[Linux dev] Re: Perl and matching several regex's

[Hendrik Visage <hvisage@xxxxxxxxxxxxxx>]

On Thu, Feb 20, 2003 at 03:22:23PM +0200, Oskar Pearson wrote:
> I normally do something like this:

> 	if ($_ =~ /somepattern/) {
> 		&processsomepattern($_);
> 	} else if ($_ =~ /otherpattern/) {
> 		&processotherpattern($_);
> 	} else {
> 		die("Unable to interpret line $lc");
> 	}

I was more thinking in terms of (but not perl syntactically correct yet)

for $i in keys %RegexHash
{
 if ($_ =~ $RegexHash[$i])
  then
   &process($i,$_);
   break
  fi
}

> Then I try and order the if statements in a way which
> is going to be the most efficient in terms of the number
> of likely matches.

The matches are of equal likenes. It's Radius Accounting log
records.

Hendrik