS=OFS=","}
{if  ($3 == "          ") print $1,$2,FS,$4 ; else print $1,$2,$3,$4}'

Thanks

pcock · 03-30-2009, 07:35 PM

It was easier to remove leading and trailing space between commas

Code:

sed 's;^ *;;;s; *, *;,;g;s; *$;;' a.txt > b.txt

found this here!