Re: checking for infinite loops

2007-03-05 Thread Bill Luebkert
Chris Wagner wrote: > At 04:01 PM 3/4/2007 -0800, Bill Luebkert wrote: > >>What's the exact message ? I added $^E in code below which may help. > > > Using this format to get the most recent Windows error message is good for > clarity's sake. $^E is context sensitive while this form is unambig

Re: checking for infinite loops

2007-03-05 Thread Chris Wagner
At 04:01 PM 3/4/2007 -0800, Bill Luebkert wrote: >What's the exact message ? I added $^E in code below which may help. Using this format to get the most recent Windows error message is good for clarity's sake. $^E is context sensitive while this form is unambiguous. Win32::FormatMessage(Win32::

Re: checking for infinite loops

2007-03-04 Thread Bill Luebkert
Chris Rodriguez wrote: > > Hi all, > Thanks very much to everyone for this list-serve, and especially to Rob > and Bill for helping me so much with this. I'm basically using Bill's > code (below) and I have it doing what I want. So yay, mission > accomplished, basically, but two peculiar glit

Re: checking for infinite loops

2007-03-04 Thread Casteele/ShadowLord
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I'm not a perl expert by any means, but.. > possibility they may have infinite loops. I occasionally get an error > message reading: Win32::Process::Create infinite2.pl : at > Bills_code.pl: line 46. That refers to the line in the code that's > simila

Re: checking for infinite loops

2007-03-04 Thread Chris Rodriguez
Hi all, Thanks very much to everyone for this list-serve, and especially to Rob and Bill for helping me so much with this. I'm basically using Bill's code (below) and I have it doing what I want. So yay, mission accomplished, basically, but two peculiar glitches gum things up and

Re: checking for infinite loops

2007-02-01 Thread Sisyphus
- Original Message - From: Chris Rodriguez To: perl-win32-users@listserv.ActiveState.com Sent: Friday, February 02, 2007 8:05 AM Subject: Re: checking for infinite loops > Many thanks to Rob for doing this! > I do in fact have perl 5.6. I typed in Rob's modifications

Re: checking for infinite loops

2007-02-01 Thread Chris Rodriguez
Many thanks to Rob for doing this! I do in fact have perl 5.6. I typed in Rob's modifications to my code, hoping that perhaps all I needed was say, the "use warnings;" line of his. But it still failed. My error message includes "The Unsupported function alarm function is unimple

Re: checking for infinite loops

2007-01-28 Thread Sisyphus
- Original Message - From: Chris Rodriguez . . for($number = 1; $number < 5; $number +=1) { $ProgramName = "infinite" . $number . ".pl"; eval { local $SIG{ALRM} = sub { die "$ProgramName timed out\n" }; alarm 9; require $ProgramName or die "Require failed\n"; ala

Re: checking for infinite loops

2007-01-28 Thread Foo JH
I wonder if it's possible to kill a thread in the Win32 implementation of Perl? Chris Wagner wrote: > I would start by executing the sub program in a seperate thread. That way > the main thread can keep an eye on it and decide if it's run too long and > then do something about it. Another way t

Re: checking for infinite loops

2007-01-28 Thread Bill Luebkert
Chris Rodriguez wrote: > > If the alarm function now works on Windows, why doesn't this code give > me the results I expect? Might it matter than I'm using Windows ME? > What about my version of PERL? It's the one from the CD that came with > that book. 5.6 I think - I can check if it matte

Re: checking for infinite loops

2007-01-28 Thread Chris Rodriguez
Thanks everyone for the responses. They're a little over my head though. I was referred to this list by someone on the PERL beginners list-serve, since my problem is Windows related. But from what I can tell, this list doesn't seem to be very beginner-oriented. :) And honestly, the cont

Re: checking for infinite loops

2007-01-26 Thread Chris Wagner
At 05:39 PM 1/26/2007 -0800, Jan Dubois wrote: >It doesn't. It will not interrupt a blocking system call. For example >you cannot use alarm() to get a timeout for reading a socket, or reading >from a pipe etc. I take it that's a kernel limitation. Any word on sigalarm functionality in Vista?

Re: checking for infinite loops

2007-01-26 Thread Jan Dubois
On Fri, 26 Jan 2007 20:29:38 -0500, [EMAIL PROTECTED] (Chris Wagner) wrote: >At 11:54 AM 1/27/2007 +1100, Sisyphus wrote: >>The alarm() function now works on Windows: > >Yes but I believe it doesn't work in all circumstances. I don't remember >seeing anything about Windows having an all working a

Re: checking for infinite loops

2007-01-26 Thread Chris Wagner
At 11:54 AM 1/27/2007 +1100, Sisyphus wrote: >The alarm() function now works on Windows: Yes but I believe it doesn't work in all circumstances. I don't remember seeing anything about Windows having an all working alarm function. -- REMEMBER THE WORLD TRADE CENTER ---=< WTC 911 >=-- "

Re: checking for infinite loops

2007-01-26 Thread Sisyphus
- Original Message - From: Chris Rodriguez To: perl-win32-users@listserv.ActiveState.com Sent: Saturday, January 27, 2007 8:35 AM Subject: checking for infinite loops > The alarm function seemed to be the way to go - but not with Windows. The alarm() function now works on Wind

Re: checking for infinite loops

2007-01-26 Thread Chris Wagner
I would start by executing the sub program in a seperate thread. That way the main thread can keep an eye on it and decide if it's run too long and then do something about it. Another way to do it is with Win32::Process and then kill them after some time span. At 01:35 PM 1/26/2007 -0800, Chris

checking for infinite loops

2007-01-26 Thread Chris Rodriguez
Hello, I have a program that runs other programs (through require statements and what not). There's a problem in that some of these sub-programs may have infinite loops, which would cause the main program to hang. I'd like the main program to just detect such and go on to the next one.