Get the output of w command in pipe delimited format

Shell Programming and Scripting
1

Since output of w command have variable number of columns I want to get the output in pipe delimited format. I tried

export OFS="|"; w

but that does not work.

Any ideas?

2

try

w | awk '{ gsub("[ ]+", "|"); print} '
3

It gives compilation error

/home/sdesai/rx/src> w | awk '{ gsub("[ ]+", "|"); print}'
awk: syntax error near line 1
awk: illegal statement near line 1

Even this is not sufficient if the field login@ contains "Mon 9am" it will be treated as two separate fields, while we want only one.

4

I tried it before I posted it no errors.

Assuming you want help:
Please post your current input from w as a sample, and then your expected output.

We also need to know the kind of UNIX because Solaris awk should be nawk instead, for an example of why we need this information.

5

Sorry, I did not mention UNIX flavor. I am running on SunOS. It works with nawk.

I would say my input would be something like

User     tty           login@  idle   JCPU   PCPU  what
user1    pts/2         7:08am     2      2      2  -ksh
user1    pts/9        Sat 4pm 2days                -ksh
user3    pts/20        7:01am    24                -ksh

I would expect

User|tty|login@|idle|JCPU|PCPU|what
user1|pts/2|7:08am|2|2|2|-ksh
user1|pts/9|Sat4pm|2days|||-ksh
user3|pts/20|7:01am|24|||-ksh