in reply to Re^4: Iteration through large array using a N number of forks.
in thread Iteration through large array using a N number of forks.

I agree. The examples for LWP::Parallel are rather intimidating.

I think this is about as easy as parallel LWP gets:

#! perl -slw use strict; use threads; use threads::shared; use Thread::Queue; use LWP::Simple; use Time::HiRes qw[ time ]; $|=1; our $THREADS ||= 3; our $DELAY ||= 10; sub ping { my( $Q, $done ) = @_; my $tid = threads->self->tid; while( not $$done ) { my( $time, $url ) = split ':', $Q->dequeue; select undef, undef, undef, 0.01, while time < $time; my $start = time; printf "($tid) %20s returned [%50.50s] and took %f seconds\n" , $url , join( ' ', grep{ defined } head( "http://$url" ) ) , time() - $start ; $Q->enqueue( ( $DELAY + time() ) . ":$url" ); } } my $Q = new Thread::Queue; $Q->enqueue( map{ chomp; time() . ":$_" } <DATA> ); my $done : shared = 0; my @threads = map{ threads->new( \&ping, $Q, \$done ) } 1 .. $THREADS; <STDIN>; print "Stopping..."; $done = 1; $_->join for @threads; __DATA__ www.yahoo.com www.aol.com www.altavista.com www.time.com www.whitehouse.gov www.parliament.uk www.scottish.parliament.uk www.europarl.org.uk www.nasa.com www.perl.com www.perl.org www.activestate.com www.bbc.co.uk www.ibm.com www.google.com www.cnn.com www.perlmonks.com www.microsoft.com

Output (truncated for posting)

P:\test>435950 (1) www.yahoo.com returned [ text/html] and took 0.88 +7310 (2) www.aol.com returned [ OLserver/3.4.2] and took 0.79 +4034 (3) www.altavista.com returned [ 9-1 1109797758] and took 0.76 +6369 (3) www.parliament.uk returned [ rosoft-IIS/5.0] and took 0.45 +3125 (2) www.whitehouse.gov returned [ 9797455 Apache] and took 0.60 +9375 (1) www.time.com returned [ teEnterprise/4.1] and took 1.10 +9375 (2) www.europarl.org.uk returned [ rosoft-IIS/4.0] and took 0.46 +8750 (3) www.scottish.parliament.uk returned [text/html Oracle9iAS] and too +k 0.812500 (1) www.nasa.com returned [ rosoft-IIS/5.0] and took 0.54 +6875 (2) www.perl.com returned [text/h/1.3.33 (Unix)] and took 0.78 +1250 (1) www.activestate.com returned [ rosoft-IIS/5.0] and took 0.54 +6875 (2) www.bbc.co.uk returned [ /2.0.51 (Unix)] and took 0.39 +0625 (3) www.perl.org returned [text/h7760 Apache/1.] and took 1.09 +3750 (3) www.cnn.com returned [ 9797821 Apache] and took 0.56 +2500 (1) www.ibm.com returned [text/h26.3 Apache/1] and took 1.03 +1250 (2) www.google.com returned [ html 0 GWS/2.1] and took 1.35 +9375 (1) www.microsoft.com returned [text/h3 Microsoft-II] and took 0.62 +5000 (3) www.perlmonks.com returned [ Apache/1.3.27] and took 3.73 +4375 (3) www.altavista.com returned [ 9-1 1109797769] and took 0.59 +3750 (2) www.yahoo.com returned [ text/html] and took 0.81 +2500 (1) www.aol.com returned [ OLserver/3.4.2] and took 0.81 +2500 (3) www.parliament.uk returned [ rosoft-IIS/5.0] and took 0.40 +6250 (2) www.whitehouse.gov returned [ 9797455 Apache] and took 0.39 +0625 (3) www.europarl.org.uk returned [ rosoft-IIS/4.0] and took 0.39 +0625 (3) www.nasa.com returned [ rosoft-IIS/5.0] and took 0.59 +3750 (1) www.time.com returned [ teEnterprise/4.1] and took 1.21 +8750 (2) www.scottish.parliament.uk returned [text/html Oracle9iAS] and too +k 0.937500 (2) www.bbc.co.uk returned [ /2.0.51 (Unix)] and took 0.43 +7500 (3) www.perl.com returned [text/h/1.3.33 (Unix)] and took 0.60 +9375 (1) www.activestate.com returned [ rosoft-IIS/5.0] and took 0.60 +9375 (2) www.perl.org returned [text/h7772 Apache/1.] and took 1.54 +6875 (3) www.cnn.com returned [ 9797833 Apache] and took 1.73 +4375 (3) www.microsoft.com returned [text/h3 Microsoft-II] and took 0.45 +3125 (1) www.ibm.com returned [text/h26.3 Apache/1] and took 2.01 +5625 (2) www.google.com returned [ html 0 GWS/2.1] and took 1.37 +5000 Stopping... (3) www.perlmonks.com returned [ Apache/1.3.27] and took 1.85 +9375 (1) www.altavista.com returned [ 9-1 1109797779] and took 0.78 +1250 (2) www.yahoo.com returned [ text/html] and took 0.70 +3125

Examine what is said, not who speaks.
Silence betokens consent.
Love the truth but pardon error.