Hi!

In an awk tutorial i was reading (from www.grymoire.com), it was mentioned that i can set the input field separator to more than one character using FS. However when i tried it out, it wasnt working.

File a :

abc/:/:as
as/:f/:f
AE/:/:as/:/:s


script a.sh

#!/bin/awk -f
BEGIN {
FS="/:";
}

{
print $1, $2;
}



command:

home/# as.sh<a
abc :
as :f
AE :


however, the output i shud get is
abc
as f
AE

can someone tell me what am i doing wrong here?

Theat awk command worked fine for me. Try replacing awk with nawk.

Thanks! it worked. But why didnt awk work? coz the tutorial was written for Awk, not nawk. The tutorial was for Solaris n my flavor is also Solaris. It was written in 1989 tho - did that make any difference?

Solaris has a crippled version of awk - which is the default. It also has nawk which is more like new awk and supports features.

There are POSIX standards for awk which came out long after the awk version Solaris has...

see:
http://www.opengroup.org/onlinepubs/...ities/awk.html

POSIX is a set of standards developers of most flavors of UNIX try to follow.
This allows, for example, somebody here running Linux to get an example from someone else on this forum - that other person runs Solaris - and there will be a good chance it works on both boxes. Solaris doesn't play POSIX with awk, but it's there to support old code. Just use nawk and forget awk. IMO.

thanks! after a cloudy day, its a sunny one again!