substr() bug?

[NeedHelp]

I use Active Perl in windows to do my perl scripting and then migrate it to a Linux machine. I've found two bugs:

1) I parse a file and I read segments to get values. One segment looked like:

$stringValue = substr($n,64,20);

And with Active perl this worked great! So my testing worked great. I move this to production and everything worked great. Then one file I get some odd values. The substr was actually assigning more than 20 characters all of a sudden. I received around 40. I could change the 64 starting value to 60, 70, etc and it would always grab more than 20. I had to end up doing this to make it work in Linux:

$stringValue = substr($n, 64, 20);
$stringValue = substr($stringValue, 0, 20);

Very odd....

2) Again I'm reading segments and everything is fine in windows. I move it to Linux and have issues. I am using a substr:

$value = substr($n, 0, 256);

But I get wierd results in Linux so I had to modify to this:

$value = substr($n,0,128);
$value = $value . substr($n, 0, 128);

Can anyone explain this to me or tell me a fix? I'd like my test and production environment to work the same.

[KevinADC]

must be an issue with the data you are running through the substr() function. There is no differenc between substr on winodws or linux that I am aware of.

[NeedHelp]

Originally Posted by KevinADC

must be an issue with the data you are running through the substr() function. There is no differenc between substr on winodws or linux that I am aware of.

That is what I always thought and truthfully I expected it to work better on Linux instead of vise versa. If I can provide the sample data is anyone willing to try to repeat the bug?

[KevinADC]

I'll take a look.

[Jordan]

I will as well. I have various Linux machines I can test on and will install ActivePerl if you tell me the version you use.

[NeedHelp]

Ok, I'll post all the information to replicate this problem next week when I'm back at work.