Thanks all, got many ideas from you.. My script took ~7 min to run with data file of ~50,000 lines with substr()/unpack() enabled and same script took ~2 min after disabling substr()/unpack(). That led me to the conclusion that substr/unpack is taking maximum of my time (and that I should reduce).
I have updated changes like using "@_" instead of "shift", with very small performance improvement. For "Devel::NYTProf", I need to check if my SA if he will allow me to install. Thanks again for your responses. Regards, Kavita :-) On Thu, Apr 11, 2013 at 12:01 AM, Rob Dixon <rob.di...@gmx.com> wrote: > On 10/04/2013 16:45, Bob goolsby wrote: > >> G'Mornin' Kavita -- >> >> Before you go off on a goose chase, how do you know that substr() is going >> to be a problem for you? Have you bench-marked it? If your file is as >> large as you say, I strongly suspect that your bottleneck is going to be >> I/O and any differences between unpack() and substr() will be lost in the >> noise band below 1%. >> > > Hint: > > Try adding > > return ""; > > as the first statement of `extractFieldValue` (i.e. so that it does > nothing and returns a null string) then run your program and see how > long it takes just doing the IO. > > Rob >
