(I bet references would be even better.)

The problem is $3 is interpolating, leaving awk with merely {print}.

You mean:

    open (MAXNUM, "awk -F: '{print \$3}' testfile |")
        || die "cannot fork awk: $!";

or

    open (MAXNUM, q{awk -F: '{print $3}' testfile |})
        || die "cannot fork awk: $!";

There's (nearly) never any need to cat just one file at something, btw.

For the record, doing it all in perl would be:

    open (MAXNUM, "testfile")         || die "cannot open tesfile: $!";
    while (<MAXNUM>) {
        print +(split(/:/))[3], "\n";
    }

That silly plus is to inhibit perl from thinking I'd said:

        (print split(/:/)) [3], "\n";

which wuldn't have worked well at all.

--tom
--

    There is, however, a strange, musty smell in the air that reminds me of
    something...hmm...yes...I've got it...there's a VMS nearby, or I'm a Blit.
        --Larry Wall in Configure from the perl distribution

Hi,

I need to be able to pipe the output of a command to open().  I understand
how to do this, but is it possible if the command itself contains multiple
pipes?

Example:

open (MAXNUM, "cat testfile | awk -F: '{print $3}' |");

I want take each line in testfile, pull the third token (delimited by :),
and have Perl process it.  When I do this though, Perl stops
execution on the command after the first pipe.  So all it does is
'cat testfile', but it doesn't parse out the third token with awk.
I know I can parse it out inside Perl, but out of curiousity, I'd
like to know if there's a way to do it like this.

Is there a way around this?  I don't want to have to redirect this
output to any intermediate file.

Environment:  Perl 5.000 on Solaris 2.3

thanks for any pointers,

tim ma
DIGIT Lab, university of utah