http://www.perlmonks.org?node_id=964622


in reply to Re: trying to get timeout to work (easier with threads)
in thread trying to get timeout to work

That sounds promising. I will have to check threads and threads::shared as I in fact need (or wanted) to:

  1. Run several system commands in parallel
  2. Store the results in an sqlite DB
  3. Time out if a command runs too long

As it seemed to be not too easy to get the data from the parallel commands back to the parent, so that I just have one writer to the sqlite DB, I skipped this part for now and concentrated on the timeout part.

It seems with threads::shared, I can achieve what I wanted in the first place. Thanks BrowserUk!


s$$([},&%#}/&/]+}%&{})*;#$&&s&&$^X.($'^"%]=\&(|?*{%
+.+=%;.#_}\&"^"-+%*).}%:##%}={~=~:.")&e&&s""`$''`"e

Replies are listed 'Best First'.
Re^3: trying to get timeout to work (easier with threads)
by BrowserUk (Patriarch) on Apr 11, 2012 at 21:20 UTC

    Here's an improved version.

    • The old version always waited for at least the timeout period, even if the command completed early.

      This one returns as soon as the external command terminates.

    • The old version returned nothing if the process timed out.

      This version will return as much as was available before the process was killed.

    • It also addresses a potential "Uninitialised" warning if the system is heavily loaded and the thread takes a while to start.

      If the thread or process was slow starting due to system load or other factors, the old version could try to use $pid before it was set. This corrects that error.

    #! perl -slw use strict; use threads; use threads::shared; $|++; our $N ||= 11; my $TIMEOUT = 10; my $extApp = q[ perl -lwe"$|++; print $_ and sleep 1 for 1 .. $ARGV[0] +" ]; my @results :shared; my $pid :shared; async { $pid = open my $fh, "$extApp $N |" or die "$!, $^E"; push @results, $_ while <$fh>; }->detach; sleep 1 until $pid; sleep 1 while kill 0, $pid and $TIMEOUT--; kill 3, $pid and warn 'Command timed out' if $TIMEOUT; print "command return \n'@results'"; __END__ C:\test>timeoutcmd -N=1 command return '1 ' C:\test>timeoutcmd -N=10 command return '1 2 3 4 5 6 7 8 9 10 ' C:\test>timeoutcmd -N=11 Command timed out at C:\test\timeoutcmd.pl line 21. command return '1 2 3 4 5 6 7 8 9 10 11 '

    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".
    In the absence of evidence, opinion is indistinguishable from prejudice.

    The start of some sanity?