I posted a similar solution here, but with one difference - the 0 was quoted.

At first glance, you might think "So what?", but in benchmarks it actually was 25% faster than the unquoted version.

I assume this is because perl has to do less work to figure out the context of 0 and transform it into a string.