Before getting too enamored of your belief that we should offer corrections to the code in the OP, note that OP's script appears (to me, anyway) to hew closely to the example code in the Synopsis at http://search.cpan.org/~mfrankl/HTML-HTMLDoc-0.07/lib/HTML/HTMLDoc.pm ... but still fails.

That, alone, seems to me a good reason to suggest alternatives... especially when the offers came from respected Monks who assert some familiarity with how to achieve the bottom-line objective.

Then, seen the open bug list etc at CPAN RT.

And what you're assuming in regard to the "sample code" requests appears to be no better supported than the assumption (about authorship) which you (correctly) ascribe to me. Your line of thought hadn't occured to me, but upon mature possibly jejune reflection, I regard it as improbable. I may be wrong, but then, I'm not alone in being wrong (and in these cases, not on points open to varying interpretation):

"... despite you're accusations that he/she can't be bothered to use Google or rtfm."

My suggestion that OP read the docs was not, technically, an accusation. It certainly implies, or lends itself to the inference that I suspect OP has not followed up on any of the alternatives offered by other respondents, but -- narrowly -- is *NOT* an accusation.

OTOH, narrowly or otherwise, the "you're" in the quoted snippet should be "your."

Best regards.