in reply to How to set pipe first and then use the forkmanager?

I think I must establish 4 pipes first(right?),

Why? Why not fork first and have each fork establish its own pipe?

Untested example:

for my $file ( @data_files ) {
    $pm->fork and next;
    open my $h, "| ...." or die $!;  
    print $h ...;
    ...
    $pm->finish;
}
$pm->wait;

[download]
With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
"Science is about questioning the status quo. Questioning authority". I knew I was on the right track :)
In the absence of evidence, opinion is indistinguishable from prejudice.

Replies are listed 'Best First'.
Re^2: How to set pipe first and then use the forkmanager?
by mlin (Novice) on Sep 22, 2016 at 10:51 UTC
    Because establish the pipe to the program may be a little bit time-consuming. So I would like to move the open pipe outside the foreach() in the serial type. Similarly I consider may be I should establish pipe first and then use fork.
[reply]
      Presumably, you're forking because your input data files are very large. The time taken to open a pipe is then negligible compared to the time needed for processing your data.

      Don't forget that premature optimization is the source of almost all evil.

[reply]
        Thanks! I'll use this form at the moment, maybe I solve the problem when I know more about the fork. I have a new question: ----- complement: while the program A runs, it usually prints some information or warnings to the screen? Will it be a trouble for the forked processes? or I must throw the output of A, like: 
        foreach (@data_files) {
   open $h, "| program_A 2>& /dev/null" or die ...;
   print $h ...;
   print $h ...;
    ...
}
close($h);

        [download]
        Can I hold those information? Thanks!
[reply]
[d/l]

In Section Seekers of Perl Wisdom