Re: fastest pattern match

[neilwatson]

in reply to fastest pattern match

Read the file one line at a time. When you hit your match exit:

open( FILE, "file" ) || die ("The file not there ");

while (<>){
    if (m/$regex/) {
         print "matched!\n";
         close (FILE);
         exit;
    }
}

Neil Watson
watson-wilson.ca

[flounder99]

You should use m/$regex/o or the regex will be compiled every iteration, slowing things down considerably

--

flounder

[d4vis]

Good advice in general,
but in this case there'll only be one interation.

Update: Fastolfe's right...my bad...mea culpa

~monk d4vis
#!/usr/bin/fnord

[Fastolfe]

Why? If we were sure the match would always occur on the first line, why search for it at all?

[japhy]

Incorrect. It will not be compiled on every iteration, but its string form will be compared with its previous string form, to see whether it needs recompiling.

_____________________________________________________
Jeff[japhy]Pinyan: Perl, regex, and perl hacker, who'd like a job (NYC-area)
s++=END;++y(;-P)}y js++=;shajsj<++y(p-q)}?print:??;

[flounder99]

Thanks,

I remember reading that it still did the recompile even if the variable in the regex did not change. Of course I read that a LONG time ago. Apparently it has been fixed.(probably a long time ago) So would the /o faster by avoiding the extra string compare? I know this would probably be insignificant unless there where a bajillion iterations. But I was just curious.
Do you even know anything about regexes? (see sig)

--

flounder

Do you even know anything about perl?
--Anonymous Coward to Tom Christiansen on slashdot

[frankus]

That's JAN-AR-IK..(like generic) Just another Regexp Question.

Also it might be worth mentioning that meta characters in $regex will still be parsed. Make sure you intended that to happen.

--

Brother Frankus.

¤