This may be overkill, but for this I would use HTML::Mason to mark up your pages. It is designed for dynamic HTML generation ala PHP, with perl code being able to be inlined in your HTML.

It can be called from the command line, and a script to render out your documentation shouldn't be too tough.

Mason calls each HTML file a 'component', and components can have attributes. I would envisage your problem being solved by using the attribute tags to define what the file has in it, then have a list that goes through components looking for matching attributes to what you want.